VALUING PRIVATE COMPANIES:

A DEA APPROACH

Burcu Anadol

A Thesis submitted in conformity with the requirements for the

Degree of Masters of Applied Science
Graduate Department of Mechanical and Industrial Engineering
University of Toronto

@Copyrightby Burcu Anadol, 2000

National Library

Bibliotheque nationale
du Canada

Acquisitions and
Bibliographie Services

Acquisitions et
senrices bibliographiques

The author has granted a nonexclusive licence allowing the

National Library of Canada to
reproduce, loan, disniiute or sell
copies of this thesis in microform,
paper or electronic formats.

L'auteur a accord une licence non

exclusive permettant la
Bibliothque nationale du Canada de
reproduire, prter, distribuer ou
vendre des copies de cette thse sous
la forme de microfiche/film, de
reproduction sur papier ou sur fomat
lectronique.

The author retains ownership of the

copyright in this thesis. Neither the
thesis nor substantial extracts fiorn it
rnay be printed or otherwise
reproduced without the author's
permission.

L'auteur conserve la proprit du

droit d'auteur qui protge cette thse.
Ni la thse ni des extraits substantiels
de celle-ci ne doivent tre imprims
ou autrement reproduits sans son
autorisation.

Abstract
-

Valuing Private Companies: A DEA Approach

M.A.Sc. 2000
Burcu Anadol
Department of Mechanical and Industrial Engineering
University of Toronto

ABSTRACT
Traditionaily Company valuation methods are based on discounted cash tlows and
liquidation values, but aii have a number of known shortcomuigs. The application of

Data Envelopment Analysis (DEA) for finding comparable fims and predictng the
market values of companies is an extension of the market-based approach and replaces
the traditionai method where DEA is used tor assessing relative eEciencies.

For the inefficient units, the DEA-Valuaton analysis correctly classified the
market ranges for 70% of the companies. An upper bourid was predicted for the
inefficient companies based on their efficient peers. Of these, 75% were below their
upper bound. For the efficient peers, DEA was able to tind a lower bound for their
market value based on peers being compared to them 50% of the t h e . Although some of
the ranges specified by the peen are quite wide, the method still does a relatively good

job in c l a s s i m g market values.

VALUINGPRIVATECOMPANIES:
A DEA APPROACH

II

Dedk ation

- -

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

III

Acknowledgements

ACKNOWLEDGEMENTS
1 wish to sincerely thank rny supervisor Dr. Joseph C . Paradi for his advice,

guidance, and insighti comments throughout this research. It is a very wondemil group
he has set up here at the University ofToronto and 1 have had a fntastic time working at

the CMTE. The deepest appreciation also gocs to the other members of my cornmittee,
Professor McCabe and Professor Vicente. for their comments and suggestions.
1 would aiso like to thank Fai. Sandra. Heather. James, Robin, Taraneh. Kerry.

Zijiang, Paul. John, Ozren. Audrey and al1 the CMTE members for their support and
Inendship. Also many thanks to great friends. Elcin. Cigdem, Aviva, Nihan. Wendy, Joe,

Eric, Jot, Semih, al1 the Indy 9T8's, and to Def and Fun.
Very special thanks to my farnily and iiends. My father, mother and brother

Onur were of great support during this work and 1would not be here without them. With
the exception of the bout with pneumonia this has been a wonderhl experience.

VALUINGPR~VATE
COMPANIES:
A DEA APPROACH

Table of Contents

TABLE OF CONTENTS

LITERATURE REViEW ........................................................................................ 6

2.1. I

Intrnsic Vulue............~...~~~~.~~~~~~~.....~~..~~~~.~.~~.~~..~.
7

2.2 VALUEPURPOSES
.................................................................................................7

2-3 VALUATION
APPROACHES
..................................................................................

2.4 ASSETAPPROACH....................................
.
.
...............................................9
5

DISCOUNTED
CASHFLOW (DCF) .......................................................................... 10
75.1 DCF Pupers ..................................................................................................
2.1 2

Cupikdizution Rutes ......................................................................................

11
12

2 - 5 2 1 CAPM ....................................................................................................... 12

2.5.2.2 WACC .......................................................................................................13

2.5.2.3 Analyst Estimates of DCF Parameters ...................................................... 14

2 5.3 Di.su~untugesof the DCF M&thod.............................................................. I 4

2.6 PROBLEMS
WlTH ACCOUNTMG
RAT~OS
...................
.......

...........................15

2 - 6 1 Free Cu.di Flow (FCF) .................................................................................

16

2-7 DCF AND PRIVATECOMPANIES

..........,..................................................................

17

2.7.1 A Belu C'oe@cient for Privute corn punie.^.................................................... 1 7

FIRMS
OR THE MARKET-BASED
APPROACH
....................................19
2-8 COMPARABLE
2 - 8 1 P/ERurio .......................................................................................................

19

2.8.3 P/E Rurio und Compru6ie Firms ................................................................ -70

2.8.3

Vuhing IPOs Using CF................................................................................

-7-7

2.8.3.1 M/B Ratios .........................................................................................

- 2 5
-

VALUING PRIVATE COMPANIES: A

DEA APPROACH

Table of Contents
2-832 issues with this study ................................................................................ 26
Ls.sue.s with the CF Appruuch ........................................................................ 16

2.8.4

2.9 PRIVATE
AND PUBLIC
COMPANIES
.......................................................................... 27

2.9.1

Ls.sue.~in Vaiuing Privute Cumpunies........................................................... 17

2.9.2

A&mtrnenf.sto Privute Cornpuniesjor V'iuution pur pose.^......................... 29

INITIALPUBLICOFFERINGS
(IPOs) .................................................................... 31

2.1 O

2 O I Underprricing und Siapm2inng Fuctors............................................................ 31

2.1 0.2 Cr,rnpuny Chvmrship..................................................................................... 33
2.10.3 Isstres wirh 1PO.s............................................................................................

34

2.1 1

INTASGIBLE
ASSETVALUATION
........................................................................ 34

2.12

SCOPEAND L I ~ ~ I T A T IOF
O NTHIS
S STUDY...................................................... 35

...................................................... 37

DATA ENVELOPMENT ANALYSISe(DEA)

3.1 METHODOLOGY ......................................................................................... 3

3

3
3

.2 IDENTIFICATION
OF THE FACTORS
.......................................................................... 40
3

............................................................................
MATHEMATICAL
FOR~IULATION

41

3-3-1 The CCR Rurio Modd ................................................................................... 42

3.3.1.1 CCR Input Oriented .................................................................................. 43
3

. ~ . -2
1 CCR Output Onented .............................................................................. 43
-)

3.3.2

BC'C' Ratio LModd..........................................................................................

43

3 .3.2.1 BCC Input Orientation .............................................................................. 44

3 -32 . 2 BCC Output Orientation ................... .
.
.
.............................................45
DIFFEREYCES
BETWEEN THE BCC

AND THE CCR MODEL

..................................
.
46

INTERPRETAT[ONOF DEA ...................................................................................... 46

AND DISADVANTAGE OF DEA ........................................................ 47
ADVANTAGES

PRACTICAL
R E Q C I R E ~ IOF
~ TDEA
S .................................................................... 48
AND FIRMVALL'ATION
...........................................................49
DEA ANALYSIS

NOVELAPPLICXT
IONS OF DEA .................... .
.
.
.
............................................... 50
4

DEA-VALUATION MODEL FORMULATION

................................................

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

54

VI

Table of Contents
. . . . . . .

4-3 DEA MODELNG NEWDEA COMPANY

VALUAT~ON
MODELS
............................ 55

4.4 SELECTION
OF FACTORS
.......................................................................................

57

The Inputs......................................................................................................

58

44I

4.4.1.1 Total Assets & Total Liabilities ............................................................... 58

4-42

TheOutpntts...................................................................................................

59

4.1.2.1 Cash tlow fiom operation: ........................................................................ 59

4-4-22Profit and Revenue.................................................................................... 60
4.4.2.3 Book Value ............................................................................................... 60
4.4. 3

Vuriubkes Not Included ................................................................................. 61

1.5 FORMLLATION OF THE DEA MODEL ................................................................

63

..................................................................................................

67

5.

DATA ANALYSIS

5.4 CORRELATION
........................................................................................................73
5.5 PRIVATECOMPANIES
................... .-..................................................................... 74

......................................................................... 76
5.6 PL'BLICAND PRIVATECOMPANIES

ANALYSIS & DISCUSSION

................................................................................

78

OF COMPANIES
WlTH SIC CODE3600s ........................................ 79
6 1 BCC ANALYS~S

6.1.1

Summury S'utisticv ........................................................................................

6.1.2

.
Ineffrcient Cornpuny dnulysis ......... .............................................................. 79

79

6.1 .2.1 Subset Hammond .....................................................................................79

6 - 1 - 2 2Upper Bound for Subset Hammond.......................................................... 81

6.1.2.3 Subset 2 - Company Inverpower............................................................... 83

6.1.3 Efficient Cornpunies...................................................................................... 93

6.1.3.1 Lower bound for Efficient Company Camco............................................93

OF THE COMPANIES
WlTH SIC CODE3700s ..................... .
.
.
.
......... 99
6.2 ANALYSIS

..

E#zciency und Murket Vu1ue.......................................................................

IO0

6.2.1 Efliciency und Total As.se!s.........................................................................

101

6.7 1

6.2.3

Ciruph o j I. irtuul 0utput.s ............................................................................102

*

v-

VALUINGPRNATECOMPANIES:
A DEA APPROACH

VII

Table of Contents
6.7 4

I ~ u / ~ ~ u tof
i o!SC
n 3 700s...............................................................................

IO4

6.3 ANALYSE
OF THE COMPANIES
(S1C2000) ........................................................... 108

6.4.7

Efficient Cumpuny CYCPCUSI Cimpured IO d l 452 Cumpnies ................... I I 1

6.5 PRIVATECOMPANIES
........................................................................................... 111
..
63-I I>t@cient Privutr Cornpunies.....................................................................
112

6.3.2

EJlicient Privute Cornpunies....................................................................... 112

7. CONCLUSIONS & RECOMMENDATIONS

...................................................

7.1 CONTRIBUTIONS ...........................................................................................

113
115

7.2 LIMITATIONS OF THE DEA APPROACH .................................................... 115

7.3 A M A S OF FUTURE WORK............................................................................ 116

7.3.1 .Veie App2icotion.sfor DEA .......................................................................... 118

REFERENCES: ............................................................................................................ 119

BIBLIOGRAPHY......................................................................................................... 124
GLOSSARY .................................................................................................................. 126
LIST OF ABBREVIATIONS

......................................................................................

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

133

--

List of Tables

LIST OF TABLES
Tubk 2-1 VuiuutioniIpprouch - Advuntuges und Disudvuntages

Tubk 2-2 Desiruhility cf the Dzjj'ierent Vuluarion lCfethoJs.

Tctbk 7-3 Exumpk

of

Esrimuting the Discount Rute

I8

Tuble 5- I Finunciul Informution Sources

69

Tubk 5-2 Stufisrcul Testhgjiw the Public Compme.~

6Y

Tubk 5-4 Correlufion Beween the P u b k Company Dutu

73

Tuble 5-5 Privute Company Stuti.sficul Information

-4

Tuble 5-6 Correlution Mairii jbr the Priwte Cornpunies

-3--

Tuhk 5- 7 List of'the MC codes und the Ci,rre.sponding Nume of the i'ndus~ry

,'

Tuble 6- 1 Hurnrncmd Subse Stuti.stics

80

Tcrhle 6-2 Distunces to Hummoncl

XI

T~rhle6-3 - Subset Hurnmond und the Predicred lbfurket Values

82

und Distunce to other peers

T~lhie6-4 Cumpuny l n v e r p o ~ w

83

Tcrhle 6-5 Suhset lnverpower crnJ the Pmrlicted bfurket Valtres

83

Tuble 6-6 Inemcient DMUs und PrdicteJ Murke Vctlzres

86

Tuble 6- 7 Eficient Cornpuny Lirmco h c (# 7)

94

Tuhle 6-8 - Efficient DMUs & Nurnber of Peers (SIC 3600.s)

95

Tuhle 6-9 Eflcient DMUS SIC3 700s

Y9

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

List of Tables

Runge _for hlurket Vulues (SIC3700)

104

Runge for itfurket Vuiues -E@cient Cornp~nie.~~(~~IC2000)

1IO

Efficient-Rungefor :bfurket Vulues (Privute CTompunic'.s)

113

VALUINGPRIVATE
COMPANIES:A DEA APPROACH

List of Figures

LIST OF FIGURES
Figure 3-1 DEA Frontier

39

Figure 3-2 Projecting the Inefficient Unit w the Fmntier

42

Figtrrc. 3-3 Gruph of VIE und CRS

46

Fi~qrr3-4 Zhu S DEA ~bfodelfir

CUrnpany Performunce

51

FiCqre4- 1 DEA Vufuution~iModel

61

i - 1 Distribution oj-the Tottd .4sSs~t.sforthe Public Cutnpunies

70

i - 2 Distribution of-the Tord A.s.w~sjor the Cornpunies Under S I 00. U1!!1.000 - 71

Figure 5-3 Gruphicut Represenrtrrion of the Distribution of Marker Vulue:

Figure 5-4 Distribution Of ' t k A.s.~c.rsofthe Privute Compunies

7.5

Fsrre 6-1 Ineflcient Cornpunies t ~ n dtheir Predicted und Act ual Mwket Vulires - 88
Figure 6-2 InefJicirnt Cornpunies- Red MC versus Predicted Upper Bound

89

Figure 6-3 Ine-fficienl CompuniwPercenr Diflerence (Red dG Predicted Upper)

90

Figure 6-4 Efficieniy und Murker Vulue (SICj'600)

91

Figure 6-5 Efliciency und Tbrd rl.v.sets (SiC3600)

92

Figure 6-6 EfJic~ientCompunie.~-Red MC ver.su.v the Predicted Lotver Bound

Y7

Figure 6-7 E#icknr Cornpunies-Percent D13C.rt.nce ( R e d & Preikted Lower MI.] - 98

Fi,qire 6-8 Eflicknqv und Murkt V u h (SIC3700)

IOI

F i ~ p r e6-9 E#iciency und T ~ wAs.set.s

l
(SIC3700)

IO2

Figtire 6- 10 C'irturrl 0ufput.s ofCi,nzpunie.s ( X 3 700)

103

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

List of Figures

Fibwre 6-11 Range for Pre Jicted Murket Vutues und the Reul Vuiue (SIC3 7110) -,105

Figure 6-12 Inrflcient Compuny-Reui h Predicted Upper Market Vdue (SIC37OU) 1O 6

Fihwre 6-13 Efficient Firms-Red und Preclictrd Upper Boz1nd(SK3700)

--- - -

107

--

VALUINGPRIVATECOMPANIES:A DEA APPROACH

XII

List of Appendices

LIST OF APPENDICES
Appendix A Public Company Data - 1998
AppendLx B DEA R e d t s & Analysis of heflcient Cornpunies SIC 3600

135

145

-4ppendix C Anulysis of Eficient Conpunies SIC 3600

188

Append; D EBcient cmd Ineflcient Companies SIC 3 700

199

AppendUr E Eflcient and IneHcient Companies SIC 2000

203

Appenduc F Anulysis UMizing ail 452 Companies

207

Appendix G EBcient und Ineflcient Cornpunies SIC 2800

212

VALUCNG
PRIVATE
COMPANIES:
A DEA APPROACH

IX
II

Chapter 1 Introduction

1. INTRODUCTION
This chapter presents an overview of this thesis by providing background
information on the corporate valuation tield, and establishes the motivations for this
research. Detailed problem definition and research objectives are presented as well as

solution approaches and contributions. The chapter concludes with a summary of the
organization of the remaining document.

1.1

Background
Valuation arose out of the need for appnisals of land. its management. and use.

As a consequence, there have been multitudes of surveyor organizations in existence

throughout history, but in the recent years, as businesses have become more cornpiex, a

new f o m of valuation was required. Business valuation in Canada consisted of people

being trained by the provincial govemments to operate in what was known as the "death
duty field." This came about due to the need for additionai tax revenues. and gave rise to
the valuation of not only property. but of intangible, tangible and business interest items.
The Canadian Association of Business Valuators was formed in 1972 to organize the
ditierent protssions dealing in valuations under one umbreila organization.
The business of valuation is a multi-billion-dollar tield. In 1996. $650 billion

(USD) was invested in mergers and acquisitions (M&A). This figure was double that of
the peak M&A boom of the 1980s [ETT097]. The reasons for business valuations difkr.

but most businesses are appraised to obtain an irnpartiaI opinion on the value of the
business. Irnproper valuations c m lead to losses adding up to hundreds. thousands. and
o%enmillions ofdollars. Moreover. the range in prices that different buyers will oRer is

large. Ofien, the difference between ofirs may be 200 percent or more [DENNOO].

Chapter 1 Introduction

Company vaiuation studies iaclude the use of the asset-based approach, the
discounted cash flow method and the comparable f i m s approaches- In comparable t-ms.
the subject Company is benchmarked relative to specific comparable companies fiom
which usetl public vaiuation data can be obtained. When using discounted cash flows.
public companies' risWreturn characteristics are analyzed. adjusted and applied to the
subject Company.

I -2

Motivation
The motivation for this research stems fiom studies in several areas of corporate

finance and from applications of Daia Envelopment Analysis (DEA). The importance of

business valuation has received considerable attention over the past years. however. a
comprehensive review of using "comparable" firms has not been considered in the
present literature. For this reason, this work concentrates on tinding a peer group of
comparable tims using DEA; a unique application for this powertiil tool.

DEA is ideal tbr tinding comparable tirms in many ways. The basic underlying
methodology is a linear progamming based cornparison of the units under analysis. be it
b d s . hospital, or in this case businesses. DEA identifies best practice (efficiency

fiontier) units, aflows multiple inputs and outputs. is non-parametric. and deterministic
(does not reqitire the specification of a tnctional tom). DEA provides an efti~ciency

score for each DMU (Decision Making Unit), identities inefiicient units, h d s a peer
croup of similar companies for the units under analysis. and it projects the inefficient

units to the eficient kontier (targets for improvement). It incorporates al1 relevant
factors in a single modrl and no single factor is given prerence (non-biased). It

masures relative efficiency fiom the best-observed perhrrnance rather than fiom

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

Chapter 1 Introduction

averages. Also, it does not make judgments on the importance of the variables
(unrestricted weights).
This paper continues the approach started by S h a k [SIMAO], but diflrs in

several respects. First, the sample includes over 400 publicly traded Canadian tims and
over 40 private ones. The models for vaiuation differ with respect to the inpudoutput
variables- as these variables appear to be more appropnate to measure value. Simak does
not carry out the actuai analysis of the private cornpanies with DEA. so this work wiIl fi11
that gap- The data onsisted of known SIC (Standard Industrial Classitication) codes
with analysis carried out only for the companies that had the same 4-digit SIC code.
Further. one can propose to determine an upper bound on the market values of the
ineficient companies and a lower bound for the efficient companies as determined by the

Iarnbda values (resulting tiom the DEA model) found fkom the analysis.

1.3

Problem Definition
The p r i m q objective of this research is to value private companies ushg DEA.

A secondary objective is to tind the tctors that determine market capitalization for

companies and in particuiar. to tind the tirms that can be used for cornparison purposes in
a vaiuation study employing DEA. The first goal of the project is to develop a procedure

For vaiuing public companies using DEA by tinding comparable timis. hence. appropriate
market value ranges, and then to validate the ranges with the known market values of the

public companies. The nest goal is to appiy the procedures available in the public
domain tor valuing pnvate companies. Some of the most pressing issues in developing a
model are to detennine a set of relevant inputs and outputs that serves the objective and
ensuring that important variables are accounted for.

Chapter 1 Introduction

1.4

Thesis Contributions
The major contributions of this thesis are:

Development of appropriate vaiuation models for public and private companies.

Establishing a set of peer goups for the companies based on multiple inputs and
outputs,
For each individual company. efficiency scores and a difference indicator are
provided.

To gather a comprehensive data set consisting ofdata for 452 public companies and
49 private cornpanies-

Finding a market value range for the subject (inefficient) company based on its
comparative peers.
For the ineflicient companies, establish an upper bound on their market value based
on the eticient companies they are being compared with.
For the etticient companies, tnd a lower bound for their market value based on the
companies being compared to them.
To provide an additionai tool for business valuators and othea to utilize in order to
detennine the value of a company.

1.5

Thesis Oreanization
The remaining chapters of this thesis are organized as follows:

Chapter 2 provides a review of the relevant litenture in the area of vaiuation. It

identifies the different vduation methods and the shortcomings of each method. and
starts to identify why data envelopment analysis may be a good tool for this
application.
-

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

Chapter 1 Introduction

Chapter 3 explains the DEA solution approach. The development of the methodology
and the diKerent DEA models are presented. The compatibility with valuation is
discussed.
Chapter cl demonstrates the valuation mode1 ilhstrating the important variables to
utilize. [t demonstrates how each variable was selected and why certain ones were

not included.
Chapter 5 summarizes the data acquisition process and describes the data treatment
that had to be performed for the research.
Chapter 6 presents the results of the analysis and provides a discussion of the results.

An analysis of al1 the public companies with a SIC code in the 3600s is provided
using the BCC ( B d e r , Chames and Cooper) model. A complete analysis of
cornpanies having a SIC code in the 3700s and 2000s is examined including

examinations of conrlation's between efficiency and market value. A valuation

analysis of two of the companies with al1 452 companies is provided. Finally, the
lessons leamed from the public Company valuation are used to analyze three private
companies having SIC codes in the 3600s.
Chapter 7 presents the conclusions of the work and provides recommendations for
future work,

- -

VALUING PR~VATE
COMPANIES:
A DEA APPROACH

Chapter 2

Literature Rwiew

2. LITERATURE REVIEW
"ln pricing an acquisition, ive c m alwuys agree on the numhers. W%af vulue is

r m d ~ru be more in rhe reaim ofpoets thunfinanciutpeople" [1VEED86]

Valuation of private companies is a key area of tocus for accountants. lawyers.

business vaiuators, investment bankers. and venture capitalists- The purpose of a business
valuation is to fmd the financial worth of a Company. DEA's application to measure this
worth is a fairly new concept but the multidimensional nature of corporate vaiuation

makes it an ideai application area for DEA. This section highlights the literature related
to Company valuation and discusses the issues and limitations of the approaches.
The vaiuation concept can be dated back to the time of Napoleon and even King

Henry the VIII, although. its main purpose at that time was for inheritance taxation
[OVENiI]. As the business world developed. starting with the Industrial Revolution.
new methodologies for valuation became necessary. Present day valuations of IPOs

(Initial Public Offerngs) and srlling and acquirng businesses have taken centre stage in
the valuation world. but there is still a significant need to value tirms for estate. taation.
and divorce settlements. As there are many diffrent applications of vduations. so are
there methods to analyze them.

2.1

Value Definition
Value is the highest price a knowledgeable buyer. at armaslength. is willing to

pay in an open and unrestricted market. Pnce and value are ofien used interchangeably

but value generally means a willingness to sel1 or to buy. while price is the value at w-hich
a deal is carried out. The US. interna1 Revenue Service defines fair market value as
"the pnce at which a property would change hmds betwern a williny buyer and a willing

VALU~NG
PRIVATE
COMPANIES:
A DEA APPROACH

Chapter 2

Literature Review

seller when the tonner is under no compulsion to sell, with both parties having
reasonable knowledge of the facts" FRAT961.

2.1.1 lntrinsic Value

Intrinsic value of cornmon stock is the pnce that is justified for a share when the
primary factors of value are considered. In other words. it is the real worth ofthe stock,

as distinguished from the current market price of the stock. Tt is a subjective value in the
sense that the analyst must apply his or her own individual background, expertise and

s k i h to detemine it, and estimates ot'intrinsic value will Vary corn one analyst to
anothrr l [PRAT961.

2-2

Value Purnoses
The need for business valuation may arise fiom a variety of circumstances and the

diffierent circumstances wilI require diftrent valuations. Valuations of estate and divorce
settlements are quite different than valuations for IPOs. Here is a listing of the varying
needs for a valuation:
IPOs - the company intends to sel1 shares to the public
Mergers and Acquisitions (M&A)
Selling a Business
Estate and Gik Tax Purposes
Re-cripitalizing a Business

ESOP (Employee Stock Ownership Plans) Valuation

Litigrition Purposes or divorce scttlernents

' From the Handbovk t'or Financiai Decision Makers. John J. Hampton, Reston VA. Publishing Co., 1979
p543-1.

VALUINGPRIVATE
COMPANIES: A DEA APPROACH

Chapter 2

Literature Review

There are many diffrent methods that can be used to assess the value o t
business. but these methods can be grouped under three main approaches. namely. assetbased. income-based. and comparable 8 m s or market value based. The two main
approaches for valuation are the comparable firm approach and the income-based
approach. with discounted cash tlow being the most popular of the latter method. There
have been articles expouding the benetits of both methods. Whether one method is used
or several. each method has associateci benetits and disadvantages as illustrated in

Table 2-1 Valuatioa Approach Advsntages and Disadvanages

Approach
Comparable
tims ( CF)

Advantage
Works best when a highIy
comparable group is a\.ailable
Firmer theoretical fooring

DCF

L o o k at the undrrlyiny value

of a company's assets

Asset-based

1
(

Disadvantage
There is a chance that an entire sector is
under or over valued
Difiicult to estirnate future cash tlows
and an appropriate discount rate
More relevant when a signi-ifcant portion
of the assets can be liquidated readily.
For small. high-tech IPOs. this method
has littie relevance since it does not
consider gowth opportunities.
[BOOTOO]

The purpose for the valuation may intluence the method appiied. Of the diffrent
value approachrs the method most popular with companies is the DC F method. Booth
examined a survey conducted by Mohan et al on what valuation methods large conipiinies
use and recorded the results shown in Table 2-2 (based on a scale fiom 1-7, where a score
of 7 is the most desirable): [BOOTOO]

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

I
1

Chapter 2

Literature Review

Table 2-2 Desirability of the Diffetent Valuation ,Methods

iMETHOD

SCORE 1-7

DCF

6-1
5.7

MarketValueofComparable
Projected Earnings
Revenue hl ultiples
Earnings !Multiples
Book Value

2.4

4-4
4.1

4-0
3 -6

Asset A ~ ~ r o a c h
This rnethod is rnainly employed on companies that are asset intensive. for

example manutacturing and retail businesses. Arnong asset valuations, book value.
replacement cost. and Liquidation can detennine value. One of the most basic approaches
is to calculate the book value. hence the shareholders' equity. One of the problems of this
method is that the assets are based on historical costs. Another method that c m be used
relies on established benchmarks, which are used to estimate the value of a business. For
example. a small investment advisory sells for 3-4 times the current year's revenue.
Factors analyzed in asset valuation are the market value for tixed assets. leasehold
improvements. owners- benefits, goodwill (intellectual property. patents, trademarks,
copyrights. etc.) and inventory. The sum of these factors will give an rstimate of the
market value for the business.
Asset valuation is the least fvoured method and considered by some to be the
most difticult. Sometimes earnings have k e n "stored up tor a number of years ...to the
point at which two valuations are required, one for the operational side of the business
including sufiicient assets tor the maintenance of earnings. and the other for what has
become the investment side of the business" [OVEN59].

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

Chapter 2

2.5

Literature Review

Discounted Cash Flow IDCF)

The DCF method calculates the expected fiiture income Stream of a Company

discounted back to present value, at a risk-adjusted rate. It is based on the Gordon Price
Model and is mathematically represented as:

Where P is the price, D, is the dividend in year n. and k is the cost of equity or discount
rate. On the other hand. the Gordon Price Model assumes constant growth so the above
equation is adjusted accordingly. In this instance. g represents the growth, k and d are the
same as in the above equation:

Another way of representing DCF is to show that the DCF fonnuIation values
only the assets that give rise to the fiee cash flow:

Where T is the t h e penod, C represents the free cash flow, K is the discount rate. and V,
is the terminal value.
Price Earnings RatioIMultiple (PER)Formulation

Another variation of the Gordon Model called the Price Earnings Ratio/MultipIe

(PER) is based on forecasted eaniings [BOOT98a]:

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

10

Chapter 2

Literature Review

Where b is the %'s

retention rate of emings. ( 1-b) is the dividend payout rate, K is the

discount rate and g represents growth.

2.5.1 DCF PapeRuback @LJBA96]prefers the DCF approach fther than the comparable fimis
method. He investigated the valuation of 5 1 highly leveraged transactions. He showed
that the DCF valuation method produces more reliable estimates of market value. He

examined three CAPM-based approaches to estimate discount rates that correspond to

tirm-level. industry-level. and market level measures of risk. Al1 three methods
pertormed better when compared to the comparable tirms approach.

Kaplan and Ruback studied highly leveraged buyouts and re-capitalization's using
the DCF approach m L 9 5 ] . They found that a CAPM based DCF approach has
ripprosimately the same vaiuation accuracy as the CF approach. using the capitalization
ot- EBITDA?.
For most companies, tture eamings potential is more imporiant than past
performance, that is why calculating DCF is such an important issue in valuation. The

DCF method is used predominately for companies that are not asset intensive. like the
service sector. In the DCF method, certain variables and unknowns must be caiculated
which include: the amount of tture cash flow. growth potential of the cash tlow,
durarion of the tture cash Bow, risks associated with them. and the discount rate to apply
to arrive at the net present value F[CILL90]. Other factors to analyze inctude risk within
the industry. quality and size of the customer base. length of time the Company has been

in business. and the IeveI of technology.

EBlTDA - Eamings before interest. taxes, depreciation and amortization. also known as operathg cash
tlow
-

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

11

2.5.2 Capifalitation Rates

The most problematic aspect of this method is the selection of an appropriate

discount rate. For private companies, the discount rates are chosen based on the
perceived riskiness of the business. Risk is a part of any business- In 1776. from the
Wealth of Nations. Adam Smith remarked that "the ordinary rate of profit always rises
more or less with the risk" [BOOT98a]. A higher rate of return. or. a higher discount rate
is to be employed to value the risky firm's operations. Usuaily, the discount rate is a

percentage added to the current 90 day T-bill interest rate- For better-established
businesses the discount rate may be 5- 10% above the T-bill rate, and for riskier
companies about 2 0 3 % . The riskier the business the higher the discount rate. For
determining an appropriate discount rate several methods can be employed: they include

CAPM. WACC and analyst estimates.

2.3.2.1

CAPM

Markowitz. in 1959. analyzed the behaviour of a portfolio of securities and

showed t hat. "unless porthl ios were perf-ctlycorrelated, risk would decrease as

impertctly correlated securities were combined into portfolios" [BOOT98a]. Sharpe

added to this theory. addressing the issue of covariance, highlighting the measure of risk
known as a security's "beta coefticient and (it) is estimated by a simple least squares

regression of the security's return against that on the market portfolio" [BOOT98a]. Et
was Sharpe who introduced the Capital Asset Pricing Mode1 (CAPM). otherwise h o w n
as the market equilibrium model. which c m br used to derive the required rate of retum

for DCF calculations. It states that the required rate is equal to the risk free rate plus

VALUINGPRWATE
COMPANIES:
A DEA APPROACH

12

some sort of premium. It measures the nsk of a security by relating the expected return
on an asset to its beta:
K, = Rf+ (R, - Rf)P

Where. K, is equal to the cost of equity capital. Rr is the risk-tiee rate of retum R, is the
market required rate of return. and P is a measure of the systematic risk of an asset due to

the CO-movementwith the market, If a company has ri beta of one then the stock has the
same volatility as the market index it is being compared to. if beta is 0.5. than the
company is half as volatiie. and if it is two than the company is twice as volatile- In order
to estimate beta, one needs to regess the stock-s price against the level of some broad
market index such as the S&P 500 and the beta factor is just the slope of the regession
line.
2.5.2.2

WACC

Another method to calculate the capitalization rate is to use the weighted average
cost of capital ( WACC); it is represented mathematicaily as:

In the rquation. K, and & ( 1-T) are the costs of aquity and debt capital respectively,

where the cost of debt is on an after tax basis. The weights. E N and D N . are the equity
and debt tinancing proportions. WACC is the weightcd average of the costs of
borrowing and of selling shtlres. The weights are the fractions of totd capital that corne

tiom the diffrent sources. WACC is not reaily a separate method; it combines two or
more capitalization rates to gst one. It assumes that the purchase of a business c m be
partly financed with debt and partly with equity. It weighs the cost of capital to the cost
of indebtedness.
--

VALUING PRIVATE
COMPANIES: A DEA APPROACH

Chapter 2

Literature Review
--

2.5.2.3

Analyst Estimates of DCF Parameters

For the selection of the capitalization rate. vduators have intbrmation about the
capitalization rates of several companies, and they try to appiy these to the business under
consideration. The parameters they consider in a DCF approach are the time horizon, the
discount rate. and the cash tlow forecasts. Most malysts choose a time horizon of less
than 10 years and 55% use less than 7 yem. For the discount rate. 55% of analysts used
the acquiring tirrn-s WACC. and 4 1% used management judgment, Also. most analysts
apply industry growth trends and -'expecte&' pertormance to cash tlow forecasts

[BOOTOO].

2S.3 Disad'vantages of fbe DCF Method

The two major problems with the DCF procedure are the discount rate used and
the estimate of tture cash tlows. Eamings are effected by tluctuations within a certain
industry and by the state of the economy. Past earnings are no yarantee of Iture
earnings. Underlying tctors may be at play and each Company should be examined
thoroughly to insure the business is sound. In the case of Internet start-ups. no formula is
valid. Their characteristics are such that they usually have little or no emings. they do
not have historical earnings as they have existed for only a short time. most of their value

is in future growth opportunities, which are very uncertain, and their employees.
Problems manutcturing plants can have include:
Plant may become obsolete
Manufacturing and distribution outdated
Products sutliering due to cornpetition or lack of sales
Location might n o longer be suitable. closeness to market
Problems with employees
[OVEN59]

Chapter 2

Litenture Review

In 1959. according to a pamphlet put out by the CICA (Canadian Mitute of

Chartered Accountants), it was presumed that the capitalization of hture profits was not
tvidely used. Little proved that the opposite was true

CL ITTS91. He showed that it was

the most widely used methoci in Canada. He quoted Mr, Justice Douglas of the United

States Supreme Court. who urged caution when applying this method:
-'Since its application requires a prediction as to what will occur in the future,
an estimate, as distinguished from mathematical certitude, is al1 that can be

made. That estimate m u t be based on an uit'ormed judgment which ernbraces

al1 facts relevant to tture earning capacity and hence to present worth.
Including, of course. the nature and condition of the properties, the past
eamings record, and al1 circumstances which indicate whether or not that
record is a reliable criterion of hture pertormance" [LITT59].
The risk fiee rate (RFR) is usually acquired tkom business news, but it is usuaily
correct for only a short time period and it has to be adjusted to imitate the growth of the

trm for longer time horizons. The base CAPM ignores issues such as taxation and it
does not consider multiple time periods. It also does not model the real world. by
ignoring transaction costs and assuming al1 risky assets are traded. Selecting the

appropriate discount rate is a key issue in valuation. These and many other reasons make
it difticult to apply the DCF approach in trying to value a Company.

2.6

Problems with Accounting Ratios

There are also some problems associated with using only an accounting approach.

Each Company tiling an annual report has to adhere to their countrirs' Generally
Accepted Accounting Principles ( G A M L In Canada many of the larger public

Chapter 2

Literature Review

companies also. or exclusively. report according to U S GAAP. When carrying out a

complete va1uation.- the accounting nurnbers have to be adjusted for the diffrent
accounting procedures. Some accounting items to examine are whether the Company
accounts For inventory by LIFO or FIFO. In Canada most t?nns use FIFO because there
is an additional 3% inventoy write-off allowed for tax purposes [BOOT98]. Companies
using LIFO will have decreased prot-ts. reduced taxes but a stronger cash tlow since the

--

last item in is typically more expensive. Other issues to address include:

Purchase or pooIing ofgoodwll,
Capitalization or expense items (R&D, etc-)

Depreciation policy (accelerated. straight line, etc,)

Off balance sheet items-pension fund liability, w a m t i e s or litigation

Capital structure
Discounted rate differences (risk, time value of money)
For these reasons. when conducting a hl1 valuation it is important to adjust the

income statement and the balance sheet to reflect these concerns.

2.6.1 Free Cash FI0 w (FC9

Hunt [Hunt751 introduced the theory that the firm's capital expenditure (Capex),
needed to sustain operations. should be subtracted fiom the Cash FIow fiom Operations

(CFO) to get the m e cash generating ability of the firm. The result is what is now known
as Free Cash Flow (FCF). [t is mathematically represented as:

FCF = net income + non-cash items - increase Net Working Capital (NWC) Capex
Non-cash items include depreciation. goodwill. and dekrred income taxes and
afier

ta^

interest. FCF "captures whether or not the tirm actually generated cash tiom its

operations. afcr adjusting the eamings for non-cash items, the need to reinvest in NWC
VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

16

Chapter 2
-

Literature Review

and new capital expenditures to maintain the tum's opentions"~OOTOO]. However.

FCF's effects on asset poor companies and service organizations is an issue that has to be
deait with. In this sort of area it is more important to address the cash dollars spent by
these companies. such as h r advertising. R&D. planned losses. and so on.
According to Booth @5OOT00]. usinp FCF is one possible solution to the
accounting problem- By analyzing FCF. one c m get a true picture o f a company's
operating results. instead of the diluted view presrnted through the pure accounting
changes used by tirms.

2.7

DCF and Private Com~anies

How do al1 of these factors impact private companies? For estirnating risk

premiums one needs data and the data are only available in public markets. Booth
[BOOT98a] examined the eEect of size and transaction costs on the discount rate. He
looked at the total retum data on size based US portfolios going back from 1926 to 1995.
This information is available from the University of Chicago's Centre for Research into
Security Prices (CRSP). The data showed that smaller tirms constantly outperformed

larger ones. They eamed close to double that of the larger firms 22.62% versus 1 1.4 1%

respectively. Smaller tirms performing better than CAPM predictions are referred to as
the "small tirm efkct." Private companies are usually valutxi (discounted) on lower

PERS than public tirms are since they are considered to be riskier endeavon. Rules of
thurnb approximatr this discount to be around 20% for private companies.

2.7.1 A Beta Coefficient tbr Private Companies

A beta for the DCF of a pnvate tirm is impossible to acquire but Booth

[BOOT98a] suggested two alternatives. the pure play approach (PPA)and the

VALUING PRNATE
COMPANIES:
A DEA APPROACH

17

Chapter 2

Literature Review

instrumental variables approach. The PPA "tinds public companies with sirnilar
operations and use(s) their beta.. .(or) use(s) the broad industry classifications of the TSE

or S&P market indexes" [BOOT98a]. From this he found that -'a hin the general
industrial products area has ti beta of 1.2 or a CAPM risk premium of 4.M.8%. whereas
a pnvate retailer would have a CAPM equity risk premiurn of 2.7-3.l%" [BOOT98a],

The instrumental variables approach, pOOT98a1, "statisticaily estimates the

relationship between actual betas and the particiilar characteristics of tnded firms. such

as their h a n c i a l structure. growth rate. dividend policy. e t c . This relationship is then

used to approximate the beta for the private tirm by substituting back the private tirm's
t-mancial values. For the non-traded tirms, the variables act as instruments to estimate
their betas as if they were traded.
The potential discount tor a firm based on these tindings could be 15.5 to 17%. as
displayed in Table 2-3 :
Table 2-3 Example o f Estimating the Discount Rate

CAPM discount rate tor industrial products hm with beta of 1.2

Item
1 % Discount
Long Canada rate
5.5
Equities tax prerred in Canada
1
3 -5-4
Market risk premiurn in Canada
4
Small trm earning since 1926
Non-inarketability, transaction cost
1.5-2.5
15.5- 17
TOTAL
[BOOT98aj
Booth extended the CAPM to include private tirms. where the variables y1 and y2
are the small tirrn and pnvate tlrm premiums. This torces the discount rates of the small
firms to be higher than the larger tirms:

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

18

Chapter 2

2.8

Literature Review

Com~arableFirms or the Market-Based Approach

The comparable firms (CF) approach or market approach as it is sometimes

known. implies cornparison with publicly traded companies. It is implemented by

capitalizing the eamings per share (EPS) of the firm under consideration at the average or
median PIE ratio of the complementary publicly traded tnns. Earnings forecast. market
to book. and price to sales rire other factors that c m be used instead of P/E ratios-

2.8.1 P !Ratio

The P/E ratio is the relationship between the current pice of a stock to its
earnings. It shows the shares selling at so many times their actual or anticipated annual
earnings. The P/E is calcultitcd by dividing the current market pnce of comrnon stock by
the earnings per share. lnvestors use it to ascertain the market demand for stocks.

Companies with a high P/E have greater eaniings potentiat than low P/E stocks, but also
are more volatile and somewhat riskier investments.

PIE ratios enable one s hare to be compared with another. they also retlect the
views of thousands of investors on the quaiity of an issue. They indicate liquidity
earnings trend, profitability. and dividend payout dong with intangible hctors such as
quality of management. and prospects for the industry or segment [CSI89]. To rate the

P/E for a company's common shares with those of others. the companies must be
comparable which usuall y means they must be in the same industry. PIE ratios for
difirent industries are available in various media and industry sources. for example the

TSE monthly review.

VALUING

PRIVATE
COMPANIES: A DEA APPROACH

19

C hapter 2

Literature Review

The P/E ratio is probably the most usefui and widely used of ail financiai ratios
because it is. in k t , ail the other ratios combined into one figure since it represents the
ultimate evaiuation of a company and its shares by the investing public [CS189].
2.8.2 PIE Ratio and Comparable Finns
iMany studies examine the use of the P/E ratio for comparable ttirms. Boatsman

and Baskin [BOAT8 11 compared the accuracy of two diffrent types of P/Ernodels. The
tirst mode1 tinds a random company in the sarne industry. The second mode1 uses a
company h m the same industry but with similar ten-year average growth rate for
earnings. They found the accuracy of the second mode1 to be better.

Kim reported on an article by Alford F;IM99]. where Alford evaluated the

accuracy of the P/Evaiuation method when the comparable tirms are chosen by industry.

tirm size and earnings growth- He found that the 4-digit SIC code is the most effective in
choosing comparable t h s . It is also interesting to note that Alford did not find
improved accuracy when size and industry membership were used togethcr in the PE
valut ion [KIM99].

Kim cited a study by De Ange10 w M 9 9 ] that demonstrated that investrnent

bankers make extensive use o f accounting data and that they depend on the CF approach.

He did this by examining working papers and timess opinions on management buyouts.
He showed that undenvriters base their pricing decisions on an analysis of the market
price ratios. with adjustments for fimi-specific ditrerenes. and determine a range of
minimum and maximum ottring price, then, after gthering more information they set
the market price.

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

20

Chapter 2

Literature Review

Ratios other than P/E can be used in conjunction with the CF approach. One can
have any sort of ratio. even those that are specific to an industry, for example. market
value divided by cable subscriber. Amir and Lev. cited fiom [KIM99], provided a
valuation study of the wireless communication industry. They showed that the primary
determinant ofcompany value is the population of the fianchise temtory. rather than
financiai variables.
The value relevance of P/Eand market to book (ME3) ratios has been
demonstrated by empiricai studies. Kun quotes two articles. one by Zerowin and the
other by Liu and Ziebert [iUM99], who examined cross-sectional variability in P/E and
found a significant relationship between P/E and growth, dividend payment and size.
According to Campbell "trading statistics with respect to shares of so-called
comparable public companies cm be. and moreover should be, a sipiticant factor to
consider" [CAMP91]. He stated that it is important to use CF where the privately held
business under valuation is similar in size. business structure and business operations to
one or more of the publicly traded companies.

In the US.. the Interna1 Revenue Service (IRS) considers the analysis of public
companies an important tool in determining the value of private companies. although this
is mainly tor income-tau purposes. Ruling 59-60 urges "using cornparisons fiom
quotations h u n d in the prices at whih similar companies are selling in tiec and open
markets. including not only listed securities but also actively traded over-the-counter and
unlisted securities" [OVEN59].
Moms [MORRV] reatirrned the tct that the best multiple may corne from a
public market comparable. If it exists it should be given full and cornplete considention.

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

21

Chapter 2

Literature Review

He points out "it is diffTcult to depart from price. or multiple, set in the public
marketplace. These public companies are one's best empuical data" F[ORR91].

2-8.3 Valuing lPOs Using CF

Systematic studies on the usefulness of the CF approach have not been pertbmed.
Kim and Ritter WM991 come closest to tilling this gap with their significant paper on

IPO valuation. A detailed explanation of [POs is discussed in a later section. Most

litenture recommends the use of accounting data in conjunction with comparable firm
multiples when valuing IPOs. Generally. younger firms Vary in terms of ratios, P/E,

WB. and price to sales, and may need further adjubtments. The best way to employ PIE
ratios is through the use oF forecasted eamings rather than trailing eamings. Another
ratio that can be used in valuations is enterprise value to sales and enterprise value to
operating cash tlow. Enterprise value is equal to the market value of equity plus the book
value ofdebt minus cash. "Ratios seem more usetl especially when adjusted to reflect

dit'trences between pro titability and growth rates of the [PO in question and the
comparable firms use&' [KIM99]. However. it is dificult to forecast the future cash flow
for very young companies, which is the category most IPOs faIl into. That is one of the
main reasons why the DCF method is imprecise, and the key advantage of the
comparable tirms method. The CF method is supported by academics and practitioners
and widely used in case studies in business schools [KIM99].

Kim and Ritter [KIM991 use sales. eamings. operating cash tlow and book value
to determine value. as well as comparable firm multiples. They tested the accuracy of the
comparable firms method by using two sets ~Ccomparablefirms. The tirst set consisted
of recent IPOs in the same industry and the second set was chosen by a research

--

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

22

Chapter 2

Literature Review
-

- -

boutique3. The results showed that the second set worked better than the fim, and
applying forecasted earnings worked better than using historical eamings. They also
tested the hypothesis that young t i m s are harder to value by splitting the data into 'age'
goups and applying the second set. They found that the -taluation errors ofthe
comparable firm multiples are noticeably smaller for the older t i m s than for the younger
rms. especiaily when using emings" wM99].
The recent IPOs set consists of companies that went public not more than 12

months prior to the IPO under investigation's ot'fer date and that has the same 4-digit SIC
code. if there were more than 5 comparable tirms to compare with they suggested using

the 5 tirms with the closest sales and ignored the rest (firm = recent [POs). Earnings for
the tirst year after going public typically include substantial amounts of interest incorne
thus in the analysis EPS is used and sales numbers tiom prospectuses only, rather than
more recent financial statements that may be skewed by interest income.
The variables thry sekcted were offer date (cannot be more than 12 months

before the offer date of the new IPO), EPS, Sales (13 month revenues prior to IPO as
stated in prospectus). and book value per share (BPS) (both pre and post IPO). Another
condition placed on the variables is that they must be positive. They then computed the
following ratios: PIE. P/S, and W B . The median of these market multiples is then used
as the CF multiple for the Company. The ompany's ratios are computed based on its

ofter price (post-issue).

CF multiples are typically higher than the multiples for the tirms going public.
partly because they are using market pices for the comparable tirms and oftkr prices for
the IPOs. Also, because of the short-run underpricing phenomenon (average tirst-day
'Company that specializes in valuing l POs

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

23

Chapter 2

Literature Review

retum in the sample is 12%). one would cxpect the [POmultiples to be discounted by
about 12% wM993. The fimctional fixation hypothesis asserts that the market
mechanically capitalizes reported EPS numbers. without adjusting for the quality and or
persistence of the eamhgs. Friedlan et al IIUM991 suggested that the market does not
h l l y incorporate the inibnnation content ot'discretionary accruals in valuing new issues,
consistent with the fiinctioniil fixation hypothesis,
The tirst mode1 of the CF approach is surprisingly weak for the tollowing reasons.

First. because historical eamings of IPOs are transitory in nature they have little value
relevance. Second, using market multiples and past data have an intnnsic limitation in
that they do not retlect future growth. This leads to the third issue, which is allowing for
adjustment of diEerences in growth and protitability. Finally, there is an element where
the comparable firms may have been chosen improperly w M 9 9 J .

LI-'boutiquey firm. Renaissance Capital. chose the second set of comparable

tims. The research boutique chose the CFs based on the major cornpetitor tirms
mentioned in the prospectus of the [PO. While historical earnings contained in the
prospectus are available to al1 market participants. practitioners aiso tiequently use
eamings forecasts for valuation purposes. tn addition, Renaissance Capital did not
restrict the comparable t i m s to the s m e SIC codes.

Kim and Ritter used --the geometri mean of the Renaissance Capital comparable
firm PIE multiples as the explanatory variable in regressions using the three P/E ratios of
the IPOs as the dependent variables" [KIM99]. The three P/Eratios. using the midpoint

of the POP (preliminary oEer price) for the IPO. are the last 13 months. current fiscal
year's forecast. and next year's forecast of the EPS. Comparing the regressions, the

Chapter 2

Literature Review

average prediction error fdls from 55% to 43-7% to 28.5% respectively, and the
percentage of rms that are valued within 15% of the actual multiple increases. Younger

firm valuation using forecasted eamings works beer than using historical earnings. as
there is a potential ~ o w t opportunity
h
that needs to be addressed,
They also tcsted the presumption that older t-urns will be easier to value since
capitalized earnings represent more of their value than expectations about ture growth
rates. Consistent with that presumption. the mean absolute prediction error t'or the twms

that are less than 10 years old at the time they go public is higher than tor older fimis.
3 1.9% and 23% respectively.
2.8.3.1

LM/BRatios

Renaissance Capital does not utilize iM/B ratios- They find 1WB ratios to be poor
valuation metrics. They daim that M/B ratios have a disadvantage beause the book
value is arbitra-; for IPOs since there is a large change fkom before the issue to aier
(intlux of new capital tiom new shareholden). Rather, they recommend using enterprise
value (EV) where they look at market value to sales. EV to sales and EV to operating
cash tlow (EBITDA) in their andysis-

To sumniarize Kim and Ritter [KIM99], the variation in these ntios is JO large,
both For public tirms and IPOs. that they have only modest predictive value. Many
idiosyncratic factors are not caphired by industry multiples unlrss various adjustments for
diffrences in grow~hand protitability are made. In practice. analysts place more weight
on a given multiple for some industries than others. Taking this into account would

probably show that CF multiples result in more accurate valuations. In practicr. one
starts with industry multiples and adds or subtracts adjustments of 10-10% to retlect

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

25

Chapter 2

Literature Review

ditierences in growth rates. protitability. and quaiity of eaniings. Investment bankers use
the midpoint of the oflr pnce range resulting in smaller prediction errors than using
comparabIe tirms. They canvas the market place tor demand before setting the final
issue price N M 9 9 1 .

2.8.3.2

Issues with this study

One issue with this study is the use of a maximum of 5 comparable t i m s and the
deletion of the rest. The comparable tirms approach works best with a hi& number of
comparable Lims and exclusion of tirms to simpliQ a study may not give meaningfl
results. tt is better to have as many comparable companies as possible.
2.8.4 Issues wit the

CF Approach

T h e e main issues with this ripproach are the P/Eratios employed, the use of the
right multiplier for the industry. and the availability of a highly comparable group of
companies. PIE ratios are dynamic and Vary widely over time and among companies.

The problems with PIE ratios and the right multiplier c m be remedied by looking at
industry trends and the broader market to forecast trends and adjust historicai P/E ratios
accordingly. Companies can be public and not usell for these comparisons because:

They are not listed anywhere (traded over the counter)

Not liquid
Shares are priced very high (Hathaway Corp.)
Share prices are too low because of too many shares
Small tloat

Huge stock options outstanding at low prices compared to market

Only tw very large tirms are listed. private t i m s are one or two orders of magnitude
smaller.

--

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

26

--

Chapter 2

2.9

Literature Review

Private and Public Conmanies

In order to apply the above valwtion techniques to private %ms. adjustments

have to be made to account for the differences that exist between public and private
companies. One of the issues to keep in mind is that private companies are not as well
established. and most of their value lies in their growth opportunities. Exceptions to this
include Goldman Sachs and Eritons, which were both very large and well established
betore their respective IPOs, Private companies have kss than 50 shareholders and their
shares are hiyhly concentrated, leading to liquidity risk oncems. Private companies are
under no obligation to report their earnings or to produce annual reports and hence their
value is often times subjectiveOn the other hand, information on public companies is readily available. through
t-mancial statements? annual reports. and such. Public markets rire organized places of
exchange- their p k e s are inkluenced by coverage in thc press and by analysts, and there
is more intorrnation avaiiable about them. When compared to private firms. they are:

Typicaily (but not always) larger:

They tend to participate in more than one industry (this leds to divenified risk):

They over a broader geographical area;

They have rules and reptations goveming them (BOD):
They have greater liquidity :
They have greater access to financial resources and funding sources:

They usually have better developed products. management trams. and strategirs

2.9.1 ksues in Valuing Pnvate Companies

When pricing pnvately held business interests the analyst nreds to look at dl
available data. The important issues are:

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

27

Chapter 2

Literature Review

The nature and history of the business

The industry's economic outlook
Financial condition of the business
Earnings capacity
Intangible assets
Market prices of stocks of corporations engaged in the same or similar businesses
that have stocks actively tr&d in ffee and open markets
Other issues to examine include the reasons for the sale. Are there large capital
expendinire requirements, changine market conditions, increased cornpetition or working
capital shortages? In any cax. a strict review of the h a n c i a l statements is necessav.
Redundant assets4 (RA) should be separated tiom the operating assets and either
excluded from the negotiations or sold with the proceeds distributed to the owners.
Consequences of not dealing with RAS rnay mean that the -'vendor does not receive full
value when the business is sold. The purchaser may have an opportunity of disposing of
the RASand applying the net proceeds to the business, either in its operations or as a
redution of the acquisition drbt incurred' [HEND88]. As well. real estate assets should
be separated edom operating assets. Returns on real estate may be less or more than the
business return but seldom the same and therebre. may not hlly be taken into account in
pncing the business. Moreover. in businesses with a long history. many oftheir real
estate ssets are sigiificantiy undervalued because they were bought for much less than
they are worth now and evaluating thrm at market is otien not possible [HEND88].
Information packages must be prepared containing the qualitative and quantitative
information about the pnvate Company. These are listed in Table 2-4.

VALUING
PRIVATE
COMPANIES:
A DEA APPROACH

Chapter 2

Literature Review

Table 2 3 tnformation Requirements on Private Companies for Valuatioa

History of the company

. Description. of current operations
S ummary ot- financial periormance
Key tinanciaI assumptions
Fixed assets description

1 Reasons for selling-

Products/services
Corporate profile
Customer pro tile
Or~anizationaichart

Opportuni ties
Market analysis
Cornpetition
Supplier pro tile
Ernrrlovee base

2.9.2 Adjustments to Private Companies for Valuation Purposes

Wise examined the discounts that apply when valuing shcues ofprivate
corporations [WSE90]. He observed tctors that could afict the outcome of a
valuation. This included portfolio, dis-synergy. key-men. and income tax discount.
Portfolio discount d e d s with the tct that while the "value of each property may
represent. on an individual basis, the highest price obtainable. the m k and quantity of the
property may result in the bundle en bloc k i n g less attractive to a purchaser than each
one viewed individually" [WISE90]. Thus, a discount to the price has to apply. Synergy
is a benetit expected fiom a purchase, for example. economies o f scale, reduction or

elimination of competition, or increase in market share. Dis-synergy discount is the

opposite of this. and would account for the disadvantayes that corne from increasing in
size. suh as increased operations, management, and such. Key-men discount recognizes
that there are some employees that are vital to the business. and this value, sometimes 25-

15% of the price. must be discounted in arriving at the ir market value of the business.
Finally. income-ta.. discount considers the impact on share valuation of axes and takes
into account the recapture of tau depreciation, capital gains. and tax on proceeds of
intangibles.

Redundant assets are assets owned by the Company that are not required in its day to day operations,
examples include propertirs, investment in subsidiaries that rire not of the core business, and investment
porttolios. [HEND88]

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

29

Chapter 2

Literature Review

Weaver looked at discounts caused by lack of control and lack of liquidity and
marketability F A V 9 8 1 . Perception dictates that risk is less if the investor has the ri@
to control the business- An adjustment to the pnce of a business has to be made for lack
ofcontrol. He quoted many studies that try to determine an appropriate amount for this
rate. The important issue here is that a rate actually sxists but quoting values fiom other
studies is futile; each case is unique and must be exmined individually to arrive at the
discount rate. Studies included Lyons and Wilczynski KYON891 who found the range to
be 27-28%. This amount was tound fiom premiums paid in M&A tiom 1968- 1987.
Pratt IPRAT931 studied takeovers of companies and the premium paid by the company
taking over. The premium is converted to a percentage that acts as an estimate for lack of
control. and is directly abstracted tiom actual market transactions. The discount level
was found to be 26-33%. Weaver noted that it is "interesting to see this level of discount

(that) exists even for SEC regulated companies where minority shareholders have more
WAV981.
protection than with small partnership~'~
Lack of marketabiky and liquidity is concerned with the lack of a real market for
the shares and the fact that investors pretr liquidity. Again. quoting many studies. one
can conclude that this rate exists b u t it is specific for each situation and there is no easy

formula to apply. ChaTe [CHAF93] used Options Pricing Theory to tind this amount.
He applied a European option. which results in a Iower price than using Amcrican
options. and showed a minimum applicable discount. AppIying the Black-Scholes Put

Formula the amount is 28-4 1% when the marketability is constrained for 2 years. It is
32-49% using 4 years. Srnriller trms with weaker earnings trade at much higher

discounts than larger tinns- Management Planning Inc. a valuation company, prescnbed

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

30

--

Chapter 2

--

Literature Review

live factors that are predictive o f restricted stock discounts. The more stablehigher these
variables are the lower the market discount @4ERC97]:
Revenue
Recent earnings
Market pnce per share
Price stability
Eamings stability

2.10

Initial Public Offerings (IPOs)

An initiai public otfering is a company's first sale of stock to the public-

Companies generally issue an IPO to raise money and to increase their f i m i a i base and
ability to make acquisitions. Most IPO literature recommends that to pnce an IPO one
should initiaMy look at and compare its operational and tinanciai performance with that of
several public companies in the same or a similar industry. In order to compare private
manutacturing companies one shouid look at the industry as a whole. and then at the
specitic 4-digit SIC Code. For companies going public. the factors that id--uencepnce
include the ownership retention in the business. and the risks associated with the POs.
Al1 this information should be applied back into and included in a complete valuation
procedure.

2. 10.1 Underpricing and Signaling Factors

There is an extensive amount of literature on the underpricing phenomenon of

IPOs and their average first day retum and the s i g n a h g tctors associated with them.
Signaling factors are issues such as percentage retention of stock by the original
owner(s). A number of studies tind that IPOs espenence large increases in value dunng
tiieir tira day of trading. Ibbotson et al PBB0941 tound an average retum of 16%, while
others have tound 1.9 2% [LECr195]. Most found that IPOs underperhrm by the end of

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

31

Chapter 2 Literature Review

-

their t i m year or within 3 yean [LEVI95]. Underpricing is due to unintormed investors

and the uncertainty surrounding the true value of the issue. The underpricing can be
explained by the 'winner's curse,' that is, smaller less informed nvestors suffer winner's
curse because they receive a smailer fraction of the coveted issues and more of the
unpopular issues [IB B0941. Stoughton [STOU98] found a relation between the fraction
of shares owned by large investors and Company performance. He concluded that the

value o f a h n ' s [PO is detemined by its ownership structure. If larger investors buy
stakes in the company the offering price and revenues raised are higher- Firth also tound
a positive ownership retention signal [FIRT97]. Jan [JAiN99]. on the other hand.

investigated whether the performance of IPOs is improved by lead investment bank

monitoring He found that bank reputation is positively correlated with post-issue
pertomance of IPO f'irms. I-ie used a Carter and Manaster [JAN991 study on ranking
undenvriters according to reputation. The "reputation proxy is based on the relative
placement of underwriters in tombstone advertisements over the period of 1979-1984'[JAIN99]. It is d s o important to investigate if the underwriters have warrants in the

company as this can inf'2uence an increase in monitoring. Warrant holders are tied to the
long-term pertrmanee of the company, and will want to insure that the company is doing
well,

[LJUN97] investigated the underpricing of IPOs. He found that stock market

retums. the macroeconomic climate, ownerstup retention. and inverse ofter size have a
positive efict on underpricing. He also showed that German IPOs experience losses o f

12% over the tirst three years of trading when compared with the market (exclusive o f
the initial underpricing retum).

VALUING

PRIVATE
COMPANIES:
A DEA APPROACH

32

Chapter 2
-

Literature Review
-- --

-- -

w M 9 5 ] investigated the information contained in the prospectuses for new

issues and found important fuiancial variables that affect the pricing of IPOs. These
variables included EPS, offer size, offer type. and industry-wide prospects. [BYRN94]
fiund initial returns are reiated to tirm size. method of issue. the market in which the
issue is made, and the concurrent retum on the market index. The average retums can be

in part modeled by P/Eratios.

[SEGU97] found a higher rate of mortality for lower priced (cg.penny stocks)
stocks than r higher pnced securities and that investors were not properly compensated
for this risk. However, mortality is not linked to market capitalization. Market

capitalization acts as a discriminatory variable for retunis. volatility and Iiquidity. if

Iower ot'fering pnces signai a geater degree of Iower returns. he specuiated about why
undenniters choose to signal this lower risk. -'If an underwriter intends to make a market

in an issue following an [POI both spread and commission revenues could be enhanced
by setting a lower offer price. It is at least conceivable that the increase in revenues due
to a lower ofir price would be sufficient in magnitude to outweigh the additional

marketing costs inherent in signahg lower quality" [SEGU97].

2.1O. 2 Company Ownership

One other issue to note, there have been dozens of articles on the pertormance of
companies that go public and the role of management ownership. Jensen and Meckling
CJEXS761claimed that the interests of managers and other stockholders become less
closely aligned as managers' stakes decrcrise and ownership becomes more dispersed.
They documented that the ownership stalie of ofticers declines sipifcanttly from the year
previous to going public to ten years later. They also observed that the median operating

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

33

Chapter 2

Liaerature Review
--

return on assets declines tiom the year previous to going public to the end of the first year
of public trading, but thereat'ter. performance does not decline in ten years. However.
they found that oprnting pertbrmance both within one year of the offerng and during the

tirst ten years of public trading is unrelated to ownership by officers.

When specitically dealing with IPOs of private companies. insiders have better
information about espected values. Ownership retained by insiders is used as a signaling
mode1 where the higher the percentage of retained shares, the higher the market

valuations [RITTW].

[KLEI961, and [LELA77].

2.10.3 Issues with lPOs

It is important to note that there is hidden value in receiving an influx of

funds/cash, either in the process of the IPO, or in a secondary oftring. For example.
when the company rrcrives tnds, they may now be able to attract better supplier
commitments that wrre not possible before the addition of equity. Alternately. additionai
tnds could change the financial nsk of the company and irnprove its balance sheet.

2.1 1

Intangible Asset Valuation

Valuation is not only about h e d assets. Mullen looked at valuation of intangible

assets such as brand names, trademarks, and patents [MULL90]. Intangible assets are
valued j ust like regular assets: find the net present value of the future cash tlows for the

intangible assrts. He suggested using the market transaction comparative method to

derive multiples of sales or eamings tiom sales of acquisitions of similar assets. and
applying these multiples. atier making appropriate adjustments. to the sales or eamings of
the asset to be valued.

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

34

Chapter 2

Litemture Review
-

- -

The cost of a well-recognized brand name can add billions of dollars to a

company's worth. The most reco~rnizablename in the world Coca-Cola is estimated to
be worth $72 billion followed by Microsoft at $70 billion [CBCOO].
Valuation Summary
To summarze, there are many diffrent methods to conduct a valuation. but each
valuator or person who values companies is quick to point out that a combination of these
methocis works best. Theretore. no study is done in isolation. different methods are
applied and a range o f values reached. as supported by W L L 9 0 1 . FITC881, and

[OVEN59].
Although the market-based approach tinds comparable

it is not a

cornprehensive search. the DEA-Valuat ion method will fil1 this gap.

2.12

S c o ~ and
e Limitations of this Studv
Al1 these factors should be taken into account when conducting a private company

vaiuation. However. al1 this information is not readily available and only applies to a
complete valuation study. As well, the earnings outlook and financial conditions of the
pnvate company are not known. other than what can be interpreted from the financials.

This fr exceeds the scope and premises of this study. This study is intended to be a
guide for valuaton and should be considered as another tool they have at their disposal.
Use of the DEA-CF method is to determine the best and most comparable public
companies. Calculation of the market value is attempted by considering the lambda
values to be a percentage, and multiplying this by the known market values in order to

tind unknown market values and also by infemng a market value range from the

VALU~NG
PRIVATECOMPANIES:
A DEA APPROACH

35

Chapter 2

. - -

Literature Review

groupings. The methodology is discussed in detail in the anaiysis and discussion section.
Another application of this approach could be to first, fmd the similar companies then,
instead of trying to calculate the market value, the known multiples/earnings of the public
companies can br applied to the pnvate companies. Thus one c m project the earnings of
the private companies using the multiples from similar public companies- Again this is

information beyond the scope of this study but it is an issue to consider in a fture
research project. The valuzition approaches e x h e d and the andysis conducted herein

are intended mainly ror companies that are being acquired or sold and for P O s -

Valuation of prokssional practices, dentist oEices, travel agencies and such, is not
included in this analysis as more ofien than not, rules of thumb procedure is applied to
these businesses.

VALUMG PRIVATECOMPANIES:
A DEA APPROACH

Chapter 3

--

--

Data Envelopment Anaiysis

3. DATA ENVELOPMENT ANALYSIS (DEA)

Data Envelopment Analysis (DEA) is a linear programming technique for
calculating the relative performance of a population of operatng entities referred to as
decision-making units (DMUs). This technique considers multiple inputs and outputs of
DMUs simultaneously. DEA handles multiple inputs and outputs by measuring the
efiiciency of a DMU as the ratio o f a weighted sum ofoutputs to a weighted s u m of
inputs. subject to the constraint that the ratio is less than or equal to unity. Essentially.

DEA -'compareso' the int'ormation rivailable from a group of DMUs with al1 others in the
group optimizing each individual DMU with the objective of calculating a discrete.
piece-wise linear, &%estpractice" frontier.
The origin of DEA is tiom a Ph-D. dissertation done in 1978 by Rhodes
[CHAR781 supervised by Cooper and Charnes. In this dissertation, Rhodes exarnined the
efficiency of educational institutions in the New York City district with regards to their
abiIity to improve the developrnent of disadvantaged children. The resulting model.

named CCR after the three authors, allows calculation of relative efficiencies on the units
under investigation based on a constant-retums-to-scale mode[. The CCR model of DEA
was an extension of Farrell's single outputlinput technicd efficiency rneasure [FARR57].

Following the CCR model was the introduction of the BCC model in 1984. by Banker.
C h m e s . and Cooper [BANK84]. The BCC model contains an added constraint of
convexity. which allows DMUs to be measured on a variable-retums-to-scale basis.
In this way, DEA's original design was to measure the relative eficieny within
not-for-protit organizations. Through the yeae. DEA has been expanded to sni- the
eficiency of over 50 industries inciuding banking, tst food retail. healthcare. insurance.

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

37

Chapter 3

Data Envelopment Anabsis

--

~reditunions. highway road maintenance. and capital budgeting projects. Over 1.300
papers have been written by academics on a wide variety of theoretical and pncticd
advances in DEA. DEA can be considered to be a combination of econornics, statistics.
engineering, operations research. and management techniques.

3.1

METHODOLOGY
DEA is a t o m of frontier analysis that recognizes that some DMUs pertiirrn

below optimum 1eveIs. In this method, each DMU is vptimized against al1 othrr DMUs.
That is. the linear pro_mam assigns weights to the variables so that each DMU looks the
best it can. DEA constnicts the eficiency tiontier and alculates the efficiency score for
each DMU based on its distance from the fiontier. The score indicates the --proportion by

which the unit c m increase its outputs without consuming any more inputs, or conversely
the proportion by which it c m decrease its inputs and maintain its current output lever
[S 1MA001The graphical representation, refer to Figure 3-1. of the efficiency scores

identities the efficient as well as the inefficient DMUs in the observed population. The
efticiency iiontier. or envelopment surace, consists of the best practice DMUs.
Connecting the efticient units forms the frontier and the ineficient units are enveloped
within. DEA links the estimation of the technical et'ticiency and the production Frontier.
Somewhere above the efticiency fiontier is the theoreticai tiontier. This tiontier is
beyond the reach of human pertormance. as there is always a chance for perhrmance
improvement when human labour is nvolved. The DMUs that are below the best
practice tiontier art: considered to be inefficient. For each ineflicient DMU, the DEA
anaIysis identifies the sources and the levels of inefflciency for both the inputs and

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

38

--

Chapter 3 Data Envelopment Analysis

outputs by comparing the inefiicient unit with a single. hypothetical, retrencr unit
formed fiom the linear combination of efficient units that operate under sirnilar
conditions. This cornparison determines the potential improvement obtainable by the

Theoretical Frontier

est-Pradice Frontier

INPUT

Figure 3- t DEA Frontier

Efficiency is typially evaluated using a ratio of output to input. The tlndamentd

idea behind the DEA nicthodology is the generalization of the single-output to singleinput ratio for a DMU. using a fractional linear progmming technique. The etficiency
of a DMU is the ratio of the weighted sum of its outputs to the weighted sum of its inputs.

This allows the DEA mcthodology to maximize the relative etficiency score of each

DMU, given that a set of assigned weights for a particular DMU is also feasible for al1
the other DMUs included in the analysis.

VALUING PRIVATECOMPANIES:
A DEA APPROACH

Chapter 3

Data Envelopment Analysis

--

3.2

- -

- -

Identification of the Factors

DEA provides an anaiyticat tool for detennining effective and ineffective

pertormance, in particular, when mu1tiple measures of performances and various

discretionary and exogenous variables are involved. The first step involving the
formulation of the DEA model is to identiQ d l relevant and signiticant resources hat are
utiIized by the DMUs. These resources are generally the inputs used by the DMU to
produce its outputs. The outputs are the factors, which represent and describe the amount
of desirable goods or services produced by each DMU. There are also many other

environmental tctors or exogenous factors that may intluence the formulation of the
model. For any of the non-quantifiable or qualitative factors, certain proxies may be
assigned and then be included in the analysis.
The generai desire by management when conducting a DEA analysis is to

ma~imizetheir outputs while minimizing inputs used in producing the outputs. This is a
particularly important issue dunng the formulation of the mode1 and the identification of
the factors.

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

40

3.3

Mathematical Formulation
The Charnes et ai. [CHAR941 model for DEA was formulated to maximue the

efllciency of each DMU subject to the constraint that the efficiency of al1 DMUs be less

than or equal to one. The algebraic representation of the model is as tollows:

C Ur-,
2 1
C

subject to ,-

for each unit j

Y b 1,

pc.Vi

where

2E

u i = the weight given to output i

y, = arnount of output i fiom unit j
vi = the wei-ht given to input i
Xij = amount of input i to unit j

The two most prominent models of the DEA methodology are the CCR and BCC
ratio models. The CCR model is also referred to as constant returns to scaLe (CRS), and
the BCC model is used interchangeably with the name variable reninis to scale (VRS).
Both of these models address managerial and economic issues and provide usetl results,
but the orientations are somewhat dieerent. Within both the CCR and BCC model a
decision can be made as to whether an input or output orientation is required for the
analysis. This iniormation is usually derived from the objectives given by the
management or the executives of a panicular organization. The brmulation of the model
can be oriented towards output maximization or input minimization. If a par<icular
output level is demanded. o r more specitically. i f a corporation has a production standard
for these DMLTs.an ineficient unit can be made efficient through the reduction of its
input levels while holding its output levels constant; this is referred to as an inputVALUINGPRIVATE
COUPANIES:
A DEA APPROACH

41

...

--

Chapter 3

Data Envelopment Anaysis

oriented model. I f the corporation has established amounts of input resources available to
these DMUs. the inefficient units can be made eficient through increases in its outputs,
while holding input resources constant: this is the output-orented model.
Figure 3-2 represents how ineficient units are projected ont0 the frontier
depending on whether the analysis requires input rninimization or output maximization.
The peers for the inefficient company C are company A and company B for input
minimization and company B and company D for output maximization.
Figure 3-2 Projecting the inefficieat Unit to the Frontier

-=
3Y

=
6

C,

y c

thit

.----------I
I

I
I

.1

I
I

Efficient DMUs: A, B, D
Peer Grout, for DM&
Input minkization: (A.8)
Output rnouimization: (B.D)
E ~ c i e n c vof DblU,
Input minimization = X,,,JX,
Output maximization = Ywg,JYc

Inputs

3.3.1 The CCR Ratio Model

The CCR ratio model focuses on the identitication of the overall efiiciency of the

DMUs studied. it also identifies the sources and estimates the potential improvements
available for the inetticient DMUs. It is important to note that the CCR model provides
the envelopment surfafe that represents CRS. This assumption illustrates that any

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

42

Chapter 3

Data Envelopment Analysis

increase in one input results in a proportional increase in at least one output or a

proportional decrease in one or more inputs must hold true among al1 the DMUs.

CCR Input Oriented

3.3.1.1

Inuut-Oriented CCR Prima1

- E- 1

-*

Min.

z, = 0 - E. 1 S-

Yh - s-

s-t,

Inout-Oriented CCR DuaI

S-

= Y,,

Mms.t-

BX,-IU-s-=o

w,, =

dy,,

vTx,= 1
$Y-vrxs0

CCR Output Oriented

3.3.1.2

The two approaches o f CCR lead to the same envelopment sufice. the only

diEerence is how the inefficient DM is projected to the fiontier.

Oumut-Orientcd CCR Prima1

s-t.

$Y,, - Y A + sXh+s-=Xo
.

Output-Oriented CCR Dual

s.t.

=O

-pTY,, = 1
opTy+ v-'x 2 O

s-. s- l O

3.3.2 BCC Ratio Mode1

The BCC mode1 [BANC841 distinguishes between technical and scale
inrtlciencies by estimating pure technical cfnciencies at the given scale of operation and
by identi fliny whether increasing. ciecreasing, or constant retums to scale possibilities are

present for further exploitation. This mode1 interprets the etftciency of the DMUs with
the underlying assumption that a VRS mechanism exists among the population of

observations. It incorporates the notion that an increase in the inputs does not necessary
translate to the same proportional increase in outputs throughout al1 scales of operation.
-

--

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

Chapter 3

Data Eaveiopment Analysis

This is different tiom the CCR model. which assumes the input and output relationships
of al1 DMUs e'xhibit a proportionai or a linear relationship for al1 scales of DMUs. The

BCC model allows the DMUs with varying scale sizes to be compared in the same DEA
analysis. This is a result of the BCC mode1 being technically efficient while CCR is scale
and technically efficient. Technical efficiencies are efticiency measures that ignore the
impact of'scale-size by comparing a DMU only to other units of sirnilar scale. On the
other hand. scale eEciency depends on the size of operations and makes this optimal.
A n y moditkation to the size will cause the unit to be less efficient.

3.3.2.1

BCC Input Orientation

The input oriented BCC model focuses on maximizing the DMUs peribrmance by
reducing the inputs 05 in other words. produces the observed outputs with a minimum
resource level. The linex proprnrning mode1 for the BCC model is:
Input-Oriented BCC Prima1

s.t.

Y h - s - = Y,,
ex,-Xh-s-=O

Input-Oriented BCC Dual

s-t.

v T x ,= i

JY - vTx +

-iC

(Constraint of Convexity) 1 h 2 1
, sC. s- 2 O

15 O

-pr 5 -c- 1
4

- v'r s -&. 1

Where s-and s-are the input slacks and output slacks respectively; p and v are the
multipliers: p,, is the measure of scale et'ticirncy. If pl, is negative then it indicates
increasing retunis to scale (IRS). If it is positive then it is operating at decreasing retums
to scale (DRS),and if it is zero then it is operating CRS. Theta. 8, is the proportional
reduction which has to be applied to al1 inputs of the DMU beinp evaluated to improve its
efliciency. The E allows optimization to occur in two steps. First. the inputs are
-

--

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

Chapter 3

Data Envelopment Anaiysis

minimized by the optimal 8.Then the optimization involves a movement to the tiontier
by the input and output slacks. Lambda, ktindicates the level o f contribution o f efficient

firms to the virtual DMU created by the projection of an ineficient DMU ont0 the
fiontirr. Those units that contribute to the construction of the imaginary composite unit
will have non-zero dual weights. AL. and make up the reference set for unit DMUo (the

DMU under consideration). The refrence set for DMUo is thus solely comprised o f
efiicicnt iinits and serves as a basis for computing the eficiency score of unit DMUO.
The magnitude of k indicates the "degree to which the characteristics of the efficient

DMUs art:used to constnict the virtual DMU on the tiontier to which the inefiicient unit
is projected" [SIMAOO].

3.3.2.2

BCC Output Orientation

The output oriented BCC model focuses o n maximizing the DMUs by

proportional augmentation of the outputs. The objective is to maximize output
production while not exceeding the yiven resource levels. The envelopmrnt surtaces for
either the input or the output-oriented BCC model are identical. The only diftrence is
how the inefficient DMU is projected to the frontier. I f 4 is greater than one. than the

DMU is inefficientOritput-Oriented BCC Primai

Output-Oriented BCC Dual

min
v-"

q , = V'X,

- v',
-p-r Y', = L

s.t.

-.

-pry+ vrx v,, 1 2 O

r1 ~

v' 2 s 1
v,, tiee

VALUING
PRIVATE
COMPANIES:
A DEA APPROACH

dS

Chapter 3

3.4

Data Envelopment Anaiysis

Differences between the BCC and the CCR model

There are some key diffierences between the BCC and the CCR model. The BCC

fiontier does not have to start tiom the origin while the CCR does. In the CCR model.
there are Iess efficient DMUs. the values o f 6 are generally smaller, and the values of $

are generally larger. To convert the BCC input oriented mode1 to the CCR input orientcd
4

model the constraint of convexity, 1h 2 1. is removed from the prima1 model, dong with
+

the p4,fiom the dual model. in a similar manner, 1h = 1 and v,, are removed kom the

BCC output model to arrive at the CCR output model. Figure 3-3 depicts the diffrences
between the models.
Figure 3-3 Graph of VRS and CRS

CRS (constant rcturlr$ to serle) CCRf!..

Output

Input

3.5

Inter~retationof DEA
Using these formulations, DEA identifies the potential improvement for the

inefficient DMUs. The intormation derived ffom the analysis is intended as a guide for
policy making and managerial decision making in an organization. It provides a better
-

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

Chapter 3

Data Envelopment Analysis

understanding of et-ficiency. and guides the management in establishing benchmarks for

similar opentions. Most importantly, the analysis helps direct management's attention to
the DMUs that are not pedbrming as well as their peers in similar environments. This

knowledge directs the organization towards a more et'fective management process.

3.6

Advantaees and Disadvantaee of DEA

DEA calcufations give new insight and permits new ways to examine data. Some

of the new insights are:

Focus on individual observations in contrast to population averages

Produce a single aggregate measure for each DMU in terms of its utilization of input
hctors (independent variables) to produce desired outputs (dependent variables)
C m simultaneously utilize multiple outputs and multiple inputs with each k i n g

stated in diff'erent units of rneasurement

Can adjust for exogenous variables

C m incorporate categorical (dummy) variables

Are value fiee and do not require specification or knowledge of a priori weights or

prices for the inputs or outputs

Place no restriction on the nctional fonn of the production relationship
Crin accommodate judgment when desired

Produce specific estimates for desired changes in inputs andor outputs for projecting

DMUs below the eficient tiontier ont0 the efficient frontier

10. Are Pareto optimal

1 1. Focus on revealed best-practice fiontiers rather than on central-tendency properties of

fiontiers
12. Satistj. strict equity criteria in the relative evaluation of each DMU

[CHAR94J

Disadvantaees
Major issues with DEA are that there are no possible improvements o r targets for

and weights are

the etticient DMUs. correlation among the varables c a ~ oexist,
t
assigned fieely and may thus ignore important variables.

Chapter 3

Data Envelopment Analysis

-

3.7

Practical Reuuiremenb of DEA

There are certain issues in DEA that need to be addressed. mainly the number of

DMUs. the environment, the weights, and the Merence between the data First, DEA
works best when there is at Ieast three times as many DMUs as the sum of the number
inputs and outputs. if this requirement is not met, the majority of DMUs will be on the
efficient tiontier and the assessrnent will not convey proper results. Second- it is essential
that the DMUs evaluated corne from a sirnilar environment. it is not appropriate to
compare manufacturing and service industries against each other since they both have
very diffrent objectives and management styles. Some frther insights can be gained

about DEA ikom the weights used. DEA assumes equally proportional improvements of
al1 inputs or al1 outputs. This assumption becomes invalid when a preference structure
over the improvement of inputs (outputs) is present in evaluating inefficient DMUs
[THAN92]. This leads to the third issue. the unrestricted weights may mean some of the

inputs or outputs, especially if the DMU is doing poorly in that dimension may be
assigned a weight of zero, wherc in reality al1 of the variables contribute in some way to
the overall efticiency. This stems fiom the fact that each unit is assigned weights so that
it looks the best it can. To address this problem, the weights can be restrcted using

analyst and management input. Bounds c m be placed on the importance of each of the
factors so that the weights assiped are more realistic- Finally, if some of the data points
are extremely large when compared to the rest of the variables, these DMUs have
potentials to be outliers and may be anomalies in the study.

VALUINGPRXVATE
COMPANIES:
A DEA APPROACH

48

Data Envelopment Anaiysis

Chapter 3

3.8

DEA Anabsis and Firm Valuation

The different valuation approaches and the different applications of DEA have

been discussed, but so fm a discussion of the usetlness of DEA and valuations tberetkom
has not been covered, The main motivation for this research cornes tiom a Ph-D.

dissertation by Sunak [SIMA001 on this subject. In this work. Simak examined DEA and
private company valuation and proposed a DEA model to identif'y similx public
companies to the one under scrutiny. The resulting peer groups fiom the DEA analysis
identifL the sirnilar companies. which when combined with their efticiency scores, the
"peer" cornpanies can be tound. The model uses data on 5 1 public manufacturing
companies, The peer group for each inefflcient Company is identitied and other
companies that have the same peer groups are collected. A difference indicator variable,

dij. brtween company i and j is introduced:

'

( h k - Aj&) - where k is the coefficient of al1 the efficient companies.

dij=
k

The lower the dij value. the more similar the companies are. He also examined the
closeness of the efficiency scores. By looking at the market capitalization (MC) of these
public companies and testing if the DEA analysis found peer companies with similar
MCs. the factors c m be combined to reved the best possible peer(s) for the subjectcompany. The results were encouraging provided adequately sirnilar public cornpanies
were availabie for reference.

DEA coupled with valuation will allow multiple inputs and outputs to be used in
the same model. The DEA comparable tirrns model is an extension of the market

approach in order to ind comparable cornpanies. DEA is ot'ten used to h d the most

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

49

Chapter 3

Data Envelopment Anaiysis

efficient unis but now it can be used to nd proper peer groups. This paper will extend
the approach started by Simak [SIMAOO], but d a e r s in several respects. The ttors
used for valuation are different, Also, two new methods are introduced. one to value
efficient companies and one to detemine upper bounds for the ineficient companies. No
estimation of the value of efficient companies has ever been attempted betore.

3.9

Novel A~alicationsof DEA

Kim and Hendry [KIM981 utilized DEA to assess NATO burden sharing among

each member nation in order to determine relative size of their net-burden. This is a
novel application of D E A which examines a net-burden index for each member nation
(the ratio between contributions to an alliance and benefits received therefiorn). instead

of calculating relative cfficiencies. In this work, two different models are evaluated and
three ofthe nations appear on the frontier in both models. hese nations are givrn a rank
of one. Subsequently. each mode1 is examined individually and the remaininy nations are
ranked according to descending net burden starting at four. Next, a cluster analysis is
applied to both the models to determine the relative position of the DMCs on the t'rontier.
ALso. the et-ticiency score fiom both the models were =paphed against e x h other to get a
simulation of the fiontier. The results from the cluster analysis support the tiontier found
%om the plot of the et'ticiency scores.

Zhu [ZHU001 conducted one of the best studies on DEA and Company efticiency.

In this work Zhu examines tinancial petiormance of Fortune 500 companies using DEA.
The data is From the April 1996 issue of Fortune Magazine. The investigation found that
the top-ranked companies by revenue do not necessarily perform etriciently. and that
only 3% of these companies were on the best-practice fiontier.

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

50

Data Envelopment Anaiysis

Chapter 3
-

Fortune magazine used 8 factors in evaluating companies, mainly revenues,

profits. assets. number of employees, stockholders' equity (SE), market value, earnings
per share and total return to investon. Zhu did not deem utilizing aU of these factors in a

single DEA model for performance satisfactory Instead a diffrent combination of these
tctors was sought. Zhu tested three models in ordrr to obtain an overall perfbrmance
index. nie tint model. called protltability. had three inputs (employees. assets. and
stockholders' equity) and two outputs (revenues and profits). The second rnodel. named
marketability, had two inputs and three outputs. (namely the outputs fiom the tirst rnodel
were considered as the inputs into this one). and the new outputs were market value, EPS
and total return to investors. The third rnodel combined the tirst two models, it used the
inputs tiom the tirst model and the outputs tiom the second model. and thus it had three
inputs (employees. assets, and stockholders' equity) and three outputs (market value.
earnings per share and total retuni to investors). This third model was meant to represent
the overall performance/efficiency- The models are shown in Figure 3-4.
Figure 3-1 Zhu's DEA Mode1 for Company Performance
Stage 3

Em pioyees

Market Value

Profitabil ity
Shrireholders'
Equity

Eamings Per
Share

4
Stage 1

Stage 1

- -

VALUINGPRIVATE
COMPANIES:A DEA APPROACH

51

Chapter 3

Data Envelopment Anaiysis

Zhu showed, with this DEA analysis. that the larger companies exhibit decreasing
rehrrns to scale (DRS). Testing the data for errors using a sensitivity analysis proposed
by Seiford and Zhu reveals that the best-practice tiontier is reliable [ZHUOO]. An input-

onented CRS model was employed in which the DEA rnethodology was modified to
show performance measures that better retlect the way businesses nnk against each
other. Results showed that substantial input savings could be achieved as evidenced by
the distribution of the CRS scores. Most industries perbrmed better with model-1
(prot'itability) than

with mode1 2 (marketability), and the model 2 scores were better than

eftlciency scores for mode1 3 (overall efficiency).

In addition to the CRS rnodel Zhu dso investigated the VRS mode1 in order to
examine returns to scale. The VRS version of model 1 revealed an eficiency of 0.46 as
opposed to 0.4 for CRS. In model 2 it was 0.29 with CRS and 0 5 4 with VRS and in
rnodel 3 it was 0.25 with CRS and 0.3 1 with VRS. The t-test is applied to the CRSineftcient companies, "the results show that the mean of the paired difierences betwern

CRS and VRS scores for the three-mode1 evaluations are si~mificantlygreater than zero
(the values are equal to 13.5.2 1.5 and 14.3 for model 1, model 2 and model 3,
respectively. at ievei 0.0 1 )" [ZHWO]. The t-test indicated that the ratios were
signifcantly less than 1. Therefore serious scale inefficiencies were present in the

Fortune 500 cornpanies. 4 summary of the above results is displayed in Table 3-1 Table 3-1 Summary results for Zhu analysis
-

VRS

CRS

t-test (0.01)

iModel 1

0.46

0.4

13.5

Mode12

0.54

0-29

21.5

Mode13

0.3 1

0.25

14.3

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

52

Chapter 3

Data Envelopment Analysis

Upon further investigation into scale eficiency and returns-to-scale. Zhu found
that the top-70 companies were not oniy scale inefficient, but they also operated in a DRS
region. while most relatively smail companies operated in an IRS region.
For the pertormance of inefcient companies, "one may use the average factorspeci tic scores within each industry to characterize the tactor-specific industry eficiency.
However. dit'trent companies with difierent sizes may exist in each industry. Therefore
arthrnetic averages may not necessarily be a good way to characterize the industry's
efticiency. usually. one expects bigger input and output levels, tiom relatively big
companies-' [ZHUOO]To summarize, DRS are found mainly among the top ranked companies by
revenue. Sensitivity analysis shows which production factor will not affect an efficient
company's financial pertormance given that ail other production tactors are kept at their
current levels. This information is usetiil when a Company considers its tiiture policy
against other companies.

VALUING PRIVATECOMPANIES:
A

DEA APPROACH

--

Chapter 4 DEA-Valuation Model Formulation

4.

DEA-VALUATION MODEL FORMULATION

This chapter presents the development of the DEA valuation model dong with the

selection of relevant inputs and outputs. Afier the appropriate variables are ascextained?
the DEA valuation mode1 is constnicted. The variables and the model are based on

available literature. expert opinion and known valuation criteria Possible new DEA
applications to business are presented as well.

4.1

Mode1 Criteria
Due to insuftiient arnount of knowledge about the geographic locations, social

contexts and structures and other qualitative inIomation, this study of the tiontier
analysis tocuses more narrowly on the available financial information. A study of just
the tinancials leads to a much more uniform study and compensates for the dit'ferent
management policies and procedures that may exkt among the cornpanies.

4.2

Identification of VaIuation lndicators

Following the collection of the data. the nect requirement in formulating the DEA

problem is the specification of the model's inputs and outputs. The three sources of
information that will be used for this purpose are available literature, expert opinion and
known vaiuation criteria,
Articles aimed at identitjring the most important indicaton of value include Zhu.

Zhu identities number of employees, revenues. profits. assets. shareholders' equity,

market value. EPS. and totai return to investors as relevant variabies [ZHUOO]. The
current study does not have access to some of this information, h r e m p t e nurnber of
employees and totd r e t m to investor. but makes use of the remaining variables. Kim

[KIM991 makes use of the: prelirninary offer price (POP). the ot'fer price (OP). first

VALUING PRIVATECOMPANIES:
A DEA APPROACH

--

Chapter 4 DEA-Valuation Mode1 Formulation

market price. proceeds. EPS, preissue book value per share (BPSp,i,uc).
sales. P/E (OPEPS). M/B,, (OP/ BPSp,,).

value per share (BPS,,stiss,)-

(OP/pc,stisue),P/S (OP/sales). and ,

,
V
M

postissue book

ME!,,

at OP in his analysis of the value of IPGs

P(IM991. Most of these variables are specitic to IPOs and were not available to this
study. Many other studies quoted earlier in the literature search section examine similar
financial ratios such as P/E and W B . Investment bankers utilize accounting intortnation.
comparable trms and also market dem@

to try and reach a more accurate price. The

key accounting variables as detennined by them are sales. earnings. operating cash flow
and book value [KiM99].

Another source for idornation regarding value was tiom communications with
professors that teach DCF and M&A courses. Free cash tlow (FCF), 7- year revenue and
profits. competing companies were some issues identified as deterninants of value.
There were also some qualitative issues to consider such as the quality of manqemrnt.
strength of market position. quality of products and services. level of customer
satisfaction. strength of corporate culture. quality of investor communications. and
effectiveness of new product development PARB971. The collection and quantification
of these types of value indicators will need to be evaluated if they are to be included in
the DEA modefs.

4.3

DEA Modeling New DEA Comaanv Valuatioa Models

There are difierent approaches to developing DEA models for cornpanies. One

c m evaluate the service efficiency of a company and as an added feature the conversion
ability of the company. or One can examine issues related to predicting market value or
rkk. Each of these perspectives require different variables. The choice of inputs and

----

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

55

Chapter 4 DEA-Valuation Mode1 Formulation

outputs has to retlect the objectives of the business. In the case ofvaluation, one wants to

examine the variables and factors that optimize the ability of the company to turn its
assets into profits.
Borrowing fiom studies done on DEA and banking eficiency. one c m apply the
sarne models to cornpanies with some adjustments. If a bank is rneasured for service

efticiency. i-e. the Production Modet. one looks at variables such as labour. oftce space.
loans and deposits. Thus. the cost to produce services is considered as an input and the

amounts of services are outputs.

If the bank is more concemed with maximidg protit. the major concem is the
Intermrdiation Model. the key variables being interest expense. drposits. interest incorne,

and loans. Thus al1 types of costs should be treated as inputs and al1 types of revenues as
outputs.
Tulkens and vanden Eeckaut [BERG981 list five approaches to efficient bank
management. The approaches are:
1. Profit maxirnization

2. Service provision

3. Intermediation

1. Utility provision
5. Risk management

One c m adjust the above approaches to retlect the concerns of companies instead
of banks. When detemining the value of a Company, the kry variables are not objects
such as number of employees. number of officeslstores. square footage of ofice space. or
number ot'products produced, time to produce and amount of sales. These are more
appropriately associated with the service efficiency of a company. and thus will be
deemed the Company Model. This measure can identi* how good the company is in

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

56

Chapter 4 DEA-Valuation Model Formulation

-

--

- - -

- -

marketing and selling their products. For example, this model can be applied to
companies such as McDonald-s and other fast food restaurants to compare efficiencies.
The Value Model on the other han& consists of the variables that make a
company more profitable, valuable, and competitive. These variables can consist of net
income. net sales. total assets. total liabilities. shareholders' equity. cash tlow. capital
expenditure. retained eamings. working capital and other quantitative and qualitative
factors.
AIso, risk management is a crucial concern for companies. One can view risk as

an input. and the return tiom the risk as an output. Thus. if given a hi@ risk more
income is preferred to less, "minimize risk for a given level of g r o s revenues.'

[BERG981. For a company, the DEA Company Risk Model can utilize the net income
and net sales as outputs. and the credit rating, i-e. credit risk. will be the input to the

approach.

4.4

Selection of Factors

Be fore the formulation of the DEA model. numerous assumptions regarding

inputs and outputs has to be incorporated into the valuation model. Some of the tetors
involving the valuation mode1 have been stated earlier in this thesis, however. there are a
few attributes that need to be discussed, These are variables that are not readily
identitiable as being relevant to either inputs or outputs in a financial operation. The
characteristics of these items and other tctors are discussed in this section.

n i e following were deemed to be the important factors used to establish the woah
of a company: total assets. total liabilities. shareholders' equity, net income, net sales and

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

57

Chapter 4 DEA-Valuation Mode1 Formulation

cash flow. These were chosen partly based on the literature review and partly because of

their signiticance in a valuation study. Their significance will be descnbed below.

4.4. f
4 . 41

The Inputs
Total Assets & Total Liabilities

Total assets and total liabilities are included as inputs to the operation because
these are items one wants to minimize. There are some -'off-balance sheet" issues with
assets that need to be addressed. Total assets are composed of things such as tked assets

and some of these tixed assets may be perceived to distort the study, A separate appraisal
may be required for certain tixed assets, for example, real estate. This is due to the fact
that the valuation approaches used for the operating entity might not be appropriate for
the appraisal of the real estate. Most srnail businesses and prokssional practices do not

own the real estate they occupy. Real estate should be taken out of the balance sheet so
cornparisons can be made more readily. Thus. the related items fiom the incorne
statement should be adjusted and instead a rent value included in the expenses.
Total assets have the characteristics of both inputs and outputs. Total assets

consist of items such as cash, receivables. loans. inventories. raw matenals, work-inprogress. prepaid expenses, total current assets. tixed assets (plant and equipment.
leasehoId improvement) and intangible assets. It can be a raw material in providing the
funds for investment or collateral, while it c m also be seen as a major contributor to
liquidity. a highly desirable state for the company. Since the primary objective is to
establish company value, it is more important to analyze how one can convert these
assets to cash and hence profit. As a result. it is considered an input into the operation.

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

SS

--

--

-- -

Chapter 4 DEA-Valuation Mode1 Formulation

Total liabilities are genenlly regarded as the "cost of doing business" a d
typically comprise accounts payable. rent. dividends payable, totai current liabilities and
long term debt. It is considered an "expense" in the financiai operation. for that reason
this measurernent wouid be considered an input. Placing total liabilities on the input side

of the mode1 works well with the pnmary objective of the DEA analysis. output
maximization and input minimization. Since the rnaximization of it on the output side of
the efiiciency model is liighly undesirable, this item is appropnately detined as input.

4.4.2 The Outputs

4. 4.2.1

Cash flow from operation:

Cash flow. as detined in a business and finance sense, is equal to the net income
afier al1 cash expenses ignoring the effect of taxes. This means d e r interest costs but
before non-cash items- such as depreciation and amortization. This c m include
amortization, amortization of detrred financing costs included in interest expense. noncash assets paid for services rendered. and others (minority interests. gain on disposal of
capital assets, development costs written off. issuance of shares to settle outstanding
debt). Defined another way. it includes items such as amortization of goodwill.
amortization of detrred costs. amortization of capital assets and income producing
property. Cash How from operations does not include changes in non-cash workinp
capital items, such as accounts receivable. inventory. income taxes payable. accounts
payable. accrued liabilities. prepaid expenses. share proceeds receivable. and deferred
revenue. Cash tlow from operations is an item that any Company would desire to
maximize. Accordingly. with the central theory of DEA of output mavimization and
input minimization. it tits well on the output side of the model.

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

59

--

Chapter 4 DEA-Valuation Mode1 Formulation

4.4.2.2

Profit and Revenue

The outputs of the financial operation include net income and net sales. Net
incorne includes revenues, interest and other income, depreciation and amortization,
investment income. and income before tax. Net sales are revenue minus cost ofgoods
sold- These two terms are the cornmon items in studying the earnings potentid of a
company. Generally, the income tiom these two sources determines whether a business
is profitable or not. And, of course. in the formulation of the DEA model, the
maximization O t' these two items indicates a better pertormance of the business.
Moreover, the inclusion of these items also allows the analyst to study the conditions of
operations in these corporations and to determine their viability.
1.4.2.3

Book Value

Shareholders' equity. often rekrred to as book value. typically consists of equity

(common and other types of shares), retained eamings. and miscellaneous shareholders'
equity. To provide shareholders e m i n g s is a paramount goal for ail businesses. It can
provide information on how eflxcient these companies are in u t i l i ~ gtheir assets to eam
proth. However. this value goes down whenever the Company pays a dividend. a
desirable feature. When measuriny return on equity. one wants shareholders' equity to
go down for a certain level of profit. It could be considered as input into the operation or
an output tiom the operation. In this research. company valuation from the perspective of
the company and not the shareholder should be esamined: thus it is more appropriate to
place shareholders' equity on the output side of the operation. Shareholders' equity is an

output as it is desirable to maximize it.

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

60

Chapter 4 DEA-Valuation Mode1 Formulation

Finally, the model constructed aims to represent the opentions o f the companies.
Inputs and outputs selected for the model are represented in Figure 4- 1. This operational
representation is the basis for the formulation of the DEA model. The specitication of the
formulation of the DEA model to measure the relative et'ficiencies is discussed below.
Figure 4-1 DEA Valuation Modcl

1 Total Assets -l

4 Shareholders' Equity 1

1 -

1 Companies

Total Liabilities

l-

Net incorne

Net Sales

Cash Flow

4.4.3 Variables Not lncluded

Items that are important but w-ere not included in this analysis are market value.
price to earnings ratio (PIE), retained earnings, and working capital.
Market capitalization was not directly included in the model. It is another key
variable that should be examined in a valuation analysis. However. by de finition, market
value only esists for publicly traded companies. To be able to adapt the mode1 for both
public and private companies the variables selected have to exist in both domains.
Tlieretore. market value was not included in the initiai analysis. Total assets can be used
to account for the size differences among the companies. Market capitalization is added

to the analysis after the DEA-Valuation models are run. The market values are used to
VALUINGPRIVATECOMPANIES:
A DEA APPROACH

61

Chapter 4 DEA-Vkluation Mode1 Formulation

-

validate the results. by not including them in the analysis they can be used for tbis
purposeFinancial ratios are often utilized to evaiuate corporate performance. The firm's
ratios are compared with some benchmark value. Size is ignored and difterent sized
tims are compared. This leads to the issue of the choice of the value to compare with. of
the dozens of ratios in existence, which ratios to use. and the choice ofthe iinns to

compare with. Companies should preferably operate in the same industry and when
constnicting the ratio there is an assumption of proportionality among the numerator and
the denorninator. However, there may be instances where this does not hold [SIMAOO],
and this leads to problems when cornparin- difkrent sized companies.
Retained earning is another very important variable in determining value. It
represents the reinvested eamings over the yean that have not been paid out in dividends.
Many corporate raiders bought companies ihat had a lot of unused retained earnings. then

afier taking dividends resold the Company. In general one wants to use the least amount
ofcash in the company while maintaining profit levels. Aithouyh retained earnings is not
directly incorporated. the inclusion ofshareholders' equity in the rnodel accounts for it.
Next, finns need cash to pay for al! their day-to-day activities. They have to pay
for wages. bills. raw materials, and so on. The money available to them to do this is
known as the tirni's working capital. Working capital consists of current assets: these are
the short-tenn assets that the firm c m use to convert into cash. Current liabilities have to
be taken into account as well to determine how much working capital a firm has at its
disposal. Working capital is theretbre current assets kss current liabilities. An
interesting issue e s e s when exmining the assets of a company. Are the assets owned or

Chapter 4 DEA-Valuation Mode1 Formulation

leased? Companies with leases rnay be ovenralued. thus more eEcient than companies
owning their assets. This cornes about because assets are located in the denorninator of

the model, so the smaller it is the better the DEA results. The companies that own their

own property will have l q e r total assets than companies that lease. It is important to
know if a Company owns land or not. since it inadvertently adds to the value of the
company. Thus it is justifiable to use total assets and totai liabilities rather than cwrent
assets and current liabilities.

4.5

Formulation of the DEA mode1

In this section. the characteristics of the businesses are examined to design a DEA

rnodel that would best describe the behaviour of the industry. The available data on these
companies will ~ l s o
be evaluated to help detemine the most appropriate DEA mode1 for
valuation measurement.
A private company is a company that is not a "public" company as detined by the

various securities acts. it tends to be smaller in size. and typically has one or few owners.

There may be signiticant diftrences in the size of the total assets ofprivate companies
versus public companies. Actually. a private company may very well have larger assets
because it does not have the motivation to use al1 the xcounting tricks avaihble to show
a larger pcr share emings. ALthouh the market values of these companies are not

directly involved in the evaiuation oftheir cfiiciency scores. their behavior can be
inherently difierent due to diffrences in size. Applyiny some banking theory. it was
suggested by Piesse [PIES951 that the scale and technical eficiency rneasurements of
tinancial institutions are inversely proportional, that is. the larger institutions are
commonly less scale efficient than their smaller counterparts. This may be due to the

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

Chapter 4 DEA-Valuation Mode1 Formulation

- -

--

requirement for a significantly larger employee group in the larger unit~.One assumes
the sarne holds true tor companies. With such a significant range of magnitudes in asset
values of these companies, it is unwise to assume that they c m al1 operate with the same
scale efticiency. Thus. the focus of this thesis will be only on the technical efficiencies of
the companies. using VRS technology.

Technical efficiency is concerned with the relationship between inputs and

outputs, where the rneasurement ignores the impact of scale size when comparng DMUs
to others on a similar scale. A more detailed description of the concept of technicd
eEiciency can be found in the DEA section. As mentioned above, the inclusion of
various sizes of companes in this study requires the assumption of scale efiiciency.
Banker, Charnes and Cooper's BCC ratio model is used to measure the technical
efficiency where the convexity constnint ensures that the unit under consideration is
being compared oniy to other units of snilar scale size.
The BCC ratio model, unlilie the original CCR ratio model, allows for variable

retums to scale where the proportional increase in the input does not translate to a
proportional increase in the outputs for the entire range of DMU sizes. The efficient
fiontier produced by the BCC ratio model is typicdly a piece-wise linear surface. The
slopes of the fiontier representing the ratios of the eficiency scores are different in the
various combinations of input and output. But al1 of the DMUs that are on the frontier
are considered to be technically efficient.

Afier the determination of the appropriate DEA model, its orientation should also
be decided. Companies could improve their performance by an increase in their output

VALUING
PRIVATE COMPANIES:
A DEA APPROACH

64

Chapter 4 DEA-Valuation Mode1 Formulation

levels. or by a decrease in their input levels. It is observed that with the limited
intiormation on Company performance. industry trends can not be reliably predicted.
Financials are available only for a single year, therefore, information regarding the

gowth of individuai companies can not be determined. With this limitation on the data,
input-orientation becomes the most appropnate approach. The output-oriented mode1
wouid project an inefficient unit by increasing its total output. But, the lack of

knowledge on whether this is physically possible makes such a projection inf'easible.

In the analysis, the CCR model is not exarnined due to the existence of negative
variables in the data set and because it is scaie invariant. The BCC model can handle
negative data because it is both scale and translation invariant. Scale invariance deals
with scaling the inputs o r outputs of a DMU by a certain amount where the efficiency

scores are not affected- Translation invariance retrs to the efiiciency scores being
invariant to the transiation of inputs and outputs by a scalar. The BCC input oriented
model is translation invariant in the outputs and vice versa. Thus the efficiency ofa

DMU wiil not be affected if the outputs of al1 the DMUs are translated by a scaiar
amount.

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

65

Chapter 4 DEA-Valuation Modet Formulation

-

Summary
Fomulating the DEA model with the available financial intonnation is one of the

goals of this thesis. Different articles cite different methodologies and uidicators that can
be included in a valuation model. The most accurate DEA model was identified dong
with the corresponding variables. In the analysis of the model no weights will be
assigned to the variables since DEA does not require pnor knowledge of the importance

of each variable and will assign weights in accordance with the Company appearing the
best it cm.

--

--

VALUINGPRIVATE
COMPANIES:A DEA APPROACH

66

Chapter 5 Data Analysis

5. DATA ANALYSIS
The data consist of financial statements for 457 Canadian publicly traded
companies in the manutacturing sector al1 having %ancial records from the year endng
in 1998. Each Company has a SIC code between 2000-3999. which entails the entire

manuhcturing sector. The industries range fiom machine rnanufacturing to food and

kindred products (refer to Table 5-7 tor a complete list). Most companies have a
different tiscal year-end, however. for this study the year-end was assumed to be
standard. As well. data on both private and public companies were collected.

The manutcturing sector was chosen for the purposes of this analysis because of
the availability of the data. A substantial number of the companies listed on the TSE are
in the manutcturing sector. Service companies. specifically hi&

technology, are hard to

monitor since most of these companies have not been in business long enough and
reliable tinancials do not exist or go back tr enough.

The data from the year 1998 were chosen because it was the most recent,
available and cornpIete. Also, when thdint comparable h

s for the private companies,

the bigger the set of data. the better- According to Booth [BOOTOO], it is better to

exclude companies that deal in resources such as pulp and paper, since they are very
cyciical. and he suggested looking at the 7-year prothbility of the companies.
Availability of this inf'orxnation is an issue. Perhaps for future research someone can
examine only a specific group ofcompanies (50 public and one pnvate Company, al1 with
the same SIC Code) and get data on the past or fture 5 years for these companies.

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

67

Chapter 5 Data Analysis

Stock Prices

5.1

To compute the market capitalization. which is equal to the outstanding shares

multiplied by the stock pnce, historical stock quotes for the public companies had to be
obtained- An arbitrary date at the end of November was chosen ter this purpose.

Public Com~anies

5.2

The data for the public companies were gathered fiom a varety of sources.

Mainly. ~rirnark's' --Compact D" for December 1999 and January ZOO0 for Canadian
publicly trded ompmes. The inf'ormation consisted of total assets, total liabilities, net
income, sharehotders' equity, net sales, date of fiscal year end, SIC codes, and so on.
Orignally. there were over 700 cornpanies' financials available. Unt'ortunately. due to
irregularities in the data. some filings from the Iate 80's & early 90's were still included.
These companies were omitted fiom the study and h a l l y 586 companies were lefi in the
data set. Missing and suspicious accounting numbers were checked and corrected if
necessary. After resolving the variable selection (inputs and outputs to use in the model),
the missing information was collected- Most of the rnissing information pertained to net
sales, shares outstanding. and net income. For the case of a Company having two SIC
codes. chietly the primary and secondary SIC codes, only the pnmary was considered.
Some of the cornpanies reported in diffrent currencies so everything had to be canverted
to Canadian Doilrs using idormation on historicd exchange rates. This insured that the
data were consistent,

Primark is one of the leading providers of disclosure information.

--

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

Chapter 5 Data Anaiysis

The tollowing sources in Table 5-1 were used to find the missing information:
Table 5-1 Financial Information Sources

tnformation
Financial statementrr,
1 annual l financial renods

1 The following web sites

1 CDNX homepage

1 were used to find the

closing market pnce for the
stocks.

Historical exchanne rates

5.3

Source

1 SEDAR homepage w~wsedarxorn

1

(Canadian Venture Exchange)t\rw.~cim.com

Yahoo Finance /~ca.f?nance.yahoo.com
FIS Online www. tisonline.coiiitn
Montreal Stock Exchange www.rne.org
Stockwatch w-\LXstockwatch.com
Microfiche copy of the Globe & Mail newspaper
fiom December 1998

[ Oanda www.oanda.com

Statistical Testing
Statistical testing of the data was c&ed out by calculating the minimum,

matimum, average. standard deviation, and percentiies (25. 50. and 75) of the data. This
information is recorded in Table 5-2. Total assets (T4), total liabilities (TL),
s hareholders' equity (SE). net income (Ni). net sales (NS). and cash flow (CFl) values are

given in millions of dollars.

,ble 5-2 Statistical Testiag for the Public Companies

iMin.
iMax.
Mean
Stm, Dev.
Pcrc, 25
Perc, 50
Perc. 75

TA 1
SIC
code '
0.006
2-01 1
3.999 395.298
2.368
3.153
572 30,934
5
2.759
33
3.3 12
23
1
3.663

TL

0.006
372.186

1.71 7
18.530
2
14
108

SE

NI

S(in millions)
-367
-841
64.1 75
9.7 18
66
65 1
3.993
599

-2

18
104

0.021
5

NS

-0.927
236.54 1
1.933
14.806

3
28
22 1-

CF1
-225
18.969'
180
1.42 1
-0.548
1
15

Chapter 5 Data Analysis

--

As c m be seen fiom the variation in the standard deviations. across ail factors. the

data set is large and accounts for a large number of smail and large companies, that
causes the variation in the data. The data contain negative values since some of the
companies had negative eamings, The range in total assets is wide and varied. from a
minimum of $6.000 (tbr Primero fndustries LTDJ to a maximum of$395 billion (for
GenenI Motors Corp.). Figure 5-1 represents the distribution of the totai assets (log scale
is used for the graphs to show a clearer picture):
-

--

Total Assets ( $ 0 0 0 ~ )

----

--

TotalAssets
-

. --

Figure 5-1 Distribution of the Total Assets for the Public Companies

As can be seen from the graph, most of the companies. 305 to be exact, have total
assets of Iess than S 100 million. Looking more closely at these 305 companies, one can

see that 68 companies have total assets of less than $2 million and 45 have assets bqeater

than $56 million and the rest lie in between, as demonstrateci by Figure 5-2.

Chapter 5 Data Analysis

Figure 5- Distribution of the Total Assets for the Companies Under S100,000,000

Statistical analysis of the firms' market capitalization(M1) was also carried out,
as recorded in Table 5-3. Market capitalization and market value are used

interchangeable in this research, Market value consists of the outstanding shares

multiplied by the stock price.
Table 5-3 Market Value Statistics

CURRENT O / S

M.C

SHARES

(SOOOsI

P R K E TSE
Dec-98

mea an
Standard deviation
Percentiles 25
Percentiles 50
Percentiles 75
1

156.723 1

3 1,133.2131

9.58

The distribution of the market value varies widely tiom $190 thousand to $6 19
bilIion. The top tive companies by market value are BP Amoco, IBM, Mobil Corp.. GM,
and Sony, their market capitaiization are $619. $240. $105, $86 and $53 billion

respectively. Ignoring the top tive companies which are multinational conglomerates. the
maximum market value in the analysis becomes $50.5 billion, which represents Norte1

Chapter 5 Data Analysis

--

followed by Thomson Corp. at $22 billion. The tive conglomerates will d l be included
in the DEA analysis because most likely they wiil be self-evaluatoa shce they are so

large and they will thus not have any negative effects on the analysis. Figure 5-3 shows a
gaphical representation of the market values.
Figure 5-3 Graphical Representation of the Distribution of Market Value:

100
80
60

Number of Cornpanier vs. Market Value

-

Ei Frequency
-- - -

40

20

O
Log of Market Value (in thousands)

There were 16 companies that have been de-listed by the TSE, CDNX or the other
Canadian exchmges. so no stock prices could be found for the year 1998. These data
were removed from the analysis. Upon M e r examination of the market value some
anomalies were found. For example, there were tive companies with market values of
less than $100.000. Subsequent analysis showed that at l e s t one of these companies.
Scott Paper. had been bought out by another Company in the preceding year ( 1997 in this
case) and still had 100.000 of its shares outstanding. The other four. mainly Consolidated

General Diarnond Corp. Difiacto Ltd., Lenox Polymers Ltd.. and Vendtek Systems.
upon M e r investigation were also excluded from the analysis.

Chapter 5 Data Anaiysis

Another important step is the examination of potential correlation arnong the data
items. In a typicai DEA analysis. correlation among the data means that some of the
variables c m be eliminated tiom the anaiysis in order to simplifj. the model. Their
inclusion in the analysis may place extra importance on these factors and skew the study.
Table 5-1 Correlation Between the Public Company Data

Correlation Matrk for 457 Cornpanies

As is evident Iiom Table 5-4,there is a hi& degree of positive correlation among

some of the populations of the data. Specitically, total assets to total liabilities have a
correlation coefficient of 0.98. High correlation exists because the data set includes only
the manufactwing industry. Manufcturing companies tend to mostly fnd their assets
through borrowing, thus total assets and liabilities will be highly correlated. Because
they are highly correlated, they would have similar correlations with equity. Neither

assets nor liabilities tend to be highly correlated with equity. as equity tends to be much
smaller. The amount of total assets gives an indication of the credit worthinesdequity of
a company. while total liabilities give an indication of the debt of the company. They are
both necessary in the valuation sheme. Net sales and cash tlow are also highly
correlated. This is partly due to the fact that companies cannot have cash tlow unless
they have revenues. More notably. cash flow is an indicator of tinancial well being,

ibqoring the impact of this variable on market value will render an ineffective analysis.

VALUINGPRIVATE COMPANIES:
A DEA APPROACH

73

Chapter 5 Data Analysis

These variables are essential in company valuation, thus. it is justitiable to use them

together in an analysis of this nature. The same argument can be used to contend the use
of the other variables in this DEA-Valuation mode!.

5.5

Private Cornnanies
One of the big major Canadian Banks provided the private company information.

[t contained statistics on the companies tkom the year 1988-1003. including tinmcials.

SIC codes. statement date. and other financiais. From this data. al1 the information on the

companies for the year ending 1998 were extracted. For companies with SIC codes
bctween 1000-3999. No information about the real names of the companies was known,
due to the contidentiality of the intorrnation. Thus, the companies are listed by number.

The data were cleaned and any duplicates elirninated. There were a total of 84 private

companies alter al1 anomalies were sorted through. Statistical intocmation about the
private companies is located in Table 5-5.
Table 5-5 Private Company Statistical Information

SIC
Code

TA

1 TL 1

NI

SE

NS

CF1

(SOOOs)

I ~ t n Dev.
.
1537
I

Perc. 50
Perc. 75

3271
3577

1324.739

1 113,326 1225.9 1 4

13,599
26.803

15 1.298

1326.308

165.855

595

27.614
40.598

1933
12.866

7,099
12.138

5.567
14.534

2.471

As private cornpanies tend to be smaller in size, the data appears to be appropriate

with the ssception of a w large private companies. The existence of this tirly large
pnvate companies will enhance the DEA analysis since there will be ample public peers.
The large private tirms are namely company #156 and #198 with total assets of $2.8

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

74

--

--

Chapter S Data Analysis

billion and $1.2 billion respectively. and net sales of $2-6 and $1.5 billion. Their net
income is quite reasonable at $457 million and $103 million respectively- The next
highest values for total assets, total liabilities, shareholders' equity, net income, net sales.
and cash tlows are smaller at $145. $11 7. $72, $68. $416, and $68 million respectively.

Below in Figure 5-4 is a histogram of the total asset distribution for the private
companiesFigure 5-4 Distribution of the Assets of the Private Companies

Table 5-6 Correlation Matrix for the Private Companies

A correlation analysis for the private companies was carried out in Table 5-6 and
it shows a high degree of correlation among the variables7especially net income to cash

tlows (0.99). The preceding section on correlation among public companies pointed out
the main reasons for this. Ayain, the purpose of this analysis is to determine the value of

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

Chapter 5 Data Analysis

the private companies. For the application of the DEA methodology, the private
companies must have the same variables as used tor the pubiic companies.

5.6

Public and Private Cornnanies

The data was sorted by their SIC codes tor both the private and the public

companies. The first group included companies with SIC codes between 2000-2099:
there were 18 other groups. thus aitogether 19. In the tirst g o u p there were 44 public
companies and 4 pnvate ones. Each of the private companies were added one at a time
into the DEA model, so in this case there were 45 DMUs in the DEA analysis. The

companies were analyzed one at a time in the DEA analysis so as not to change the
tiontier too drastically. Aiso. one does not want a private company to have another
private company in its peer group since this would not yield any usefl Iformation. The
grouping chart is shown below in Table 5-7:

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

76

Chapter 5 Data Analysis

Table 5-7 List of the SIC codes and the Corresponding Name of the Industry

[2000-2099[Food& Kindred Products

2100-2199Tobacco

1
. -L...u"...

- --

. -7 , .,.
- .J.

;>.-+.:.

O
--;."6.;

. ..

. . ,'. -0 .-..; - .
-.

..-.:;$
:
..,;

;'

..:-:-.,;
3..
: .- ( -:"-? - - ..
..

-x

240092499 Lumber & Wood

20

2600-2699Paper & Allied products

270002799Printing & Publishing's

19
22

3
5

47
85

9
3

22
36

5
5

3500-3599Machinery Except Electrcal

3600-3699Electrical equipment
3700-3799 Transportation
3800-3899Meas., Analyzing. & Control. Inst.
. --.
390W999.M&:mg;<@oys,
,-J&we%Ly;:etc:) . : - .'
-<--

a+,

-*;

4.2-

'

- . ..-.
i'
. .

g
452
-

;7'

;"'.
II_

6
84

Since it is essential in a DEA study to have at least three times the number of
DMUs as the sum of the inputs and outputs (6 in this case), some of these private

companies can not be investigated for market value. Therefore, one needs at Least (6'3)
18 DMUs for the DEA models. For the ones with less than 18 DMUs, a DEA analysis

could not be carried out, as was the case for 36 of the private companies and 91 of the
public companies. The shaded areas in Table 5-7 represent this. There are some models
with ody 19-22 D W s . This is rather small but nevertheless a DEA analysis would be
able to be carried out. The data on the public companies c m be found in Appendix A.

VALUlNG

PRIVATE
COMPANIES:
A DEA APPROACH

77

--

--

Chapter 6 Analysis and Discussion

6. ANALYSIS & DISCUSSION

Based o n the models previously discussed, this chapter details the results of the
DEA. The analysis is presented in five separate sections. The first section deals with

public companies with SIC code 3600s only. and calculates market vaiues for the
companies. Section two is a comprehensive study of the companies that have a SIC code
in the 3700s. Section three reports on companies that have a SIC code in the 2000s.
Section Four examines an etlkient and an inefficient company compared with d l 452
companies. Finally, section tive examines the methods developed tor valuing public
companies and appIies them to private companies.
The BCC model is run to classi@ the companies as efficient or inefficient. Then
the two classifications are examined separately. For the inet'ticient companies, peers are
found based on the other companies ha\:ing the same e f k i e n t companies as their peers.
The distance equation is applied and a range for the market value for the subject company
is revealed. An upper bound is found for the inefticient companies as well. The upper

bound is based on the lambda relations of the inet'ticient Company to the eflicient

companies multiplied by the market values of the efficient companies. For the efficient
companies, a lower bound c m be found by tinding al1 the companies using it as a peer.

summing these values and nonnaliziny them then rnultiplying these by the market values
and summing them to get the lower bound.

Since there are negative variables in the data, the most negative variable is found
and one plus this value is added to al1 the outputs in the BCC model, making them
positive. This c m be done since BCC is both scale and translation invariant.

VALUINGPRIVATE
COMPANIES:A DEA APPROACH

78

Chapter 6 Anabsis and Discussion

6.1

BCC Analvsis of Com~anieswith SIC Code 3600s

In this section, public company analysis was carried out using the results of the

DEA BCC model. The BCC model was nin for the 452 companies gouped by their SIC;
thus 9 models were run in total. one for each SIC grouping. The tirst malysis was
conducted on only cornpanies with SIC codes in the 3600s (3600-3699) and consisted of

85 DMC's. The data For these companies can be round in Appendix B- For each
cornpany one now has their associated eficiency. lambda, slack and multiplier values.
The market capitalization is incorporated back into the model to validate the t-mdings.
Once a value is estimated for the company. it can be compared back to the known market

capitalization of the company to see if the results are redistic. In this grouping, there are
also three private companies with the sarne SIC code. These are added one by one into

the analysis and their results are discussed in section t'ive.

6. 1.1 Summary Statistks
The a\.erage efticiency of the companies is 0.73 with a standard deviation of 0.27

and a median of0.79. There were 23 eficient DMUs and the eficiency scores ranged
tkom as littlr as 0.0653. for company Noron, to 1. Noron had $57 1 thousand in total
assets and $4.4 million in total Iiabilities along with -% 1.3 millions in net income and
-$ Z .2 million in cash tlow. These account for why its efficiency was so poor.

6.1.2 Inefficient Company Analysis

6.1.2.1

Subset Hammond

A subset of the 85 DMUs was chosen for this anaiysis. The peer p u p for

cornpany. Hammond was found. The peer g o u p consisted of al1 the efficient companies
Hammond was being compared to as well as al1 the inefficient companies that were k i n g

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

79

- -

Chapter 6 Anaysis and Discussion

compared to the same combination of eficient companies. in the case for Hammond the
peer companies were LG Technologies, Camco Inc. TS Telecom, Calian, Primetech. and
Electrohome. Table 6-1 illustrates the subset p o u p dong with efficiency scores and
lambda values.
Table 6-1 Hamrnood Subset Statistics

IDMU

6ILG TECH
7 C'4iUCO
32 TS TELECOM
1

!Score

1Market 1 Eff~cientPeen (A):

1 0.58651

0.73021
1

O
1

741ELECTROHOME 1 0.67331

27,7071 6.03E-031
11 ~0O.i)OO
1
il
11.832
O

O
O

1 1.3987 7.52E-021

0-43571

0.499

Carnco. TS Telecom and Primetech were on the efficient frontier and thus had a
score of 1. h e inefticient companies had scores mging fiom as low as 0.59 to 0-8. The
lambda values going across the table for crich DMU should sum to unity in the BCC
model; this is not the case with the CCR niodel. In the BCC mode[, this is due to the
convexity constraint. the CCR model does not have this constraint so its lambda values
have to be adjusted: refer to section two for an explanation of the adjustment. This subset
was then d y z e d with the Simak [SIMAOO] distance equation. The distance for

Hammond to LG Technologies is cornputcd as:

'

(0.33154.2637) = O. 1315

VALU~NG
PRIVATE
COMPANIES:
A DEA APPROACH

80

Chapter 6 Analysis and Discussion

In a similar rnanner the r e m a h h g distances are computed for Hammond's six peers,

Table 6-2.
Table 6-2 Distances to Hammond

1 Hammond

10

LG Technologies
Camo
TS Telecom
i Calian
1 Primetech
1 Electrohome

1 0.1315

/ 0.9152
1 0.4652

1
1

0.048
0.6956
0.0490

Al1 the companies that had a distance of less than 1-0 tkom Hammond were then

compared in terms of their eficiency scores and market capitalizations. Caiian was the
closest Company with a distance of0.048 followed by Electrohome, LG Tech., TS
Telecom, Primetech and Camco. Caiian and Electrohome were also closest in tenns of
efficiency score, 0.80 and 0-67 respectively. compared with Hammond's efticiency OF
0.76. EIectrohorne was second closest by market value. 9 1 . 4 million. followed by TS

Telecom at $1 1-8 million. Although Primetech and Carnco had a distance of less than
one in relation to Hammond. they were greatly diffierent from Hammond in ternis of their
efficiency scores and market values and thus can be ignored as peers. nie analysis
deterrnined a range for Hammond's market value basrd on its peen of $1 1.1 to $33
million. Hammond's reai market capitalization of$lll.4 million is in this range.
6.1.2.2

Upper Bound for Subset Hammond

In this part. a new method ofdetemining value basrd on the reference units tor
the inefficient companies is introduced and it is argued that this is an upper bound for the
inefficient Company. To calculate an upper bound for Hammond one uses the market
VALUING PRIVATE
COMPANIES:
A DEA APPROACH

Chapter 6 Analysis and Discussion

--

values of the efficient companies that it was k i n g compared to. namely, Carnco, TS
Telecom. and Primetech, multiplied by the lambda values of Hammond in relation to
these efficient peers. The lambda values can be found in Table 6-1. This formula is
exemplified by:
= (0.22 l7* 100,000,000)

+ (O.U67* 11,832,000) + (0.3315' lJ3,36S,OOO)

= S74,988,000

The $75 million value thus round is considered as an upper bound on the market
capitdization of Hammond, since it is based on the market values of the efficient
companies. Similar analysis is carried out on the rest of the ineft'icient companies in this
subset. The results are summarized in Table 6-3:

Table 6-3 Subset Hammoad and the Predicted Market Values

For the Ineff~cientPublic Companies(%)

Predicted Market Cap Range

Real iMarket
Cao

Company

i~lectro
home

Minimum

Maximum

1 1-39%-O001 1 1,832,0001

Predicted Upper
Bound

33.191.000(

84.103.0001

The tint two companies' actual market values falls within the predicted range. so

the mode1 appears to work well. Calian on the other hand is not within the predicted
range with a diflerence of SZ.5 million fiom its predicted upper range value. so it could

be overpriced, but it is less tliiin its predicted upper bound. Electrohome seems to be
undervalued, but again it is less than the predicted upper bound. It is only $434,000 less
than its predicted minimum range value.

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

82

Chapter 6 Anabsis and Discussion

6.1.2.3

Subset 2 Company Inverpower

The next subset in the 3600s group consists of Inverpower and its peen. which
are Avante. MM, TS telecom, Advantedge, Fifty-Plus, Kasten Chase, Wi-Lan. Plaintree,
Ne.xtron. WE XL. in total there are 11 companies in this subset- RIM. TS telecom and

Fiflty-Plus are the efficient companies. Table 6-4 represents this set,
Table 6-1 Company Inverpower and Distance to other p e r s
Score !Market Efficient P e r s (A):
Cap.
S(00s) JO-MM 132-TSTel.

ZU
Name

1 . I6E-02
INVERPOWER
0.545
7,179
AV ANTE
2-302 2-76E-03
0.583
RESEARC14 IN MOTION 1
1 585,308
If
TS ELECOM
1
1
11.832
Ol
ADVANTEDGE
0.180
19.928 1. I 7E-O2
OI
FIFTY-PLCS
2.404 1
1
1
1
M S T E N CHASE
1 O-9951 23.5691 O. 1 129991
I

42-Fife
0,920842 6.76E-02
2.1 O

O
1
8.9 1 E-02
O1

0-76 17091

0.976265
6.18E-16
2.78E-17
0.8992 19
1

O . 12529 1

I~istance
(from
D~MUrnvct.)
O
1 -635
1,829
0-0109
I -383
1.717
0.0389
1.122
0.308
1 -548
I-679

TabIc 6-5 shows the predicted market values for these companies.
Table 6-5 Subset Inverpower and the Predicted Market Values

For the inefficient Public ComnaniesC5)

1
- .
Predicted Market Cap Range
Predicted
Company
Real Market Minimum
Maximum
Market Value
upper bound 2000 (march)
Cap (1998)
17.848.000
Inverpower
1 1.832,OOO
23,569,000
3 -793,725
7.1 79.000
339-339.000
4.2 11 .O00
2.404.000
Avante
2.3 02.000
2.28 1.016
10.063.000
2.302.000
3 29.3 3 9.000
1 1-385.000
Advantdge
19.938.000
75.454.000
329-339.000
7.179.000
322.862.069
Kasten Chase ! 23369.000
24.838.000 707.841 -877
2.302.000
33-369.000
Wi-Lan
1 329.339.000
58.607.000
Plaintree
1 14-946,000
2.302.000
329.339.000
126.6 12.983
329.3 39.000
1,320,900
4.707.000
Nextron
1
1,426.000
2.302.000
I

Inverpower tlls short of the predicted range for its market vatue, this could mean
it is undervalued. It is also very inefficient and an improvement in etliciency score may

VALU~NG
PRIVATE
COMPANIES:
A DEA APPROACH

83

Chapter 6 Anaiysis and Discussion

enhance its market value. The predicted upper bound is 2.5 times as g e a t as
Inverpower's real market value. Avante's market value falls below the predicted range.
however. DEA found two very similar peers to it. namely, Nextron and Fifty-Plus.
Especially Nextron. which had a similar score and market capitalization. Avante does
seem slightly undervalued but again it is highly inefficient an increase in efticiency
might help Avante improve its market value. The predicted upper bound is 1.8 times
greater than the real market value of Avante. Ignoring Wi-Lm. which was greatly
diffrent fkom Avante in terms of the efficiency score and its market value. one c m find a
more appropriate range for Avante of $2.3 to $20 million. The same is true for
Advantedge; if one ignores the etTect of Wi-Lan. which was not even close in t e m s of
etticiency or market value. Its range becomes $2.3 to $17 million. Also. there are closer
companies like Avante. Nextron. and WE XL in terms of distance. Advantedge had the
Iowest eficiency score in the g o u p at 0.18, so this seems to be a very poor Company in
t e m s of tuming its assets into profits. That also explains why the upper bound is so 10w.

By al1 accounts, Advantedge is overvalued. It is just 0.5 times less than its upper boundKasten Chase. too is not very sirnilar to Wi-Lan. in terms of distance, eficiency score,

and market value. The closest peers are TS Telecom and Plaintree. Thus. ignoring WiLan the upper range becomes $14.9 million. Kasten Chase then becomes ovrrvalued, but
it is still operating less than its upper bound and it is highly efficient at 0.994. The
predicted upper bound is 3.2 times greater than Kasten's real market capitalization.
In all. 62 inefficient companies were exarnined and a range for market values
predicted as displayed in Table 6-6. Subsequently. it was found that eleven of the
companies did not yield any information due to too few peers. Of the remaininp 5 L

VALUINGPR~VATE
COMPAN~ES:
A DEA APPROACH

84

Chapter 6 Anaiysis and Discussion

companies DEA predicted a suitable range 59% of the time, although some of these
ranges were quite large, Another 16%were within 35% of their actual market value,
either sIightly above or below. There is evidence that DEA works better when there is a
bigger peer group.

- -

--

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

Chapter 6 Analysis and Discussion

Table 6-6 Inefficient DMUs and Predicted Market Values
Company Name

Score

Market Cap
1998

Predictd Range
Maximum
Minimum

, UPP-

Bound

0.757
0.545
0-435

1 HAMMOND
3 INVERPOWER
4 UNITEC
6 LG TECH
8 NORON
9 AVANTE
I O SEMI-TECH
l IIDANBEL

0.587

0,065
0.583
021 5
0.755
0347
0.656
0,SU

I2'AVVA

I 3 TIR
14 DIGITAL PROC
18 ABL
19 UCS TECH
20 BELL CAN
21 ClRCA
22 DBA
23 EIGER
34 KING
25 MEMOTEC
26 MICROCELL
28 NHC
3 1 - R A D E WIND
33 ADVANTEDGE
34 CALIAN
36 CELL-LOC
37 COM DEV
39 CYGNAL
40 DIMENSIONS
41 DYNAMIC
43 GEMSTAR
45 KASTEN CHASE
46 MARATHON
38 SENSE
49 SIMMONDS
50 SPECTRUM
51 SR TEL
52 STRATEGIC
53 SYSCAN
54 TEKLOGIX
55 TRIPLE CROWN

O .Z 99

1
1

0-677
0,254
0-859
0,640
0-81 5
0-472
0.649
0.066
0.67 1
0.676
O- 180
0.800
0-980
0.863
0-589
0.1 I O
0.939
0.232
0.9Ujl
0.604(
0.3 7-1
0.766
0.787
0.447
0.983
0.657
0.8 1 O
0.526

74,988.000
33-191.O 18
18,426,080
I 1297.939
17,848,000
7,179,058
1 1,832,000
33-569.233
5,057.000
2,404.000 one peer
5,570,707
47,055,000
33.19 1 .O 18
27,706.969
l 1.397.939
2-383.000
1,104,000 one peer
7,039.3 1 O
4.2 1 1.000
2,404,000
329.339.3 22
3.3 02.125
I00,000.000 one peer
720,885.000
10.025.976
44-09 1,000
30, 1 56.000
289,000
5,797.5 76
1 3 -3 82,000
2,669,000
1 -710.000
3,429,279
I
5
-3
83,000
3-53
7,000'
1,709.64 1
1.956,OOO
1 54.005.000
28.123.144
1 1.832-000 one peer
7.609.000
1 1,832,000
1 -7 10.000
13.381.932
1 5.097-570
3-376.000
622.999
1 -026,687
3.60827 1 .O00 2,448,639.000
1-371-1 19,948 1-031.559.000
46-09 1 .O00
22,797.000
6,600,085
3.855.000
1 3-3 82.000
3,782,000
1-7 10.000
1,956.000
5-512,000
3,878.000
2,404.000
8289.000
1,854,080
1,072,000
307.000
7-738-941
76.67829 I
48,488,000
4,952.53 1
1 1,832,000
2 1-615,000
3.878.432 one peer
307.086.660
44.4 1 1 .O00
2,404,000 one peer
7.3 1 7.082
1,485.000
15,697.570
307.000
9.699.54 1
I0,064.000
329.339322
19,927,644
2,302,425
62.705.000
27-706.969
33-191.018
1 1.397.439
23,996.000
2,404.000
1 -969.999
28,669,000
85
1.037.000
1
.
3
9
.
0
3533 15.304
1 12,896.7 I O
50,650,000
32.734.620
1 1.832,OOO
8-637.174
13.34 1-000
28.669.000
2.404.000
1-969.999
89.29 1-000
80.6 15.000(one peer
1 10.353.695
3 -327.000
6 1220.528
1,729.263
4,887,873
75.454,000
329.339.322
7.1 79.058
23369.233
2.447-000
15,697570
7 1 7.0 1 0
622.999
4.887373
1 5.085.000
1,739263
6 1220.538
7-73
8-94
I
3 -334.000
3.878.000
2.854080
76.67829 1
1 00.858,OOO
38,637.268
4.952.52 1
1 .O3 1.559.000
532269.000
44.404,l 12
1 12896.710
l3,222.000
6,597.000
2.404.000
3,943. 1 54
5.65 4,000
3,855.000
2.404.000
739.505
1 79,974.000
143.365.000
243.50 1 .J58
44.404.1 12
4.8 18.000
I 3-382.000
1.7 10.000
4978.95 I

Chapter 6 Analysis and Discussion

56 TSI TELSYS
57 WESCAM
1

61 SIERRA

63 STRATEGIC
64ISY MPLEX
651WI-LAN
WORLD
661C'IRC'UIT
,
69 COREC'O
71 SHELLCASE
71 T U S D M
73 c'-MAC
71 ELECTROHOME
75 LAMBUS
77,PL.4INTREE
78 TRI-VISION
80 BLCESTAR
8 1 COMPRESSION
83 GSI LLMONICS
83 NEXTRON
84 ROYAL LASER
I

0 21 1
0,628
0.7571
0,996
0,534
0,562
0,873
0.753
0,983
0,085
0,983,'
0-817
0,673
0-7i 0
0.866
0,743
0,555
0.3 I O
0.9 t 7
0.534
0.684

-- --

622,999
12,192,753'
15,697,5701
2.3 04,000
76.678.29 1
4-952-521
I43.365.0001
169,2 12.000
9-421.6651
2,404,000 1one peer
1
14, 108.00a
3,855.000
32,734,620
1 1.832.000
1 07.539-000
622,999
1,719,718
15,697.570
2,004,000
5,879226
1-71 0.000
1 3-382.000
2-657,000
339,339,323
23,569.233
24.838-000
2.3 02,425
188.73 7
46-001,000
24.592.000
5,798,000
.dl-922-938
2,404.000
5 1-361.300
1 53.377.000
622.999
15,697.570
13.192.753
2-14 1.000
120,705.32436,523,000
3.855.3001one peer
29 1,126.880
100,000,000~ 1 .O3 1359.000
969.673,OOO
1 1,397,939
339108
84,103,000
1 1,832.000
2,404,000
1,729,263
6 1220,528
7.566,000
14,946,155
329-339.322
58,607,000
2,302.425
2,404.000 one peer
39,930,177
53,470,OiiO
44,404,112
112,896.7lO
717.010
622.999
1 5,697,570
3-391-000
129.625.600 1-031,559.000 one peer
1-037,907.000
4,707,000
2,426, I 40
2302,425
329.339.322
5 1.56 1.300
1 5 1,693.O00
2,404.000
U.922.93 8
1

- -

VALUINGPRIVATECOMPAN~ES:
A DEA APPROACH

Chapter 6 AnaIysis and Discussion

- --

Figure 6- 1 demonstrates the predicted and the actual market values:

Figure 6-1 Inefiicient Companies and tbeir Predicted and Actual Market Values

1 Diamonds represent actual market caps

The real market values were also graphed against the predicted upper bound- Of
the total 62 companies, 41 companies had r d market caps that were less than the
predicted upper market capitalization. thus DEA is correct 66% of the time in predicting
an upper market cap that is truly higher than the market cap of the Company. This is the
result one would hope for since the upper market value is based on the etticient
companies. and intuitively, the market caps of the inefficient companies should be below
these values. Another 13% of companies were within 35% of their predicted upper
bound.

Figure 6-2 represents the predicted upper bound versus the a c ~market
l
capitalizations.

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

Chapter 6 Analysis and Discussion

6.1.3 Efficient Companies

In this part. a new way to analyze eficient companies is introduced. The anaiysis
for the efficient companies was carried out in a diflferent mamer than for the inefticient
cornpanies. Since these efficient cornpanies are rarely compared to other companies. one
has to calculate the market value fiom the cornpanies being compared to them, whether

they are inefiicient or efficient. One c m hypothesize that this way a lower bound on the
market values can be tound. Since it is an efficient company, it should presurnably be
doing better than the market value cdculated from the inefficient companies in its peer
group. The lower bound was calculated based on the lambda values of the ineficint
companies being compared to the efficient company. Adjustrnents must be made for the
h c t that the lambda values on the column of the matrix do not s u m to unity.

6.1.3.1

Lower bound for Efficient Company Camco

Camco was analyzed tirst: it had 17 cornpanies being compared to it. The lambda
values of the ineficient companies in relation to Camco were surnmed then this sum was

used to divide each lambda value to get a new normalized Iarnbda (A / Ch). For example.
Harnmond has a lambda relation of 0.222 to Camco. the sum of ail the lambda relations to
Carnco equaled 1 -961. Hammond's new lambda score is equal to (0.22l7/l .%OS) 0.1 13 in
relation to Camco. This same procedure was repeated for al1 the other companies' lambda
values, thus the new sum of the lambda values equals one. The new, normalized tambda's

of the firms were then multiplied by their market values and summed. The resdts are
s h o w in Table 6-7:

VALUINGPRIVATE COMPANIES:
A DEA APPROACH

93

Chapter 6 Analysis and Discussion

-

--

Table 6-7 Efficieat Company Camco Inc (#7)

DMU

Score

#
1 HAMMOND
6 LG TECH
10 SEMI-TECH

15 SONY CORP
17 AASTRA TECH
23 EIGER TECH

26 MICROCELL
24 CALIAN TECH
37 COM DEV
5 1 SRTELECOM

0.7568
0.5866
0.2 145
I
i
0.8151
0.0664
0.800 1
0,8633

0.447

52 STRATEGIC
0.9832
54 TEKLOGIX
0.809V 1
1
CELESTICA
INC
68
73 C-MAC
0.8171
71 ELECTROHOME
0.6733
80 BLUESTAR
0.5547
X2 GSI LCMONICS
0.9 167
SUM:

Market Cap DMU #7 Normalized

S(000)
(old A)
New k (A / CA)

18426
27707
10026
5.3&+07
4609 1
8289
307087
33 191
152315
1 12897
3943
24350 1
2854867
29 1 127
1 1398
44404
129626

0.2217
0.006(
0-7746
3.1E-15
6.3 E- 1 8
0.0107
0- 1845
0.0599
0.08231
0.13631
0.00 111
0-0755
2.8E- 16
0.2202
0.0752
0,0926
0.0 196
1.96053

0-1 13 1
0,003 1
0,395 1
1-6E-1 5
3.2E-18
0.0055
0,094 1
0.0305
0,042
0,0695
0,0006
0-0385
1-4E-16
0.1 124
0.0384
0.0472
0-01
1

M.C. J(0s)
Calculation
2083-92
85.22
396 1.35
8.343E-OS
1 -483E- 1 3
45-2-1
38902-05
1 O 14.08

10604.66
7835.93
2.45

9377.23
4.092E- 1 O

32707.43
437.19
2097.3
1295.9
S100,460

For Camco, the predicted market capitalization is $100.459.000; its real market
value is $100.000,000, DEA did an exceptional job at predicting the value of the public

company. Similar analyses were carried out on the remaining 22 eficient companies and
are recorded in Table 6-8. A complete malysis can be found in Appendix C.

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

94

Chapter 6 Analysis and Discussion

--

Table 6-8 Enicient DMUs & Number of Peers (SIC 36009)

For the Efficient Public Companies (%)

Company

Real Market Cap

(9
Turbo Genset
Global Therm
Camco
Sony
C inrarn
Aastra Tech
Lui itel
Nortel
Res. In Motion
TS Telecom
Cdn Marconi
Comm. Sys
Fi@-Plus
In t. Datacast.
Newbridge
Jemtec
Pacitk E-link
Snif Security
Primetech
Celest ica
Gennum
Leitcii
Bal lard

80-615,000
1 3,121,000
100,000,000
53.93 5-385,000
1.075-69 1,000
46-091 .O00
1.257-3 78,000
50,495-406,000
585-3 08.000
1 1 ,x33.000
426,508,000
3.855.000
2,404.000
3,893,000
8.125-48 1,000
330,000
3,878,000
3 07,000
1 1 3-365.000
2,854,867,000
576,578,000
1 .O3 1 ,559.000
3.60827 1.O00
--

Predicted Market Cap

(Lower Bound) (S)
79-4 13,730
2-3 58,240
100,439,049
Self evaluator
Self evaluator
4,427.000
Sejf evaluator
Seif evaluator
48,895.000
25,878.000
Self evaluator
43-170.000
3 I ,030.000
5,074,000
5 1,858,075,000
3.943.000
86.2 16.000
7-164.000
73-681.000
487.320.000
50,868.000
287,982.000
1.3 7 1-1 19.000

Comment

# of

Peers
Above lower
Above lower
Above lower
no info
no inb
Xbove lower
no info
no into
Above lower
Undervalued
N o info
Cndervalued
t'ndervalued
L ndervalued
L'ndervalued
Cndervalued
Undervalued
L'ndervalued
A bove lower
A bove lower
Ahove lower
Ahove lower
A bove lower

6
4

17
O
O
5

O
I8
29
O
9

45
4
2
1
18
11

15
6
5
21

Excluding the tive self-evaluators, DEA does a tirly good job of tinding a
reasonable or appropriate lower bound for the companies. as evidenced by the majonty of
companies having real market values hiyher than the predicted lower market value. DEA
finds a lower bound that is less than the tme market value 56% of the tirne. In some cases,
there are too few peers and the calculation maybe exaggerated or the peers may be vasrly

ditferent fiom the Company under analysis. The seven companies that had market caps
below the predicted lower bound cm be classitied as undervalued, since accordinp to the

Chapter 6 Anaiysis and Discussion

---

--

analysis their value can be greater. Efficient Company Snif Securty was the most

undervalued. It had a predicted lower market cap of $7 million when in reality it only had a
market cap of $307.000. It had a 2234% difkierence between its predicted and real market

value. Figure 6-6 displays the red market values compared to the predicted lower bouod?

and Figure 6-7 shows the percentage dimerence between the real market value and the
lower bound.

--

VALUINGPRJVATE
COMPANIES:
A DEA APPROACW

Chapter 6 Analysis and Discussion

Efficient CO'S Real vs Predicted Lower Bound Y.C. ($)

Company Name

--

m ~ e aM.C.
l

m~redictedLower Bound MC

Figure 6-6 Eincient Companies-Real MC versus the Predicted Lower Bound

VALUING PRIVATE
COMPANIES: A DEA APPROACH

--

-.

--

- .

-- -

Chapter 6 Analysis and Discussion

Analvsis of the Cornonnies with SIC Code 3700s

6.2

A similar anaiysis as tor SIC 3600 was caned out on the companies with a SIC

code in the 3700s- This group consisted of 22 DMUs. This group was tested to see if the
number of DMUs in the analysis really has an effect on the number of peers that DEA
tinds. AH 22 DMUs were analyzed. The average efficiency of the BCC mode1 was 0.8 1
with 8 efficient DMUs. The efficient DMUs are listed in Table 6-9.
Table 6-9 Efficient DMUs SIC3700s
#

-7
5

6
7
Il
I I
21

-73

IDMU
GM
Budd Canada
Linamar

# of Peers
O
12
6

Magna

Spectra
Wescast
Templar
Bornbardier

6
8
1I
O

DMUs Budd and Templar are the most employed peers with 12 and 11 peers
respectively, however Templar has a higher degree of intluence over its peers. DMUs

GM, Mabqa and Bombardier do not act a reference unit for any of the inefticient units.
This could mean they are either efficient in a very narrow sector or that they possess a very
uncornmon input/output mix or they are just very different. The cornpiete eficiency scores

and lambda values are shown in Table 6- 10:

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

Chapter 6 Analysis and Discussion

DMU

IScore ISIC IMarket

Cap.

1 ~fficientP e r s (As):
1 3

Il

14

21

6.2.1 Efficiency and Market Value

Figure 6-8 shows the efficiency for al1 the DMUs in the analysis ordered by their

market capitalization. It is evident fiom this figure that the larger compaties, the ones
with the white points, are al1 efficient or nearly so. The correlation coeficient is 0.26.

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

100

22

Chapter 6 Analysis and Discussion

Figure 6-8 Efficiency and Market Value (SIC3700)

6.2.2 Efficiency and Total Assets

The correlation coefficient between eficiency and total assets of the DMUs is O. f 5 ,
illustrated in Figure 6-9. Thus there is great difference between the total assets of the
various companies.

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

Chapter 6 Analysis and Discussion

Figure 6-9 Eniciency and Total Assets (SIC3700)

<O

IS

20

T o t a l A s s e t s (LN S c a l e )

6.2.3 Graph of Vittual OuQDuts

A virtual output is simply the product of a gven output and its weight. its value

indicates the importance attached to a given output. In the case of the CCR model. the sum
of the virtual outputs is equl to the unit's efficiency score. In the BCC model. the sum of
the virtual outputs plus the mesure of scale efficiency, %. is equal to the unit's eficiency

score. For each DMU, a stacked column that has the unit's virtual outputs as its senes is
plotted in Figure 6-10. It displays the relative amount that each output factor contributes.
It is interesting to note that Company 21 has no conmbution t'rom its outputs. it is solely
made up of the contribution tiom its input. namely total liability. Figure 6-10 also shows
that the factor shareholders' equity is the most influential output.

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

102

Chapter 6 Analysis and Discussion

Table 6- 1 1 surnmaxizes the predicted values for the companies with SICs in the

3 700s and Figure 6- 1 1 shows the predicted ranges. There were 1O peer groups in total
with some companies only being compared to efficient companies. For most DMUs

there were at least duee peers. with the exception of Trailmobile and Devtek. which oniy

had two peers. and Serena which only had one.

Table 6-1 1 Range for -Market Values (SIC3700)
+

For the inefficient Public Companies(S)

lcompany

1 Real Market
Cap

A-E Ventures
Avcorp
Merch
Rose
Stackpole

Glendaie

predicted Market Cap Range

- 1
IMinimum
IMnximum
[Predicted upper
baund
1

627.000
45.730.000
f,869.000
25.150.000
1 3 1,497.000
-17.948.000

4,576,000
627.000
4576.000
2,869,000
1 16,772,000
4576,000

4.5 76.000
4.576.000
525.280,OOO
----------4, ~76,000
1 16.772.000
574300,000
i 16,772,000

55.3 27,000
16 1.222.000
One peer
52 1-65 1.O00
3 1 4.7 1 0,000

1 16.772.0001

96 1.8 1 7.000

-----------

Northside

Heroux
Serena
Mage1lan

Devtek

14.056.000
63,162,000
572.000
496-389.000
99.709,OOO

Tesma

570.121(

45.730.000
525,280,000
25.150,OOO
525.280.000
574.3 00,000
1 16,772,000

-----

1.792.5 1 7.0001

1 2.58 1 ,000
23 5.632,OOO
I 4.730.000
3 1 6.2 13.000
43 7.63 8.000
60,753.000

For the Effkient Public Compaaies(%)

Real M.C.
Preicted Market Cap. Lower Bound Comment
I 16.772.000
Budd Canada
2441 1 6.000
Undervalued
Linamar
S pectra
Wescast
Templar

GM
Magna
Bombardier

1 -792.5 1 7.000

502,534.000
64.3 57.000
383-838.000
16.3 59,000
Self Evaluators
Self Evaluators

525280.000

5 74.300.000
4376.000
86.392.582.000
7.3 74.1 03.000
7,502,049.000

Self Evaluators

VALUINGPRIVATECOMPANIES:A DEA APPROACH

Above lower
A bove lower
Above lower
Undervalueci
No info
No infb
No info

Chaptei 6 Aiiiilysis and 1)iscussion

--

- -

Figure 6-1 1 Range for JDredictedMarket Values and the Real Value (SlC3700)

The rcd dianwndu reprewnt the real market values

Ali Ven

Avcurp

Mrch

Rose

Stuck

VALU~NG
PRIVATE
COMPANIES:
A DEA APPROACH

Glend

r u

1 Icroux

12

10

Mugcl

Dccoiiia

Tesmil

1OS

Chapter 6 Analysis and Discussion

Figure 6-12 Inefiicient Company-Real & Predicted Upper Market Value (SIC3700)

A.E
Ventures

Avcorp

Merch

Rose

Slackpole

Glendale

VALUINGPRIVATE
COMPANIES: A DEA APPROACH

m ~ e aMarket
l
Cap

Trallmoblle

Norlhslde

Heroux

P r e d i c t e d u 3per bound

Magellan

Oevtek

Decoma

Tesma

Chapter 6 Analysis and Discussion

Figure 6-13 Enicient Pirms-Real and Predicted Upper Bound(SIC3700)

m ~ e a M.C.
l

Budd Canada

Llnamar

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

m~redictedMarket Cap. Lower Baund

Spectra

Wescast

Templar

Il

Chapter 6 Analysis and Discussion

in the case o f the inefticient companies, in Figure 6-12. DEA accurately classifies
64% ofuie companies, but without exception the range from the minimum to the

ma~imurnvalue is very large. In the instances when DEA camot c l a s s e the companies,
it is due to there being not enough similar peers. DEA correctly t'inds an appropriate

lower bound 60% of the time for the efficient companies. F i p r e 6-13, with a SIC of
3700. Appendix D contains the information on the companies with a SIC in the 3700s.

6.3

Analvsis of the Companies (SIC20001

A similar analysis for companies with SIC in the ZOOOS was also performed. This

~ ~ o represented
u p
food and kindred products. There were 44 DMUs in this subset, with
15 on the efficiency fiontier and an average efficiency of 0.8 1. The most popular

efficient peers are A. L. Van Houtte Ltd. with 21 peers tollowed by AMT Fine Foods
Ltd. with 19. Analysis of the lower bounds of the efficient companies shows that 67% of
the companies are tound to be undervalued, they have actual market value less than the
predicted lower bound. Canada Bread Company Ltd.. S e a g m and Petroreal Ol
Corporation are Lound to be self-evaluators. For the 29 ineftcient companies. 10 of them
do not reveal any information due to the tct they have one or no peers. The remaining
19 companies have a correct predicted market range 68% of the tirne. Also. 86% of the

tirne the inescient companies have an actual market value less than the predicted upper
bound. TabIe 6-12 shows the range of market values for the inefticient companies and
Table 6- 13 demonstrates the market values for the eEcient tirms.

--

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

Chapter 6 Analysis and Discussion

Table 6-12 Range for Market Values -Inefficient Companies(SIC2000)

For the Inefficient Public Companies($)

Efficient
Peers

Company

3 SEPP'S
6 T M L FOODS
7 SAPUTO
8 DELICIOUS
10 LASSONDE
I l GLOBAL COMM
14 AGRO PAClFIC
15 RI DLEY
18 BETA BRANDS
19 S M A L L FRY

121 BANFF

121 BIG ROCK

23 BRICK BREWING
25 SLEEMAN
116 TREE
28 ANDRES
39 MAGNOTTA
13 I VINCOR
32 HIGHWOOD
34 ARROWHEAD

1Predicted Market Cap Range

Minimum
Real
Market Cap

Maximum

Predicted
Upper Bound

4-5, f 2
27,532,000
1 1.173.0 1 5 one peer
4-33 1
1,782,000 l
1 1.1 73 .O 15 1one ~ e e r
2-24-431 2.1 07.8 19.000 632.44 1262lone peer
302.0001
3 8,002.000
4 1 -43 1
1 7.246.000
1 7.825.0001
15.889.0001 1
632.44 1.O001
I
11.4 10.000
1 -9 1 7.000
3 -93 8.000 1 6.4 1 -43
8,570.000
2-45
L 0.874.000
1 1 - 1 73,000'
2 - 4 4
i h0.05U.000 no peers
no peers
4 1
47,999.000
1 1.1 73,000 one peer
1 1.1 73 -000
15,889.000
4
36,740.000
1

1
1

I
1
1

18.037-000
10,684,000
586.998-000
19.742,OOO
328-5 12.0001
5367.000
19.65 4,000
48 1.98 1,000
19.308,OOO
29.793 .O00
w

10.33.4 Ilt
4 14
4-12-43
44.43
1 6-41 -431

1.3 78,00011
19.445.000l
1 5-42 1 .O00
126.836.000 1
1.9 1 7.0001

2.879.000I
3 -938.OOOl1
1 1,173 -0001
13,878.000[
1 1.173.000 one peer
1I,173.000
632,441.000 1
302.0001
14-41 0,0001

2.88 1 .000I
7 1 -738.0001
52-8 1 1.O00
205,764,000
2.368.0001

4,924.43
4 - 2 4
9 1-1
4.1 6.4 1-43
4- 16.43

79.826.000
1 3 -310.000
197.1 14.0001
2,373.000
962.000

1 1.173.000
633-44 1.000
126-836.000
1 1 - 1 73.O00
13.889.0001
632-44 1.0001
3.93 8,000 one peer
1 1.1 73.000 one peer

139259.000
49,850.000
288.746.0001
6.5 17.000
13,002.000

36 COTT
37 LEADING BRAND
139 URBAN JUICE
1
40 VlVANT
42 LEF MCLEAN
1-14ARCTIC GROUP 1
I

I.LX 330.647.000
8,570,000
2.4.5
4.4 1-431
12.878.000~
33
876.000)
3,381.000)
-1.5.9.1 2
4 1.43 1
3 8.002.000)
I

139.628.000
1 0,874,000

1.3 74.236.000[
1 1-1 73.000f

1 1,173,000~
19-445.0001
2.879.000 Similar to Aqua [
11,173.000onepeer
1
302.000(
17,246,0001
I

VALUING PRIVATE
COMPANIES:A DEA APPROACH

552.839.000
10,947,000
14.83 4.0001
2,879.000
14,939,000
66.865.0001

109

Chapter 6 Anaiysis and Discussion

Table 6-13

Range for Market Values -Eflcient Companies(SIC2000)

For the Efficient Public Cornpanies(S)

Predicted.
Number of Comment
Lower Bound
Peers
I
Above lower
1-374.23 6.000
95-21 7.2 1O
II
Undervalued
32 I.064.000
139.628,OOO
19
Undervalued
24,079,000
1 1,173,000
Undervalued
5
5-936,000
1 2-038,000
9
Undervalued
73-881,000
15.889,OOO
Above
lower
3
20-13 3 -000
53.129.000
U ndervalued
1 92-1 14,000
2
Undervalued
7
5-9 18,000
N o infi
O
self evaluator
449.753 .O00
Undervalued
10
1.060,747.000
N o inb
O
self evaluator
Above lower
3
!,085,000
No info
1
5,304.000
self evaluator
Undervalued
1 3 -971 .O00
11
302.000
Above lower
21
SZX,193.000
632.44 1.000

Real M.C.
1

1 1 MAPLE LEAF
2 SCHNEIDER
4 AVT FINE
5 FLETCHER'S
9 SLN-RYPE
12 DOVER
I

1 1 7 CAN BREAD

138 PETROREAL
JI EVEREST
3 3 A. L. VAN
L

Appendix E contains the information on the companies with a SIC in the 2000s.

Analvsis Utilizine All 452 C o m ~ a n i e s

6.1

Some may argue that using different SIC codes in the same analysis is incorrect.
However. there are studies that do in fact use companies with difirent 4-digit SIC codes
in a valuation analysis, such as Renaissance Capitai [ZHUOO]. To test the theory that a

relevant peer group would be found when al1 452 companies were included in the
valuation analysis. an efficient and an inefficient DMU were analyzed with al1 452
companies.

6.4.1 Inefficient Hammond compared Lo al/ 452 Companies

Comparing DMU Hammond to al1 of the companies resulted in Hammond having

2 1 peea of which 4 were efficient. The range for its market value based on these

p
.

..

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

--

Chapter 6 Analysis and Discussion

companies was between $3.4 to $142.8 million with a predicted upper bound of $57.5
milIion. Hammond's real market value is $18.4 million. The most similar companies
were Patheon. Sico, and Lassonde with SIC codes 0~2834,2851and 2034 respectively
and market values off 142.8, $71. and 5 17.8 million. respectively. The only other
company with the same SIC code in the 2 1 peers thus tbund was Aastra. Even though this
analysis resulted in a bigger peer group for Hammond. the results did not conclusively
tind a better. tighter range for its market capitalization.
6.4.2 Efiicent Company Wescast Comparecf to a11 452 Companies
An efficient company. Wescast, was also tested with al1 452 companies. This time

around Wescast had 97 companies utilizing it as a peer. The lower bound calculated

based on these peers was $408 million, with Wescast having a real market value of $574
million. This m o u n t was better than the $383 million predicted as a lower bound for
Wescast when it was compared only to cornpanies with an SIC code in the 3700s just like
itself. In this cornparison, the peer group consisted of companies with SIC codes ranging
f?om 2000 to 3900, The most comparative group were companies with SIC in the

3500as,numbering 15 in total. There were also 8 companies with SIC in the 3700s
compared to Wescast. including 6 that had the same identical SIC code. Appendix F
contains the information on this analysis.

6.5

PrivateCompanies
The private companies were analyzed by addng one at a time and re-running the

DEA BCC mode1 with the 85 DMUs. Once similar reference peer units are found, the
private companies can be assumed to have the market values of these retrence DMUs;
thus one has found a market value for the private companies based on the similar public

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

111

Chapter 6 Anaiysis and Discussion

-

- --

companies. The private companies were analyzed in the same manner as the effrient

and inefficient public companies. As mentioned, there are three pnvate companies
having a SIC code in the 3600s,namely. company #55?85, and 220.
6.5. f Inefficient Pnvate Companies

Mer the initial DEA analysis, private company #55 was found to have an

efiiciency score of 0.77 in BCC. It was then analyzed using the procedure applied for
public cornpanies above. Untbrtunately. the company had a very unique dekition of
peer groups. consisting oniy of 4 other companies. ail efficient- The peer group included
Astra Technologies. Ts Telecom, Communication Systems Internat. and Fitly-Plus.Net

International. Afier the analysis, it was found to be most similar in tenns of distance and
efticiency score to TS Telecom. The a c d values found are listed in Table 6- 14.
Private Company #220. coincidentally, had the same peer grouphg as subset
Hmmond above. In its peer group were LG Technologies, Camco Inc, TS Telecom,
Calian, Primetech, Electrohome and Harnmond. it was found to be most similar to TS
Telecom. which had a market cap of $1 1-8 million.
Table 6-1 4 Inefficient-Range for Market Values (Private Cornpanies)

For Ineffkient Private Companies ($)

-Predicted Market Cap Range (S)
Closest to: Predicted
minimum
Maximum
Company
Upper Bound
46,000.000 1 1.832.000
17,700,000
Company #55
2,100.000
37,414,000
33.191.000 11.832,O0Oi
1 1,832.000
Company #220
6.5.2 EIRcent Private Compenies
For the efficient pnvate companies the same analysis as for efficient public
companies was conducted, the lower range found is listed in Table 6- 15.
-

VALUINGPRNATE
COMPANIES:
A DEA APPROACH

Chapter 6 Anaiysis and Discussion

Table 6-15 Efficient-Range for Market Values (Private Cornpanies)

For Escient Private Companies

l~ompany

Predicted tower bound ($)

Analysis Summary

To surnmarize. the DEA-Valuation model is a way to identi @ a set of snilar

public companies that c m be used to estimate the market value o f a private Company.

DEA is traditionally used to measure relative efficiencies and this novel application
expands its tnction. DEA does a firly good job in determining appropriate peers for the
companies when utilizing the BCC model. The CCR model was not employed due to the
existence of negative variables in the model. The DEA-Valuation approach works better
with more DMUs. There must be more than 22 DMUs for DEA to find a correct range

for the company under analysis. Private companies are valued by tinding suitable public
peers and intmng their market values ont0 the private companies.
Although the range specitied by the peers is sometimes quite wide. the method
still does an above average job of classitj6ng market value. For companies that have only
one peer or none (the self-evaluators), the resdts are not binding. A larger data set, if
possible. should be collected. This will increase the likelihood of correct classifications.

Chapter 7 Conclusions and Recommendations

-

--

7. CONCLUSIONS & RECOMMENDATIONS

With a global increase in mergers and acquisitions and the number of companies
coing public, the need for valuation tools and methodologies increases. Determining

Cr

value is a si@icant part of any business. be it for selling a business, divorce settlements

or for mergers. Value estimates c m vary widely and to insure a f ' r value for the
business a number of issues need to be addressedValuation is O ftentimes a subjective task performed by valutors. business
analysts. accountants. lawyers and investment bankers. The existing methods for
detennining this value consist of forecasting g o w t h and fture cash tlows, or optimizing
the value of owned assets by liquidation or finding similar companies and adjusting their

forecasts and growth expectations to arrive at a value. The goal of this thesis is to
develop a methodology that evaluates the market value of corporations utilizing Data
Envelopment Analysis to find comparable t - m s and to apply this methodology to assess
the worth of private companies.
A DEA andysis of rnanufacturing companies was studied. Valuation models for

predicting market ranges were developed. DiRerent approaches were constnicted based
on the performance ofthe Company in the DEA rnodel. Ineficient companies had their
most similar peers identified which lead IO a predicting of the market value range. An
upper bound for their market value was determined as well. For the efficient companies.

a lower bound market value predicting approach was constructed.

The results were encouraging when a highly comparable peer p u p existed. In
total over 200 companies were examined with the DEA-Valuation method. Results
showed that overall 70% of the inetficient cornpanies were within the range predicted by

.- -

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

114

Chapter 7 Conclusions and Recommendations

the DEA-Valution method, and quite a number of companies were scarcely outside of
their individual predicted range. Overall, the calculated upper bounds were greater than
the real market value for the inefficient companies 75% of the tirne. For the efficient
companies the Iower bound was less than the real market value 50% of the t h e , however
there were instances where it was not and this could mean that the companies are
undervalued.

In summary, the DEA-Valuation mode1 was tairly adequate in predicting

appropriate market ranges. As with ail valuations, this is another tool that valuators have
at their disposal, and a combination of the different valuation methods wilI give better
ranges of value.

7.1

CONTRIBUTIONS
The contributions of this thesis include novel ways to predict lower bounds for

market value for eficient cornpanies and upper bounds for inefficient ones. This work
expands on the use of DEA for finding peer groups instead of its traditional role of
h d i n g eficirncy measures. Valuation models were developed for both pnvate and
public companies and eflficiency scores, difference indicators. market value ranges. and
comparative peers were provided where appropriate. DEA was able to handle the
multidimensional nature of companies because it ailows multiple inputs and outputs. The
work consists of a group of data. more than 500 companies al1 toether, which can be
used to determine the peer groups.

7.2

LIMITATIONS OF THE DEA APPROACH

In order to apply the DEA-Valuation Methodolo~ythere has to be a suffxcient

number offirms in the peer group. The number of DMUs in the analysis varied tiom 22-

--

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

115

Chapter 7 Conclusions and Recommendations

-

85. The number of similar peers should at Least be three in order to conduct an
appropriate analysis. lncreasing the number of firms in the andysis may improve and
tighten the ranges for predicted market values. IdealIy one wants a set with many DMUs
but with very sirnilar DMUs. Having more DMUs would not p u a n t e e more peers
however a bigger group would increase the probability of having more similar
companies.

7.3

AREASOFFUTUREWORK
The methods discussed in this thesis to determine market values represent an

excellent start and show encouraging results. The above methodology cm be enhanced
and made stronger by applying it to different areas and comparing its results with other

methods. In the following list, areas of future research are addressed and exarnined to
advance the DEA-Valution methodology.
Obtain data on ve-year growth forecasts for cornpanies and nin a DEA analysis
Analyze diflierent sectors rather than just manutcturing to see if the results hold for
the other industries.

Attempt to analyze Internet Company valution, need to somehow rank key-men and
technology contributions, and it is necessary to look at other qualitative and
quantitative factors. Although information of this nature was not available for this
analysis, if it were, the possible models would consist of the important quantitative
tactors that Internet start-ups rely on. These include their burn-rate of cash and the
amount of cash or investments on hand. While one wants to irease the amount of
cash investments in the business, and thus maximize it, it has to be in balance with

VALUINGPRIVATECOMPANIES:
A DEA APPROACH

i 16

--

Chapter 7 Conclusions and Recommendations

the burn-rate. internet companies go through an exorbitant amount of cash, cdled
the "bum-rate". This would be an important tctor to rninimize in a DEA study.
Other fctors to consider are supplier relations, marketability, %lier apps" or unique
t e c h n o l o ~and
~ customer base to make up for the fct that most Intemet companies
scarcely have revenue and typically, no profits.
The weights in this study were not restricted. DEA chose the weights according to
the Company under analysis allowing it to appear the best it could. With expert
opinion. these weights can be restricted and analyzed to test the change in market
value ranges.
According to Renaissance Capital (a research boutique) a valuation analysis can be
carried out without using ail the companies tiom the same 4-digit SIC code. The
argument is that some present day public companies are so large and diversified into
many t o m s of businesses that four digit SIC codes are no longer of significance.

For example, Molson's Canada is a brewer. i-e. Manufacturing, but at the same time
it owns the Montreal Canadians Hockey CIb. Lc. Non-manutcturing. This might

lead to an interesthg study if a lot of companies with varying SIC codes can be
included in the same studyA complete valuation using diffrent methodologies such as DCF. CF. asset

valuation. with al1 the appropriate discounts applied would be the best study. As
well as being the best way to compare al1 the methods against each other
Find and set optimal cutsff values for the distance and the efficiency similanty
when analyzing the minimum and maximum of the inefikient companies. Right
now the distance is set to 1, and only companies that have a distance of less than 1

VALUING PRWATE
COUPANIES:
A DEA APPROACH

117

Chapter 7 Conclusions and'Recommendations

are analyzed as peers. For the efficiency scores, try to determine a percentage range

that is appropriate for peer companes, For example, if the subject Company has an
efficiency score of 0.75 only examine other cornpanies that are within 10% o f this
score,

7.3.1 New Applications for DEA

The ranking of the reputahon of the underwriters are based on pricing, selling of
their allotment, post-issue pnce stability. market making, analyst coverage/publicity,
ability to underwrite subsequent debtlequity issues. The DEA approach can be used
to study [PO performance and also to determine the best undenvriters based on the
above factors.
Also, as previously mentioned, DEA c m be used just to find the best comparable
fims, and instead of trying to estimate a market value range based on the found
comparable peers, the known PIE ratios, gowth forecasts and discount rates of the
peers should be used in the estimation of value of the private ones.

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

118

References
-

REFERENCES:
[BANK841

Banker. R-D.. Chames, A., Cooper. W. W., "Some Models for Estimating
Technical and Scale Efficiency in Data Envelopment Analysis." Management
Science. ( 1984). 30(9): 1078-1092.

[BERG981

Bergendahl. G., "DEA and benchmarks an application to Nordic Banks,"

Annals of Operational Research. ( 1998), 82: 233-249

[BOATS1 1

Boatsman, J.. Baskin. E-."Asset Valuation with lncomplete Markets.''

The Accounting Review, (1 98 1) 56:38-53,

[BOOTOO]

Booth, L.. Class Notes on Mergers and Acquisitions. (2000).

[BOOT981

Booth. L-, "What D ~ v Shareholder

~ s
Value.*' Federated Press-Creating
Shareholder Value Conference. October 28. 1998.

(BOOT98al

Booth, L., "The Capital Asset Pricing Model: Equity Risk Premiums And The
Privately Held Business," Journal of Business Valuation- Proceedings of the
3rdJoint Business Valuation Contrence of the Canadian Institute of Chartered
Business Valuators (CICBV) and the Arnerican Society of Appraisers (ASA).
Montreai, (1998), 1-36-

[BYRN94]

Byrne. A., Rees, W., "The Time Series Behaviour of Initial Returns on U.K.
IPOs." Journal Of international Financial Markets, Institutions & Money,"
(1994), 4(3-4): 81-100.

[CAMP921

Campbell, I.R., "Business Valuation a Canadian Author's Perspective,"

Journal of Business Valuation - Proceedings of the 1 1" Business Valuation
Conference, Quebec City, ( 1992)- 135- 159.

[CBCOOl

CBC Newsworld Broadcast, July 22, (2000)

[CDNXOO]
[ C H U 9 3j

Chatie. D., "Option Pricing As A Proxy For Discount For Lack Of

MarketabiIity in Private Company Valuations: A Working Paper." Business
Valuation Review, December ( 1993).

(CHAR78)

Chames, A., Cooper, W.W., Rhodes, E.L., "Measuring the Efficiency of

Decision Making Units," European Journal of Operational Research, ( 1978),
2(6):429-444.

[CHAR941

Chames, A., Cooper, W. W., Lewin, A-Y., Seiford, L.M., Data Envelo~ment
Analvsis: ~heorv:Methodoloev and Atmlications, Boston: Kluwer ~ f a d e m i c
Publishers (1 994)

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

119

References
[CS1891

C -S-1.. How to read financial statements, Toronto: The Canadian Securities

tnstitute ( C S 0 ( 1989).

[DENNOO]

Dennis, R., "Question and Answer on Selling A Business,'' The Washington

Post, Febniary 21,2000-

[ETT097]

Ettore. B.. '"Too Much of a Bad Thing," Management Review. December

1997.

[~-571

Farrell. M.J., "The Measurement of Productive Efticiency." Journal of the

Royd Statistical Society. Series A. ( 1957), l20(3):2S-290.

[F[RT971

Firth. M.. Liau-Tan. C.K. --Signaling Models and the Valuation of New
Issues: an Examination of IPOs in Singapore," Pacitic-Basin Finance Journal.
( 1997). 5(5): 5 11-526.

[FISOOO]
[FITCSS]

Fitchett. G-A, "Financing Closely-Held Businesses" Journal of Business

Valuation- Proceedings of the 9th Biennial Conference of the CICBV,
Toronto. ( 1988), 11-20.

[-971

Harben. T.. "Intangibles Count in a Company's Valuation." Electronic

Business. November 1997.

mEND88l

Henderson, A.R., "Key Strategies in Selling a Privately-Held Business."

Journal of Business Valuation- Proceedings of the 9th Biennial Conference of
the CICBV Toronto, (1 988), 39-65.

[HUNT751

Hunt. P .."Funds Position: Keystone in Financial Planning," Harvard Business

Review. May-June 1975.

[IBB094]

Ibbotson. R., Sindelar, J.L.. Ritter, J.R., "The Market's Problems With The
Pricing of IPOs..' Journal of Applied Corporate Finance, (1994), 756-74.

(JAiN991

Jain, B.A., Kini, O., "On Investment Banker Monitoring in The New Issues
Market," Journal of Banking & Finance. ( t 999) 23:49-84.

[JENS76]

Jensen. M., Meckling W., T h e o r y Of The Firm: Managerial Behavior,

Agency Costs And Ownership Structure," Journal of F inancial Economics.
(1976) 3:305-360.

[KAPL951

Kaplan. SN., Ruback, R.S., T h e valuation of cash tlow forecasts: an

empincal analysis," Journal of Finance, ( 1995),50: 1059- 1093.

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

120

References
[KIM951

Kim, J-B.. Krinsky, 1.. Lee, J., T h e Role Of Financial Variables In The
Pricing Of Korean Initiai Public Offerings." Pacific-Basin Finance Journal.
( 1993, 3(4):449-464.

[KIM981

Kim, I., Hendry, L-C., "Ushg DEA To Assess NATO Burden-Sharing,"

Journal of the Operationai Research Society. ( 1998)-49928-236.

[KIAM991

Kim, M-. Ritter J.R.. "Vaiuing IPOs." Journai of Financial Economics. ( 1999).
53:4O9-43 7.

(IUE1961

Klein, A-, "The Association between the intbrmation contained in the

prospectus and the pnce of initial public offenngs." Journal of Financial
Statement Analysis. (19%), 223-10-

[LELA77

Leland. H.. Pyle. D.. "Informational Asymmetries. Financial Structure And

Financial Cntermediation," Journal of Finance, ( 1977). 32:V 1-387.

[LEVI951

Levis, M.. Thomas, D.C., "Investment Trust IPOs: lssuing Bebviour And
Price Pertormance. Evidence From The London Stock Exchange," Journal of
Banking & Finance, ( 1995). 19(8): 1437-1458.

[LITT591

Little, A.J., -'Valuation of Goodwill." The Valuation of Private Business and

Protkssional Practice, Toronto: Canadian Institute of Chartered Accountants
(1959). 1-6,

(LIUN971

Ljungqvist, A.. "Pncing initial Public Otierings: Further Evidence From

Germany." European Economic Review, (1997), 4 1:1309-1320.

[LYON891

Lyons, R.P., Wilczynski, M. J.. "Discounting Intrinsic Value." Trusts and

Estates. Febniary ( I989), 22-27.

[ME001
[MERC971

Mercer. C.. Ouantifvine Marketabilitv Discounts. @city: Peabody Publishing

(1997), 345-370.

(MORR921

Moms. I.T., -'Valuation Perspectives of a Venture Capitalist." Journal of

Business Valuation - Proceedings of the 11" Business Valuation Confrence,
Quebec City, ( l992), 83-90.

[MULL901

Mullen, M., "What is a Brand Name or Trademark Really Worth - How Can
that Vaiue be Measured?" Journal of Business Valuation- Proceedings of the
2nd Joint Business Valuation Cont'erence of the CICBV and the ASA,
Vancouver, (1WO), 203-2 12.

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

References

[NEED861

Needhant. J., Young, A., "The Leading Edge In Mergers And Acquisitions."
Mergers & Acquisitions. November I December 1986.

[OANDOOI
[OVEN591

Ovens. G.. Taluing a Privately Owned Business... The Valuation of Private

Business and Professional Practice, Toronto: Canadian institute of Chartered
Accountants (CICA) (1959), 1 1-1 8-

(OVEN721

Ovens. G.. "lntroduc tion to the Business Valuation Field." Solvin- the
Valuation Problem, Toronto: h e CICA (1972). 7-9-

(PIES951

Piesse. J-. Townsend. R.. T h e Measurement of Productive Eficiency in UK

Building Societies." Applied Financial Economics. (1 995). 5397407.

[PRAT931

Pratt. S.. valu in^ Small Businesses And Protssional Practices, znded.,
Homewood: Dow Jones-Irwin (1993).

[PRAT961

Pratt, S.. Valuing A Business: The Analvsis and Amraisal of Closelv Held
Companies, jrded.. Homewood: Dow Jones-Invin ( 1996).

[RITT841

Ritter, J.R.. 3ignalinp And The Valuation Of Unseasoned New Issues: A

Comment,'' Journal of Finance, (1984). 39: 123 1- 123 7-

[RUBA961

Ruback. S.N.. Richard, S., --The Market Pncing Of Cash Flow Forecasts:
Discounted Cash Flow vs. The Method Of Comparables." The Bank of
Arnerica Journal of Applied Corporate Finance. ( 1996) 8(4):45-57.

[SEDAOOI
[SEGU971

Seguin. P.J.. Srnoller. M.M.. "Share Pnce and Mortality: An Empirical

Evaluation of Newly Listed NASDAQ Stocks,-' Journal of Financial
Economics, (1997). 45333-363.

[SIMA001

Simak, P.C.. Paradi. J-C., "How To Estimate The Market Value Of A Private
Company - A DEA Approach." University of Toronto working paper (1999).

[STOCOO)
[STOU981

Stoughton, N.M., Zechner, J., T P O Mechanisms. Monitoring And Ownenhip

Structure," Journd of Financiai Economics, (1998), 49:45-77.

[THAN921

Thanassoulis, E.. Dyson, R.G., "Estimating Preferred Target Input Output

Levels Using Data Envelopment Analysis," European Journal of Operational
Research, (1992), 56:80-97.

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

i 22

References

Weaver, W.C., "Discounts And Other Adjustments To Fair Market Value

Estimates," Orlando: William C. Weaver ( 1 W8), 1 16.

Wise, R.M., "Discounts in Arriving at Share Values of CIosely-Held

Corporations'' Journal o f Business Valuation- Proceedings o f the 2nd Joint
Business Valuation Confrence of the CICBV and the ASA. Vancouver,
( 1 WO), 127- 148.

Zhu. J., "Multi-Factor Perhrrnance Measure Mode1 With An Application To

Fortune 500 Companies." European Journal of Operational Research, (2000)
123: 105-124.

VALU~NG
PWATE COMPANIES:
A DEA APPROACH

123

Bibliography

BIBLIOGRAPHY:
AL,AL. Lerme, CS., Seiforci, L.M., 'romponents of Eficiency Evaluation in
DEA." European Journal of Operationai Research. (1995), 80:462-473
Berger. A.N.. Hurnphrey. D.B.."Efficiency of Financial Institutions:
International Survey and Directions For Future Research.' European Journal of
Operational Research. ( 1997). 98: 175-2 12
Brocken. P.L.. Charnes, A.. Cooper. W.W.. Huang. Z.M., Sun. D.B.. "Data
Transfomation in DEA Cone Ratio Envelopment Approaches for Monitoring
Bank Pertormances," European Journal of Operational Research. ( 1997). 98:
350-268.
Campbell. I.R.. Low. RB., Murrant, N.V.. The Valuation & Pncinrr of
Privatelv-Held Business interests, Toronto: Canadian Institute of Chartered
Accountants ( 1990).
Chi1ingerian. J.A., "Evaiuating Physician E fficiency in Hospitals: A
Multivarate Analysis of Best Practice." European Journal of Operational
Research. ( 1995),80:548-574.

Cook. W., Roll, Y., Kazakov, A.. "A DEA mode1 for Measuring the Relative
Eficiency of Highway Maintenauce Patrols." Journal of Information. (1990),
?8(2):113- 124.
Dauderis, HI.Financial Accounting: An intro to decision rnakinq, Toronto:
Holt, Rinehart and Winston of Canada, Ltd. (1987).
Droms. W.G.. Finance & Accounting for Nonfinancial Manarrers. al1 the
basics vou need to know, 4'h ed., Reading: Addison Wesley Longman Inc.
( 1998).
El-Mahgary. S., "Data Envelopment Analysis: Visualizing the Results,"
European Journal of Operational Research. ( 1999, 85 :700-7 10.
El-Mahgary. S., -'Data Envelopment Analysis - A Basic Glossary." Journal of
Operational Research Society, (19 9 3 , 8(4): 15-22.
Mikkelson, W. H., Partch, M.M., Shah, K., -'Ownership And Operating
Pertbrmance Of Companies That Go Public," Journal of Financial Economics,
(1 997) W:281-307.
Morck, R., Shleifer, A., Vishny, R.,"Management Ownership And Market
Valuation: An Empirical Analysis,?' Journal of Financial Economics, (1988)
20293-3 16.
-

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

124

Bibliography
[PAPA971

Papagapoou, A., Mingers, J., Thanassoulis, E.,"Would You Buy A Used Car
With DEA?" Journal of Operational Research Society, (1997), 1O( 1 ): 13- 1 9.

[PAST97]

Pastor, L M . . Perez F., Quesada, J., "Efficiency Analysis in Banking Firms:

An International Cornparison," European Journal o f Operational Research.
(1997), 98: 395-407,

[TATJ971

Tatje, E.F.. Lovell, C.A.K., T h e Sources of Productivity Change in Spanish

Banking," European Journal of Operational Research. ( 1997), 98 :3 64-3 80.

(TRAC991

Tracy. LA. How To Read A Financial Re~ort.For Managers. Entrepreneurs.

Lenders. Lawvers & tnvestors, 5" ed., New York: John Wiley & Sons Inc.
( 1999).

GLOSSARY
Another name for the efficiency measure as obtained tiom the CCR model.
also known as eff?ciency
The act or process of detemuning the value of something
BCC Model

This model measures technical eficiency as the convexity constraint

ensures that theComposite unit is of similar scale size as the unit being
measured; unlike the CCR model. the BCC allows variable returns to scale

Beta

A measure of the systematic risk of an asset. Compares the price volatility

of a particular issue to the movement of the market in general. If the
difference is great security has a high beta; if low. has a low k a .

Binomial
Model

A model where the price of an asset is monitored over successive short

perods of time. In each short penod it is assurned that only two price
movements are possible

Binomial Tree

A tree that represents how an asset price can evolve under the binomial
model

Black-Sc ho les
Model

A model for pricing European options on stocks. developed by Fischer

BPS

Book value per share

Business
Enterprise

An organization pursuing an econornic activity on a 'for profit' basis

Business
VaIuation

The act of determinhg the value of, or the estimated value of, a business
enterprise. or an interest therein

Cal1 Option

A hancial instrument that gives the owner the option to buy an asset at a
certain price by a certain date

Capitalization
Rate

The required rate of retum that is used to convert income/cash tlow into
value. The inverse of the cap. Rate is referred to as the multiple or
multiplier.

CAPM

The capital asset pricing model is a model relating the expected retum on
an asset to its beta.

Black, Myron Scholes, and Robert Merton

. .

VALUING PRIVATE COMPANIES:A DEA APPROACH

126

Giossary

CCR Mode1

The CCR ratio mode1 calculates an overall efticiency for the unit in which
both its pure technicd efficiency and scale efficiency are agzegated into a
single value- The obtained efficiency is never absolute as it is always
measured relative to the tieid

C losely-Held
Corporation

A term used to describe a corporation whose shares are neither listed for
trading on a recopized stock exchange nor traded in an over-the-counter
market- Technically it has less than 5 1 shareholders (Canada).

Comparable
Firrns

Public companies that are sunilar in nature to private companies or other

companies,

Constant
Returns to
Scale (CRS)

A unit operates under constant returns to scale if an increase in inputs

results in a proportionate increase in the output levels. I f the inputs values
for a unit are al1 doubled, then the unit must produce twice as much outputs

DEA

Data Envelopment Andysis is a decision makinp tool based on linear

progr.ammio,gfor measuring the relative efficiencies of a set of comparable
units

Depreciation

The systematic assi_miment of the cost of tangible assets to expense

Discount Rate

The annualized dollar r e m on a treasury bill or similar instrument

expressed as a percentage of the tinai face value. Or, a rate of retum used
to convert a rnonetary sum, payable or receivable in the tiiture, into present
value

Discounted
Cash Flow

An accounting technique for estimating the present value of anticipated

future income and expenditures. such as earnings from loan principal and
interest payments, and income fiom investment securities. It is calculated
as either net present value, which expresses funue cash flows in terms of
curent money by applying a discount rate to friture receipts, or intemd rate
of return- which figures the average annual yield or return on capital of an
investment or a bank loan over its expected iittime

Dividend

A cash payment made to the owner of a stock

DMU

Decision Making Unit. This term emphasizes the fact that the focus was not
on profit generatuig enterprises. but rather on decision making entities

DRS

Decreasing returns to scde

EBIT

Earnings Before Interest and Tax

EBITDA

Earnings Before Interest, Tax, Depreciation, and Amortization

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

127

Glossary
Economies of
Scale

The decreases in combined operating expenses and/or increases in

combined revenues?which result tollowing the mergjng of two or more
business operations. They are created by hctors such as the elimination of
duplicate overhead costs. the acquisition of a secured source of supply or
enhanced combined purchasing Fower. more eEective utilization of
operating facilities. and so on.

Efficiency

A general t e m ofteo associated with pertorming activities as well as

possible.

Efficient
Frontier

A fiontier or surface determined from the best-observed or best practice

DMUs.

Enterprise
Value

The market value of equity plus its book value of cash minus cash.

Envelopment
Surtce

Please see Efficient Frontier

EPS

Eamings per share

Exogeneously
Fixed Factor

An input or output that is beyond the control of the DMU manager

Fair market
value

The highest price available in an open and uarestricted market between

infomed and prudent parties, acting at arm's length and under no
compulsion to act, expressed in t e m s of cash.

Financial Ratio

Usuaily a ratio of tinancial information. aiming to capture a single

dimension of cocporate performance

Financial Risk

The portion of overall business risk associated with the financial structure
of a business

Forced
Liquidation

The sale of the assets where an imrnediate cessation of the business and
disposition of assets is assumed on an 'as is/where is' basis. The forced
liquidation of assets at auction.

Going Concern

A business enterprise that is both conducting operations at a given date, and

has every reasonable expectation of doing so for the foreseeable tture after
that date. An accounting terni used to indicate that a Company is operating
nonnaily.

Going Concern
Value

The present value of al1 future benefits expected to accrue from ownership,
where a business operation is expected to continue to operate either for a
finite period, or indefinitely into the tture

VALUINGPRIVATE COMPANIES:
A DEA APPROACH

128

Glossary

Goodwill

The diEerence between going concern value and the sum of the net angble
assets and identifiable intangible assets
Degree of similarity between the units in the field
Indicates that an inefficient unit is made efficient through the proportional
reduction of its inputs while its output proportions are held constant. The
CCR mode1 yields the same efficiency regardless of whether it is input or
output-oriented, this is not the case with the BCC mode1

Intrinsic Value

A notional market value, based upon rates of return required by investors

gven economic and business conditions existing at the valuation date.

without consideration of possible synergistic benetits that might accrue in
ditTering degrees to am's length purchasers

IPO

Initial public oflring

IRS

Increasing returns to scale

Lambda Values Results Liom a DEA analysis that indicate the level of contribution of
efficient firms to the virtual DMU created by the projection of an efficient
of an inefficient DMU ont0 the fiontier
Liquidation

The process of converting assets into cash.

Liquidation
Value

The net amount of money, ifny, avaiiable to equity owners following a

voluntary liquidation, a reorganization of a business pursuant either to a
proposa1 to creditors, or a liquidation of a business pursuant to a
receivership or other proceeding under The Bankruptcy Act.

Liquidity

The relative ease of converting non-cash assets into cash.

Market
Capitalization

The total value of a company, derved by the price of a share multiplied by

the number of shares the company has outstandhg

Market to
Book Value

Ratio of the market value of a security to its book value. An indication of

pertbrmance of common stock equity, alter accounting for such factors as
dilution of stock value after a merger and market reaction to unexpected
gains or losses and changes in operating income

WB)
Marketability
Discount

An arnount deducted frorn what would otherwise be the value of a

shareholding in a company to reflect illiquidity, or lack of marketability

Minority
Discount

The reduction tiom the pro rata portion of the en bloc value of the assets or
ownership interests of a business as a whole to reflect the disadvantages of
owning a non-controlling interest

VALUINGPRNATECOMPANIES:
A DEA APPROACH

129

Glossary
Multipliers

Piease retr to definition for Weights

Net Book
Value

With respect to a business enterprise, the difference between total assets

(net of depreciation, depletion, and amortization), and total Liabilities of an
enterprise as they appear on the balance sheet. The term is synonymous
with each of the term' s book value, net worth and shareholders' equity.

Nominal Rate
of Return

A discount rate. or capitalization rate, that includes both an inflation

component and a real rate of return.

Notional
Market

A term used to describe circumstances where it is necessary to determine

tir market value. fair value. or some other value in the absence of open
market negotiations

Open n market

A term used to describe the market in which arm's length, negotiated

transactions take place-

OutputAn inefficient unit is made eficient through the proportional increase of its
oriented Mode1 outputs, while the inputs proportions remain unchanged

Over-t hecounter Market

The trading of securities or cornmodities not Listed on au established stock

exchange

Peer Group

The set of efficient DMUs against which inefficient units are compared.
The efficient targets are a linear combination of these peers.

Piecewise
L inearity

An efficient tiontier is piecewise linear when the underlying production

nction is approximated through intercomected h e a r segments

Price

The consideration paid in a negotiated open market transaction involvig

the purchase and sale of an asset

Price to
Eamings Ratio
(Pm

The relationship between the current price of a stock to the eamings over
any earnings period. The PIE is calculated by dividing the current stock
price by the earnings per share or dividing the market value by earnings per
share for the most recent 12-month period. It is used by investors to gauge
the market demand for stocks. Companies with a high P/E have greater
earnings potential than low PE stocks, but also are more volatile and
somewhat riskier

Privately-Held
Business
Interest

An en bloc, control, or minority shareholding in a private company, or al1 of

part of the business of a private company represented by its net operathg
assets or al1 or one or more partnership interests Ui a partnership or a
proprietorship

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

130

Privately-Held
Company

does not offer its shares to the public

Public
Company

A Company whose equity shares and/or issued debt are traded on an

established stock exchange or in an over-the-counter market.

Put Option

An option to sel1 an asset for a certain price by a certain date

Rate of Return

An amount of income realized or expected on an investrnent, expressed as a

percentage of that investment

Real Rate of
Return

A discount rate. or cap. Rate. that excludes consideration of inflation, as

contrasted with a nominal rate of return

Research
Boutique

Company that specializes in t-mding comparable companies

Risk Free Rate

of Retum

The prevailing rate of r e m on money market investments such as long

term govemment bonds. conventional tim mortgages, and hi& quality long
term corporate bonds. These are investments which are generally fke from
substantial credit risk, are priced to reflect a relatively long term holding
period. and hence. aside Eom intlation related purchasing power risk, put
the investor at relatively little rsk. These securities typically are k e d in
income as well as in principal and, hence, generally do not offer m y
potential for capital growth.

RTS

Returns to scale

Rules of
Thumb

Commonly used criteria within a specific industry to Bauge an element of

business value

Scale
E fficiency

A unit is said to be scale efficient when its size of operations is optimal so

that any modifications on its size will render the unit less efficient; value
obtained by dividing the aggegate efficiency by the technical efficiency

Shareholders'
Equity

The aggregation of a company's paid-in capital, retained earnings,

contributed, appraisal and other swpluses.

Slack Variable

Result f o m a DEA analysis and indicates the level of input minimization or

output augnentation that is required for the inefficient tim to reach the
envelopment surface following the proportionai change indicated by the
efficiency score

Technicai
Efficiency

An efficiency measure that ignores the impact of scale-size by comparing a

A business that is incorporated. has less than 5 1 shareholders (Canada), and

DMU only to other units of similar scale, computed using the BCC mode1

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

131

Glossary

Threshold Rate
of Return

A benchmark, point in time related predetermined rate of r e m on equity

criteria that is applied to a projected afier tax cash tlow that includes both
inflation and real growth, where leverage and related interest bearing debt
considerations are e x l u d e d and where purchaser perceived net economic
value-added is taken into account

Treasury Bill

A short-term non-coupon bearing instrument issued by the govenunent to

t-inance its debt

Vaiuation

The act or process of determining the value of something

Valuation Day
(V-Day)

The date proclaimed in the Income Tax Act (Canada) on which capital
gains taxation in Canada came into et'fect. With respect to privately held
business interests. this date is December 3 1. 1971.

Value

A return or equivalent in goods, services. or money for something

exchanged

Variable

An input or output included in a DEA mode1

Variable
Retums to
Scale (VRS)

Can be increasing, decreasing, or constant. A proportionate increase in

inputs can result in the non-proportionate increase in outputs

Warrant

An option issued by a Company or a fuiancial institution. Cal1 warrants are

tiequently issued by companies on their own stock

Weighted
Average Cost
of Capital
(WACC)

A discount rate that is applied to a projected atier tax (generally

discretionary) cash tlow that includes both intlation and 'real. gowth. It is
detemiined by the appropriate capital structure (Le. interest bearing
debtlequity structure) for a given business. It weighs the c o a of said
appropriate indebtedness and the cost of equity capital, where the cost of
equity capital is taken to be equal to the threshold rate of return. It assumes
an ability to deduct interest expense when calculating taxable incorne.

Weihts

Results fiom a DEA analysis and indicates the level of contribution each
variable has in the determination of the DMU7sefficiency score.

Working
Capital

The amount by which current assets exceeds current liabilities

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

132

List of Abbreviations
-

- -

LIST OF ABBREVIATIONS
BPS

Book Value Per Share

Capex

Capital Expenditure

CAPM

Capital Asset Pricing Mode1

CDNX

Canadian Venture Exchange

CF

Comparable Firm

CF1

Cash Flow

CF0

Cash Flow tom Operations

CICA

Canadim Institute of Chartered Accountants

CICBV

Canadian [nstitute of Chartered Business Valuators

CRS

Constant Returns to Scale

CS1

Canadian Securities Lnstitute

DCF

Discounted Cash Flow

DEA

Data Envelopment Analysis

DMU

Decision Making Unit

DM

Decreasing Returns to Scaie

EBITDA

Earnings Before Interest, Tax. Depreciation, and Amortization

EPS

Earnings Per Share

ESOP

Employee Stock Ownership Plan

EV

Enterprise Value

FCF

Free Cash Flow

GAAP

Generally Accepted Accounting Principles

IPO

Initial Public O f f e ~ g

---

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

133

List of Abbreviations

IRS

Increasing Returns to Scde

M&A

Mergers and Acquisitions

m
l

Market to Book

MC

Market Capitalization

NI

Net Income

NS

Net Sales

W C

Net Workig Capital

OP

Offer Price

Pm

Pnce to Earnings

PER

Price Eaniings Multiple

POP

Preliminary Offr Price

PPA

Pure Play Approach

RA

Redundant Assets

RFR

Risk Free Rate

SE

Shareholders' Equity

SEC

Securities Exchange Commission

SIC

Standard Industriai Classification

TA

Totai Assets

TL

Totai Liabilities

TSE

Toronto Stock Exchange

VRS

Variable Returns to Scaie

WACC

Weighted Average Cost of Capital

APPENDIX A
Public Companies DATA 1998
SIC 2000-3999

-- - C

TOTAL ISHARE~~NET
NET SALES CASH
LIAB-EQUITY
-ME
/REVENUE FLOWS
(in thousands)
206,480
40,532
165,948
13,952
207,067
1,995
1,779
1,181
598
-283 (

TOTAL

ASSETS

1 A. L. VANHOUTTE
L 2095
A.E. VENTURES LTD 3714

1 3661

ABL CANADA iNC

AGRO PACIFIC INDU 2048

AIMGLOBAL T
E
w 3829
1 2 AINSWORTH LUMB# 24 1 1
13 ALCAN ALUMINlUM 3353
14 ALGO GROUP INC 3729
1 5 ALGOMA STEEL INC 33 12
16 ALLIANCE FOREST 26 1 1
I 7 ALPHA COMMUNICI' 273 1
18 ALTA NATLlRAL HE: 2833
19 ALTAREX CORP
1 2834
20 ALTERRA RESOURC! 384 1
2 1 AMR TECHNOLOGIE 28 19
22 AMT FINE FOODS LT 20 13
23 ANDRES WINES L q 2084
2834
24- ANGIOTECH
25 AQUA 1 BEVERAGE 4 2086
1

11

If

PW

1 55.6 17

44,467
9,435,500
9266
108,383
563,100
3,019
6,577
7,100
340
76,024
11,173
79,826
140,743 1
2,879 1

87.1 12
434.862
15205.956
84,550
1-521,000
1.77 1,800
1,238
1,956
15,160
2.146
75-913
1,592
126,085
27,508
2 10

10.887
357.966
6.729.876
63220
1.1 85,000
8 14.000
886
613
2.5 1 3
1.945
33.760
696
72.58 1
767
97

76.225
76.8%
8,476,080
2 1330
336,000
957.800
352
1.343
12,647
20 1
42,153
896
53,504
26.741
113

4,625
-9,479
612.784
-15,700
-59,000
25.900
-308
-1 10
-13.1 16
-667
596
164
6,732
-6.73 1
-50 1

5.6 13
17.425

-3,522
18.732

-2.395
-1.016

78,478
7,62 1
349.056
24.354
1 1,962,346 l297,7S 1
196,136
-3,467
l372,OOO
74,000
1,085.100
138,200
1,927
-227
107
3,123
-12,736
50
O
-6 18
3,609
49.655
25 1
6,188
1 1,078
1 18,668
2277
-6.325
387
-485

- 1,869

AQUARISJS COATIN~3479
ARCTIC GROUP INC 2097
7

2.1 15
38.002

2.09 1
36.157

ASIA PACYFIC CONCI 3273

1,888
2,743
246
ASTAWARE TEcKNC/ 274 1
3,178
1.385
2 I4
AT PLASTICS INC 1 282 1
153,632
505.62 1
3 1 7.9 13
33 1ATS AUTOMATION 13569 1
9 9 ~ 5 3 10
423,5191
155.3131
34 AVANTE TECHNOL~3634
2.302
1,115
592
5 1.803
35 AVCORP INDUS*
45,730
86,856
3728
36 A W A TE CH NO LOG^ 3646
476
7 16
2,429
165,498
18.8 19
37 AXCAN PHARMA IN( 2834
88,823
38 B.C. REPORT M A G 272
284
2,053 1
235
~ 1
39 BALLARD POWER SA 369 1
690.348
3.60867 1
62.656
40 BANFF BREWERY ca 2082
332937
1.378
4 1 BCS TECHNOLOGY d 366 1
9 14
996
1,027
42 BELL CANADA m l 3 6 6 1
2,923,589 2385264
l,37 1,120
43 BESTAR INC
35.495
i2511
23,108
26.096
44 BESTAR INTERNATI~2836
10,769
1,488
3 -532
45 BETA BRANDS INCO( 2064
59.54 1
63,737
47,999
46 BIG ROCK B R E W 2082!
3 1,993
~
19,445
14362
PHARMA 12835 4,708260
629,990
~~IBIOCHEM
1 5 1.369
48 BIOTECH HOLDING^ 2834
7.987
22.80 1
3.778
49 BIOVAIL CORPORAT 2834
1,404,647
307,036 (
2 8 . 4 16
50 BiT INTEGRAnON T,3577
5,486
10,771 1
8,254
L

--

- -

.-

2,497
1.171
187.708
268,2061
523
35,053
240
70,004
-1.818
627,692
605
82
538.325
9,399
7337
4,196
17.73 1
478.62 1
4,209
78,6 19
2,s 17
--

-803 1
-1,881
10,768
27.3621
-625
2-810
-374
1,540
1,885
765
-43 2
-1.21 1
-65.6 1 1
5,432
-697
-3,4601
289 1
1 14,774
-3,014
69,755 1
-5,143 [
1

4.3 74
9.574

-3 67
1

769 1
469 [
229.976 f

27.984

777
66.903
1,875
43.404
2,498
25.078
1 79
2,176
77 1.506
69.155
6,280
63,870
25.185
33532 1
1,356
1 73294
2,955

-563
5,032
-3 1 1
2,240
-100
- 18,826
-3 72J
-9 13
-7 I,29 1
6,808
157
-1 16
2,3 84
129,087
-1,723
77.367
-1,3 19

-84

- 1.65 1

51 1BLUESTAR B
A-!
3692
44,4041
133,332
83,377
49,955
-8.816
135,366
-3,62#
521 BOLIDEN LIMITED 3339
40 1.3 121 2-831370 1,828,276 1,002,994 -1 16,239
1,618-176
4,195
53 ( BOMBARDIER [NC 3743 7..502..049 ,1 L 0-575200 7,685,900
2
.
8
8
9
.
3
0
0
420200
8
.
5
0
8
.
9
0
0
9
19-70f
= - -.
-54[BORDERCHEMICAL 28 19
3,7251
IO,174
1,515
8,659
352
14,387
79(
55 i BP AMOCO P.L.C
29 1 1 619,712,0001 129,775.100 65..6O0..161 64,174.939 5.006,708 128.595.606 14,732,,17- 5
56 BRAMPTON BRICK 4325 1
33,4661
64,523
20,389
44,134
7,042
34,9491
10,335
57 BRICK BREWING CO 2082
1 5,421 '
19,835
7,775
12,060
694
19,9231
1,57C
58 BUDD CANADA W C 3714
116,772
121.603
40.977
80,626
7,359
349,036) 13.72I
59 BUHLERIMlUSTRlE 3523
66.665
61,139
16.349
44.790
7,1 13
89,1941 11.194
60 CAE INC
3812
917,070
928,179
651,726
276,453
70.236
922.369
87,541
G I 1 CALIAN TECHNOLO 3663
33-191
45,533
22.936
22.597
1,408
79,928
3-00;
621 CAMCC iNC
363 1
100,000
192,375
117,134
75.2411
8,979
565.305
24.1 1 S
63 iCANADA BREAD CO 205 1
449,753
338272
107.736
230.5361
2,896
508,445
28,883
46,8981
97,932
30,659
ij52l
100.589 1
64i CANADJAN BANK ~d 2796
8.50
67273f
370
10,155
58 1
65 iCANADIAN MANOIR 3433
3.5 1 1 '
11.815
9,474
2.34 1 1
426.808
1 13,940
38 1.988 1
28,003
495.928
29 1.544
39-501
66j CANADIAN MARCO? 3663
177.740
645,753
445,882
199.871
26,066
1 ,000,700
52.0 11
67 1 CANAM MANAC GR^ 3441
68f CANFIBRE GROP LI 263 1
9 1.809
362265
33 1.5 15
30.- 749
-6.334
284
-3,564
27415 1
1.864295 1,438,247 426.0481 -203.731
1.632.273
69 !CANFOR CORPORAT 2421
'7 1,921
38.300
10,159
70i CANGENE CORPORA 2836
33 1,802
30,1831
1 1,000
45.784
15.60 1
7 1 i CANHORN CHEMICA 2869
1,7921
6.3 82
4,858
1.524(
-2.701
4,155
-1,827
-46
8.490
85
4,825
72i CANTOL LIMITED 2842
2,108
6.508
1,683
73 ;CARBITE GOLF [NC 3949
24.054
1.424
9240
12.271
11237
367
1,034
78
-2,401
212
-2.5 1 5
2.065
279
67
74 i CARDIOCOMM SOLI.384 t
75 1 CASCADES INC
26 1 1
595,430 2485.000 1,867,000 6 18,0001 45.000 7.527.000r l S5.00C
1.469.1951 1 32.4 1 8
571.417
76 CCL KNDUSTRIES IN(2844
44.394
1 .128,531
857.1 14
698,733
471
296
109
1861
-206
289)
-1 la
77 : CD ROM NETWORK 1 273 I
78 ;CEAPROINC
1 2844
10.546
- 1.654
300f
-1.088t -14,260
479
1,5671
79 CELESTICA MC
3674 2,854,867 2,513222
1,193,561( 1.3 19.66 1
-74.442 4.990.12 1 1 130326
3081
-1,403
80t CELL-LOC INC
3663
1421
2,271
-1,441
28,669
2.4 13
3.835
28 1
4.2
1
2
-191
2,8981
2,765
7-110
8 1 CERAMIC PROTECTI 3259
O
-518
-133
-79
3.053
3,1863
82!CFE MDUSTRIES
3999
7,887
5.368
150,2411
10,106
42,542
22,4651
65,007
83 :CHATEAUSTORES 233 11
45,927
15.001 1
-150
1.106
-150
8.792 1
54 j CHEMBOND LI
2.9 17
9,898
- - MIT^
- 289 1 1
85 :CIE-NERGY PLY-FOI3086
918
542
528
14
-873
2.066
436
86 !CIMATEC ENVIRON 3564
4,040
1.9 10
283
1,627
148
1.493
232
111.912
42.303
642.714
87 ' CMRAM LNTERNATI 3652 L,0%,69
433,063
423,7I1
856,774
1,472
1 ,O31
2 1.464
88 1 CIRCA ENTERPRISE 366 1
6,601
1 1.660
5,113 7..6
1.942
22.00 1
5,330
1,561
89 1 CIRCUIT WORLD COI 3672
8,962
289
14392
3.841
476
6 1.320
50,508
65,775
15,2671
90 i CLEARLY CANADIA:
2086
5,135
9 1 :C-MACINDUSTRIES 3679
29 1.127
607.83 1
347,s141 260.3 1 7
25.4 18
642,983
47,363
92ICME TELEMETRIX 384 1
27.749
5.024
1,295f
3.729
4,139
1.250
-3.826
93 !CML INDUSTRIES LI 2621
7,l09
19,702
12,2241
7,478
21
33.152
1,159
-1,511
94 I COM DEV INTERNAI
3663
252-315
105-496
56,6731
148,823
- 1 8.439
175,298
95 !COMMUNICATION 3663
3,855
4.635
1-1 101
3,525
424
8.350
717
1 96 1 COMPRESSION & EN13695I
7171
763I
753 1
101
-1.2881
791 1
-278
97! CONCERTINDUSTRI 262 1
48,917
57.951
35.977
2 1.974 1
1.46 1
12.672
3.27 1
36
-460
-71
1
-53
8
708
636
936
98 j CONSOLIDATED EC( 2844
24,7801
1,974
-2,984
8.124 1
22.3 17
14.193
991 CONSOLIDATED EN7
2875
5.178 I
1,368,150
97,730
184,255 47.985
204,373
1.63 1.982 1,447,727
1 O0 i CONSUMERS PACW 3221
14,349
4,379
2,778
3,830
24,788
28.6 18
1 O 1 1 CORECO LNC
1 3674
44.923
378.139
-2,781
43.789 I -46,646
170,934
1 14.724
362.8 16I
1 03i COREL, CORPORATI~3575
z-

- - 7 - - - - - -

-.
T.

.-.

SI

- -- - -

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH
137

104 1 CO-STEEL MC

3312

106 CPI PLASTICS GROU 3089

3569
107 CRS ROBOTICS
108~ C S CREDIT
I
SYSTEM 3577
109( cv TECHNOLOGIES 2835
11O CYGNAL TECHNOL~3663
I I 1 DANBEL INDUS3645
1 12 DATAMARK S Y S 276
~1
1 13 DBA TELECOM
366 1
1 14 DECOMA INTERNAI 37 14

385,413
330,647
60.85 1
3,765
7.6 16
12,657
8,637
5,798
35.889
1.956
560,656

1,705,264
1,829,043
3 1.709
9.222
1,732
3.608
27,797
12,460
45.5 1 1
2,369
681.894

1,032,327
1333.12 1
2 1.973
5.636
674
1.094 I
14,5631
9.024 1
18,897
1.468
427.1 16

672.937
495.922
9.736
3,586
1.O60
3514
13,234
3.436
26,6 14
901
254.778

1 I'IIDIAGNOCURE INC 1 2835

l l S :DIGITAL PROCESSW 3651
1 191 DIMENSIONS WEST 1 3663.

13.124
28.133
1.970

13,304
39.6 15
7,806

5,159
10.309.
7,Ol 1 '

8.145(
19,306i
-7951

12 1 ID~VERSITYCORPOR~2899 !
132 i DOFASCO iNC
1 33 121

1.O96
1,603,890

1,143
3,377,100

1OS JCOTTcORPORATIOI 2086

CON

CON

124 1DOMCO TARKETT 3996 1

107.1 10
1251 DOMTAR MC
1261 1 1 1.568.323
126 DONOHUE MC
242 1 1 2.604.005
127 DOREL NDUSTRIES 25 121
682.294
1 28 DOVER INDUSTRIES 204 1
53,129
139 DUPONT CANADA Il/28 19/ 4,08 1.770
130 DYNAMIC DIGITAL 3663
110.354(
57,395
13I EAGLE PRECISION T/ 3542
132 EI ENWRONMEN
592
3585
TN
8.289
1 33 1 EIGER TE CH NO LOG^ 366 1 !
1 1.398
1341 ELECTROHOME LIM 3679 1
19.094
1 35 1 ENERCHEM INTERNA 29 1 1
420,532
1 36 ENERFLEX SYS3563
9.4 16
137 ENERSHARE TECHN~3823 1
I 3S EPICDATA
357 1
149.403
139 1 EUROPEAN T E C H N ~
3596
6.067
140i EVEREST MINES AN1 2092
302
138.406
14 1 !EXCO TE CH NO LOG^ 3544 !
9,422
I 42 1 EXI TECHNOLOGIES! 3669 1
1,059
I 43 EYELOGIC SYSTEM~
384 1 !
137.839
144: FANTOM TECHNOL~3589 (
3,692
1-15 (FAR WEST INDUSW 2329!
2,404
146i FLFn-PLUS.NET IN'f 3663 j
I47!FLETCHER CHALLE? 26 1 I ! 2.1 1 1 . 17
5,936
148!FLETCHER'S FINE FC/10l3
4.750
3533
149!FLOTEK
20.332 1
I 50 IFOREMOST INDUS^ 3533

INTERNA

INDUSTRIE^

--

673
470 1
1.548.600 1 1,828,500 1

-80.745

-1 1.1 98
2,726
1,952
-5 13
-849
-824
1.O48
3,049 1
1

-722 1

29.307 1

107j10
67,969
4,160
865
-3 19

6641

-79
. - 1-

717
1322
4.685
-2.461
73.649
-

-6.052 1
1 A87 )
1.637 1

42.464 1

-7 1

-5.460-3 82
-129

-3,167)
175.700 1

983 )
2.982300 1

-484
4 16, IO0

329,lO 1
ll3,908/ 215.193
4,030,000 2,342,000l 1,688,000
3,456.865 ) 2.068,543 ) 1288.322
460.790
137,2081 223,582
54,546
19.766 i
74.3 12,
62 1,9771 1209.668
t ,831,6451
9,4331
7821
8,65 1 )
43,41 1
80,6551
37344 1
3611
996 1
635
3.064 1
9.22 1 1
6.157
68.9 16
39,775 1
39.14 1
16,8671
9,638 1
7.229
100.056
159.506
59,450 1
7.870 1
1.181
9.05 1 1
14.0 1 1)
50290
5,085 1
587
4,498
1.570 1
163;
1.406
125,4051
47,763 1
77.642 1
i.61~1
3,292 1
4,907 [
883
i .O05 1
1,888)
38,1931
45,013'
83.206
4,4071
4.674
9,080
-17
52 (
35
355.000 i 3.175.700
2.530.700
64,204
83.161 1
147,365
1 -050(
2,608
3,658
20.732
3 1.2671
5 1.999

1.5 1 1,432
2.1 52766
46.50 1
13.5071
2 2 141
2.585 I
24.553 '
24.099
82.783 1
4.3361
656.1 75 i

14.725
74,000
329249
26.985
4270
298.772
-2.062
4,364
189
84
5,116
1.817
22.568
93 /
6.5 12
-1.096
-1 -559
1 1.1 15
99
51
10.318
147
18
298.600
3.656
-3.457
-4.597

3 50.798
2,348.000
2295.805
766,607
131.987
2.025.072
3681
136.618
3,766

42.104'
2.0 1 2
168.685

1.336-7561
33,477
641.73 1

34.1 70'
58.000
450,728
47.477
8.0 I 1
282,480
-1,750
7.508
236
360
22,734
-4.342
97,873
3.208
1 6.447
3 14.496
16.042
3.3301
161
53.1 74 (
8.627
928 1
-337
OI
-527
2 1.238
1 1 8.045 /
3,027 1
311
330
958 1
13.185
177.585 (
11.7891
4 12
151;
17
284.1001 -105,000
394.0151
11.377
4.5221
-1,762
44.948 i
-4.3 71

154'G.T.c. ~ S C O & 2759

155 GARNEAU INC
3499
1561 GEAC COMPUTER CC 3575

559,473 :
19,557

2.425.988

1,153,587'
29,250
702,572

728.64 1 '
6,044
424.8 15

424.946i
232061
277.757 1

117.33r
3,163
223,762

VALUINGPRWATE
COMPANIES: A DEA APPROACH
138

IPL MC
IPSCO iNC
IRWIN TOY LIMITEI:
IVACO iNC

M CHECKM~TCO
JANNOCK LIMITED

JEMTEC MC
KAFUS INDUSTRIES
KASTEN CHASE APP
KELMAN TECHNOL(
KING PRODUCTS IN4 366 1
KRYSTAL BOND INd 289 1
LABOPHARM INC / 2834
~
LAFARGE C O R P O 3241
LAMBUS ENTERPRI~3679
LANDOVER EN ERG^ 2869

7.739
1,398
44,269
4,330,770
1,729
2846

'

227 LBL SKYSYSTEMS CI 323 1

7.1 13
228 LE PRMT EXPRESS 2759
8,776
229 LEADER INDUSTRIE^ 385 1
13,5561
Z~O~LEADL'NG
BRANDS, d 10861
8.570 1
33 1 LEF MCLEAN B R O 2092
~
3,38 1
232 LEITCH T E C H N O L ~3679 1.O3 1,559
233 LG TECHNOLOGIES 3629
27,707
234 LiNAMAR CORPORA 3714 1.792.5 17
236 LOGICSYS INC
3823
237 LUXELL TECHNOLO 3575
3088
238 MAAX INC
239 MAGELLAN AEROSP 3728
240 MAGNA INI'ERNATI~
37 14
341 MAGNIFOAM T E C 2821
~
242 MAGNOTTA WINER~3084
243 MALETE QUEBEC 2672
244 MAPLE LEAF F O O D ~10 1 1
245 MARATHON FOODS I 3663

3,337
24,493
307,970
496,389
7374.103
20,940
13.3 10
56,024 1
1374236
623

347 MARSHALL-BARWIC 344 1

248 MAXILL INC
399 1

23,227
1,570

.310

-5.44 1
-5.108
-6.027.
36 1.68 1 1
473/
-1521

11,018
3,070
1-301
3,759,953 1
88
7

4.96 1
-2.556
-4039
6 14.461
-179
-47

6,39 1
3,376
3,996
7,041 1
8.1561
171.986 1
15.888
423.290

3041
-4, I O8 1
1,4171
-7.72 1
604
33,393
-9,463
84.3 851

40,142
4,742
22,106
35,667 1
3 1.232
147,743
45,6 16
998.3 12

95 1
-3.64a
2.822
- 1339
1,570
29,672
916
150.854

1210(
1,092
118
5,596 1
2.63 7
2,959
168.735 1
84,927
83,808
23 1.548
1 76,4 12 1
407.960
3,675,300 ( 4,945,400 1
8.620.700
18.599 (
5.532 1
24.13 1
1 7,764 1
8.766 1
26.530
96.0 1O 1
187.4 13
9 f -403
399.256
1,406.595 1-007.339
1.862,
434
3,296

-338 1
-2,548
1 2724 f
32,336 1
506200 1
4.289 1
1,432 1
3.80 I 1
-23 -257/

5,868
1,372
174.84 1
426,942
9,190,800
6,640 1
16.537 1
1 1 1.829 1
328 1,464
6,350

-2,096
20,084
4 1.9 18
803.498
-2,898
2-41 5
13,498
44280
-277

7,309
12.80 1
2,463
8,099
2.974
20,896 1
4.46 1.187 2387.763
243
634
6
3 16
24.0 18
4.799
14.625
37.899 1
17.037
193,542
38,659
669.64 1

17.627'
1.423
10,629
20,858 1
8.88 1
2 1 -556
2 - 7 71
216.3 5 1
- ---

-5,492
5,636
17,932
2,173,413
39 1

-464

4 1,445
1,797

9,428
923

2,093 1

32.0 17
874

-9

5 1,589
1,166

-232

-1.1 17

134

- ---

MCGRAW-HILL RYE~373 1 1
39.533 1
62.3481
11.1931
41.1551
5.2071
68.3 56 1
8.094
MDC CORPORATION 2759
242,559
740.0 t 3 1
54 1.869
198.144
14.1 O8 1
476.727 1
45,072
MEMOTEC COMMUb 366 1
4,953
30.28 1
15,913
14.368
-2,360 1
37,926
864
174.896
15,009
9.012j
153,0191
180.265
138.46 L
333384
MERCER MTERNAT 26 1 1
498
442
-754 1
597
1.3941
2,869
1.99 1
MERCH P E R F O3714
~
1.107.126
75,783
1
05,122
f
L
.
O
6
1
-9601
1.70
1,037
1,370,334
2.762.996
METHANEX CORPO~2869
143,412 -224.5 IO
307,087
720,969 I.087.6571 -366,688 1 408,920 I
MICROCELL ELECC 366 1
MILLTRONTCS LTD ,3823,
222.793 ,
105.188 1
15.320 1
89.868 1
10,006 /
125,053,
14,499
34,343
4,6 19
35.958
6,456
29,502 1
2.63 8
258 MITEC TELECOM IN(357 1
56,40 1
888.500
125,600
802300
435,500
9 1.900
1,237,700
359 MlTEL CORPORATIq 366 1
1 ,257,378
260 MOBiL COWORATI~
29 1 1 105,489,100 65.66 1,593 37,448,947 28.2 12,646 2.6 17.003 80.076.612 8,145,883
29,118
3,922
17.796
13.362
2,865
33,407
3 1,158
26 1 MOBILE CLMATE 3585
4,173,847
133,682
1,713,937
937,061
-841,413
2,650,998
262 MOORE CORPORA'TI{2761 1;428,580
1

'

- - -

CI

VALUING PRIVATE
COMPANIES:
A DEA APPROACH
140

270 NEWBRIDGE NETWC'

27 1 NEXFOR INC

3663
2426

8,125,481
929,923

1,966,825
2,775,000

733305 1 133.620
1,493,O1 1,282,000

- 183 18
48,000_

1,620,620
2,426,000

323,961
260,000

-44 1
1,247
1,439
-3 66
274 NORDEX EXPLOS~~E( 2892
530
738
2,177
-1.317
159
-3.879
-1,201
275 NORON TECH LTD 363 1
7,029
4.450
57 1
276 INORTEL NETWORKQ 3661 50,495,406 30$04,406 12,542.879 17-761.327 -824,725 36,991,685 1,397,578
17.704
333
14.059
85 1
11.001
Z%o6or
7,036
277 NORTH AMERICAN f 3829
-59
1
2.306
-23
1
6
2,698
2,467
7.100
378 NORTH AMERiCAN 1i 3841
54,256
4.58 1
2.1 15
7.884
21295
29,179
279 NORTHSIDE GROW 1 3714
14,056
-3.1 92
-3 96
2.195
-133
693
560
380 NORTHSiDE MINE^ 3533
4,066
82.76 1
7,388
1.3 87
25.841
3 0,959 1
69,109
43,268
28 1 NORWALL GROUP 262 1
1 7O.OOO
2,075,000
18.000
1,838,4471 3,580,000 2.O68,OOO 1,5 12.000
182 NOVA CHEMICALS q282 1
612891
16.033
67.7321
9.807
34,3501
283 NQL DRILLMG TOOI 3533
102,082
7 1,671 1
34,030 1
5266
284 NTS COMPUTER S Y 3571
~
48,09 1
38.093
18.55 1 1
19,5421
3,593
1
8.033
9
1,653
4.763
C
O
R
P
O
R
~
3879
~
4
1,449
6
1,530
36,407'
25.1
33
W-GRO
285
69.887
16.039
1 2,242
286.013
82,129
2.3241
801 1
1,701 1
997
3,007
683
287 OMNI-LITE WUSTIg 3297
23,659
-2,745
158 1
-3,003
3411
7.828
8,169
2.93 1
288 1ORALEE GROI.JP, IN1 3834
2.247
20,186
-3,286
4,544
5,234
9,778
40.576
289 (OTATCOlNC
1 3533
784
47,950
-8 82
4.767
15,142
1 0,375
3,878
290 1 PACIFIC E-LINK COq 3669
10,611
213
1.490,
-59
6,048
4,558
771
29 1 / PACIFIC RANGER Pd 3443
-3 80
9,3 12
-522
9,1175,772
3345
292 1PACIFIC SAFETY PR( 3 842
5,040
-4-4
12
96.9
-39.299
1
5
4,373
38,736
34.364
5,466
293 [PALLETPALLET INd 2448

VALUINGPRIVATE
COUPANIES: A DEA APPROACH
141

PRIMETECH E L E 3672
~
PRIMEX FOREST PR^ 242 1
PRoCESS CAPlTAL 4 291 1

PROFAB ENERGY ~43533

PROMATEK INDUS^ 3841
PROMETIC LIFE SCnf 2833
PRIJDEN?AK, STEEL 3498 [
QLT PHOTOTHERAP. 2834 1
QSOUND LABS, MC 3846
QUEBECOR PiUNl'IN 2759
353 1
R-P.M. TECH INC

143.365
56,925
5,667
2,980
10,3 74
78-451

205,602
974.676
92329
3.82 1.070
6366

56,143
99,053
3,455
3,929
2,876
36,374
154.315
103,223
7,372
5,900,722
24369

--

14201
33,341
1.8 IO
1,737
402
16,463
26,004
8,145
33 1.7328
3,497,957
15.507

4 1 -942
65,7 12
1,645
2.1 92
2,474
19.9 I 1
128.3 t 1
95.078
7,039
3,402,765
8,863

1 1,074

6,447
-877
129
936
-7.497
12,886
-24,071
-1,900
245,052
851

330 1 RE-CONBUILDING P 3272

1,974
35,055
17,683
7279/
899
5.247
4 1,474
47936
89.4 I O
38,558
33 1 j REKO [NTERNAOh 3544
540
150.054
3,292
332IRESEARCH M MOTI~3661
585308
153,347
-6.2 16
DLAMOM>
3915
47.33 1
79-183
22.55 1
56.632
333
13.479 f
334i RIDLEY MC
i 2048
106.6 1 6
124Z
160.050
230.83 8
-7.845 1
240.4 18
3 35 : RIVERSIDE FOREST 1 242 1
37 1772
130.854
75.896
15.213,
378 1
336 1 ROBERT MiTCHELL 3535
10,636
55.849
19-851
18.8 171
1 1 1,740
337'ROLLAND iKC
309.6 16
197,876
149395
2631
1.1441
2 1,009
35.3 73
338jROSE CORPORATIOP 3714
56.382
25.1 50
120.0241
736.578
8 15.074
339 1 ROYAL GROUP TECt 3089 2,920250
1.55 1.652
3.30 11
5 1,561 '
37.376
340 iROYAL LASER TEW 3699
16,2441
21,132
-20,7851
59,149
94.45 1
70.7 1 61
23.735
34 1 / SAC0 SMARTVISION 3861
481 1
-74
88
569
570 /
342 iSADDLEROCK RESO 2449
12.680
14,803
I 1.379
10,483 1
343 1 S A L E PHARMACEU 2834
2,196
524,402
205.3 82 1
3 19,020
23,393
~ J J ~ s A M U E L M A N U - T E C ~ ~310,901
~~
896.662
345 SAPUTO GROUP WC 2022 2,107.8 19
369,893 (
45.745
526,769
103.58 1 [
10.524
178.877
34G SCHNEIDER CORP0F 20 1 t
1 75296
139,628
9,714
70.829
89.37 1
3471SCI lNCOME TRUST 25 15
75.0 19
1 8.542
4.73 7
1.460
348 SCI-EC INSTRUMq 3826
11,174
6.43 7
5.73 1
946.000
2084 19.977.447 22.179.000 12,863,000. 9.3 16.000
349 1 SEAGRAM COMPM
10.026 /
3.335758 3,064,548 1
270.7 1 1 -659.573
350 i SEMI-TECH CORPOR 3639
-2,4691
178
17
263121
1.757
35 1 1 SENCO SENSORS INq 3823
986
61.221!
2,820
1,834
3 52 1 SENSE TECHNOLOGl3663
-67 1 1
1591
1,209
3.938 1
1.67 1
462
353,SENTEX SYSTEMS 2048
3311
354 SEPP'S GOURMET FC 20 1 1
13.804
34.064
27,532
20,260
-334
1.400 1
233 1
1.589
93 1
355[SEPROTECHSYSTEb 3569
-755
-758 1
37
795
572 !
356 i SERENA RESOURCE! 373 1
69.624 1
3 19.534 /
556.897
357 1 SHAW INDUSTRIES 2899
237,363
767.740
-2,954
-3,248 1
2,792
6.040
1 3674
358 iSHELLCASE LTD
15.698
109,171
9,507
57.662 1
359lSHERMAG IXC
12511
5 1.509
185.957
43.093 1
6.156,
102,844
59,75 1
3601SIC0 INC
7 1,059
1 285 1
21.089l
2,529
24.8 12
3.723
36 1 ] SERRA WIRELESS, I/3669
32,735
-5,782
3337
1,103
4.340
362iSIGNALGENE INC / 2836
17,484
12.245
1 7,023
33,825
3663
16.802
363 1 SIMMONDS CAPW
2,854 1
130.806
154,349
32.811
23,544
1 19.484
364 ( SINO-FORESTC o q 24 11
9.0461
-1,428
12396
3,250
2.977
365 1 SIRIT TECHNOLOGa 3829 1
8,155 /
6 1 -603
42,537
192,8291
104.140
366ISKYJACK iNC
3537
2 10.63 7
-7.507
4 147091
203,572
84,802
367 1 SLATER STEEL iNC 33 16
51,4151
6,755
105,655
54340
368 I SLEEMAN BREWERT 2082
126,836

IREX

1
1

89,O 13
168.079
1 50
6,766
3,144
10,586
175,656,
2.0041
3,244
5,848,564
36,090 1

1 1.2011
12.45a
571
19a
966
4,807
16,766
-22.854
-1,724
653,148
1.474

1 1,908!
-3.222
9.654
67,927 !
1.732
33,159;
1 1.3621
-2.538
398,004 1
27.007
14.783
450,942
2.788
78.712
3 1392
670,669
5,799.
1,944
196.8 19
1.050.1 O3 (
4.6 16
I9,36 1 i
3,333
70.1841
II
-3 8
2.3941
-11.239
735,4751
46.800
8 1 7355 (
63,972
905,908 1
1 8,180
104,8031
14,047
12,5951
-573
9.7 14.000
90.000
1.562.5381 -2 12,843
4391
-2.42 1
O1
-51 1
2.702 1
165
43.175 (
1.509
177
3.357
-289
O
95.05 1
769.0 17
-2.587
799
13-252
109.438
1 1.067
206.086
3,193
26.400
251
-4.30 1
40.79 1 1
-8.175
2 1-678
142.3 84
-1.1 12
1 1.466
10.407
199.433
10-927
509.6 10
11,283
76,023

VALUINGPRIVATE
COMPANIES: A DEA APPROACH
142

384 iSTELLA-JONES INC 249 1

19.54Or
385 1 STRATEGICTECHNC 3663
3-943
1-720
386 [STRATEGICVISTA 4 3669
~ ~ ~ ~ S T R E S SBIOTEC:
G E N 2836
44,933,
3S8!SUN ICE LIMITED
190
2329
359 [SUN-RYPEPRODUCI
2033
15,889(
~~O~SYMPL
COMMUN
EX
3669
5,879
39 1 jSYNSORB BlOTECH 2836
77.292
392 1 SYSCAN INTERNATI~3663
7,260
393 !TANBRIDGE
3111
9.639
394iTECHblIRE LTD
3 5422,698
39.5 /TECmrILABPHARM, 2834
107,756
1

CO-4

-.

398lTELEP.4NEL SYSTEb 3571

399lTEMBEC INC
26 1 1
JO0 !TEMPLARRESOURC 3732

5 1,2341
5.1561
1-981
33,477
1,368
36,678 (
2,452

24069
2369
2.473
6,678
3,764i
32.529

29,44 1
4,340
45.35 1
8.95 1
6637 1

1 5.43 8

9,450
10.082~
2,496.900
583.665
4,576
7i

1,999

2.376
1 9.938
5-531
2 1.146

27.465
1.887
493
36.799
-2.3961
14.149 1
453
14.003
1.964
25.4 131
3.419
45,125
4

-33571

13,806)
1,535,200
580

96 1.700
-573

'

3.82 1
382
- 1,778
- 14309
-1221
1.8141
- 13231
-13,487
159
2,880
82 1
4.422

-8,9586 1,100
-94

82.0 141
7.3 181
3.408
2.626
5.064
83.8 1 1
4.352
633
5.778
78.064 /
16.5 171
3.6241

9.577
1,422,900

5.371
1.171

- 1.373

- 13.394
-1.302
5.556
425
-13.016
425
4.7 1 1
1.555

9.388

'>

-5,789
207.000
-94

6,006
20,142

-8.949
2.242

40s i THOM SON CORPORA-27

11

408 (TIR SYSTEMS LTD &8

409!TML FOODS INC
1 20 13
4 I OiTOROMONT INm=5
4 I 1 ;TRADE WMD C O W 366 1
4 13i TRAILMOBILE CANA 37 15
4 13iTREE BREWiNG CON 2082
~I~!TIUNORTHCA
1 P~]~O~
4 1 5 !TRIPLE CROWN ELE~3663
4 1 6 i TRITECH PRECISION3369

420 (TSITELSYS C O R P O ~3663 /

12,193

8.9 12

8,662

249

14.889

3,156

I 1-733

-9,3 131
473 i

422 1 N RB0 GENSET M C 36211

423 ULTRAMAR DIAMOP 29 11
424 UNIBROE WC
2082
335 UNIFORET iNC
242 1
426 UNILENS VISION I N C 385 1
427 UNIROM TECHNOLQ 3825
428 ' WC
MTEWATI( 3625
429 UNITED DOMINION 3325
430[UPLAND GLOBAL C( 2834
43 1 URBAN XnCE AND S 2086
3494
432 VELAN MC

434 WCEROY HOMES

~d2452

4351 VINCOR INTERNATI~2084

80,615
3,323,827
33,828
136,498
199
441
5371
1258,176
11,221
12,878
793,613

29,840
f 92.1 14

~ ~ ~ ! V I T R ECAPITAL
OUS
329 1
4381VIVANT GROUP INC 2086
J39!WAJAX LIMITED
] 3569
440f WARWICK COMMUB 273 1
44 I i WE X-L. HOLDINGS C 3699
442 1 WESCAM MC
3663
443 j WESCAST INDUSTRI'3714
t14) WEST FRASER TIMB.26 1 1
445 IN'FI DVDUSTRES L T ~
3567
446 ( WI-LAN INC
3669

447!WRJPAK LTD
2656
3481WOOD COMPOSITE 3089
449 1 WORLD SPORTS ME: 3949
4501EROX CANADA iNC 3579
45 1 XILLLX TECHNOLOG 3 84 1
452 ZEMEX COR~ORATI~
3299

6,325
8,162,777
29,284
409,914
3,O 16
402
23 13
3,245,637
12,415
4,055
256,957
47,0821
203.096 1
t

914
876
120,082
240
17,083
76,678
574,300
880.03
9,697
329,339
3 18.500
9,334
4,932
2.56 1,066
22,149
86,904

156
6,1691
-1,9L9
6,03720 2,125,547 -1 19.946
2,113
20.065
9,219
-22244
128296
281,618
-4,022
-979
7,03 8
-278
680
-3,096
17
580
1,933
153.10 1
1,867,593 1,378.044
8,743
171
3,672
2327
-2,342
1.728
89,449
167,508
34,156
4.129 /
100.753 1

1,689i

254.7371
122,150'
2733
4,504 [
538 1
1.075 1
2,575,6 18 2-061.952 1
16.596
1,4001
228,628
102,849

VALUING PRIVATE
COUPANIES:
A DEA APPROACH

284.483

-1,743
374.735
4,047
-8.208
-256
-108
166
220,582
1.774
1,904
42.37a

13,753
7.0 1 1

12,953
102,343

833
10,825

44.425
204.8 10

1,316
18,956

2,520
-476
20 1.827
796
1,338
42.4 19
23 7.773
897.63 1
11.338
6.333
132.587
1.77 1
-536
5 13.666 1
15,1961
125,7791

230
-1,611
9,469
36
-1.647
2,865 1
47,856 (
5,6251
-1,0171
-6,292
16.138
-732
-1.838
69,826
-2,984
8,240

S,Ol 7j
2841
992,248 1

46a
-1,517
203 13
79
-1,624
8,494
67.277
161.198
-1 19
-5,ZM
37,018
-688
-1,814
158.732

5.385
2,865 1
956
1.431!
637,973 4363461
505 1
130ll
1.523 1
1851
873 121
44,893 1
42,594 1
280.3671
2,108,930 ( 1211.799 3
16,541 1
37.88 1 1
8.022

35(
17,100.5 191
21,8141
292,357)
4.905 1
0[
2,474
3,102,890

I.I90!

593
79,9041
284,861
1.863399
53.595
5.414/
336.3971
851
1231
1,550,4 t 2
5,602
159,560/

-2,5 1 1

7,128

APPENDIX B
DEA Mode1 Results and Market
Value Analysis of the Inefficient

Efficient and Inefficient Companies (SIC3600)

DMU
Score Reference set (lambda)
1 HAMMOND 0.7568 CAMCO 1 0.2217 TS TELE!
2 TURBO
1.0000 TURBO C 1.0000
3 lnverpower 0.5446 RESEARi 0.01 16 TS TEL(
0.4349 GLOBAL 0.021 7 AASTRA
4 UNITEC
5 GLOBAL
1.O000 GLOBAL 1.0000
6 LG TECH
0.5866 CAMCO 1 0.0060 TS TELE!
7 CAMCO
1.0000 CAMCO I 1.0000
8 NORON
0.0653 FIFW-PL 0.9899 SNlF SE<
9 AVANTE
0.5830 RESEARl0.0028 TS TELEI
1O SEMI-TECH 0.2145 CAMCO 1 0.7746 CELESTI
0.7553 GLOBAL 0.0556 AASTRA
11 DANBEL
12 A W A
0.8473 TS TELEI 0.0245 AFP/-PL
13 TIR SYS
0-6557 TS TELE 0.1 198 FIFTY-PL
14 DIGITAL
0-8843 RESEARl0.0034 TS TELE(
15 SONY
1-0000 SONY CC 1.0000
16 CINRAM
1.0000 CINRAM 1.O000
1-0000 AASTRA 1.0000
17 AASTRA
0.3991 TS TELEI 0.5432 F I W - P L
18 ABL
19 BCS TECH 0.6766 FIFTY-PL 0.91 56 PAClFlC
0.2541 CELESTI 0.1402 LEITCH 7
20 BELL
21 CIRCA
0-8595 AASTRA 0.2780 TS TELEf
22 DBA TEL
0.6399 TS TELEI 0.1446 F I W - P L
23 EIGER
0.8151 CAMCO 10.0107 TS TELEI
24 KING
0.4719 PACIFIC 0.2143 SNlF SE(
25 MEMOTEC 0.6492 TS TEL( 0.8199 PRIMETE
26 MICROCELL 0.0664 CAMCO 1 0.1845 PAClFlC
27 MITEL
1-0000 MITEL CC 1.0000
0-6709 RESEAR 0.01 32 FIFW-PL
28 NHC
29 NORTEL
1.O000 NORTEL 1.0000
30 R1M
1.0000 RESEARl 1-0000
31 TRADE WIN1 0.6762 FIFlY-PL 0.4601 PAClFlC
32 TS TEL
1.O000 TS TELEI 1.0000
33 ADVANTEDC 0.1 804 RESEARi 0.01 17 TS TELEI
34 CALIAN
0.8001 CAMCO 1 0.0599 TS TELEI
35 CDN marconi 1-0000 CANADlf 1.O000
36 CELL-LOC 0.9803 TURBO C 0.2345 RESEARl
37 COM DR/ 0.8633 CAMCO 1 0.0824 PRIMETE
38 COMM
1-0000 COMMU?1-0000
0.5892 RESEARl0.0242 TS TELEI
39 CYGNAL
40 DIMENSION. 0.1 1O3 TURBO C 0.1233 RESEARl
41 DYNAMIC 0.9394 TURBO C 0.9828 RESEARl
42 FIFTY-PLUS 1.O000 FIFlY-PL 1-0000

No.

0,9208 FIFTY-PL
0-0472 FIFTY-PL

0.0676
0.9285 PRIMETEC

0-01O1
0-0210 FlF7Y-PL
0.2254
0,621 0 FIFTY-PL
0,9525 PACIFIC
0-8776 PAClFlC
0,8591 LEITCH 1

0-3099 INTERNAT
0.0230
0.0027
O. 1375

0-4001 PAClFlC
0-0417 SNlF SEC
0-4090 BALLARI:
0.0737 COMMUC
0,8450 PAClFlC
0,1798 FI-PL
0-7857
0-1655 LEITCH 1
0-8155

0.0567
0.0428
0.4507
0.6009 PRIMETEC
0.0104
0.5594 PAClFlC E-

0.0279 LEITCH TE

0-3397 SNlF SEC

0-0891 FIFTY-PL
0.5935 PRIMETE
0.0056 FIFTY-PL
0.1 168 LEITCH 1
0.6881 COMMUC
0.0003 FIFTY-PL
0.01 72

0.2612 LEITCH TE
0.8763

APPENDM B Efficient and Inefficient Companies (SIC3600)

GEMSTAR 0.2324 TURBO C 0.0118 FIFTY-PL
INTER DATA 1.0000 INTERNP 1.O000
Kasten chase 0.9949 RESARi 0.1 130 TS TELE1
MARATHON 0.6045 FIFlY-P100.8016PAClflC
NEWBRIDGE 1-0000 NEWBRlf 1-0000
SENSE TEC) 0-3741 TURBO C 0.1621 FIFTY-PL
SIMMONDS 0.7663 PACtFlC 0.8477 SNIF SE<
SPECTRUM 0-7867 TS TELb 0.8465 PRIMETE
SR TEL
0-4470 CAMCO 1 0.1363 PRIMETE
52 STRATEGIC 0.9832 GLOBAi 0.3983 CAMCO I
53 SYSCAN
0.6574 AASTRA 0.0522 COMMUt
54 TEKLOGIX 0.8099 CAMCO I 0.0755 PRIMETE
55 Triple crown 0.5260 TS TELE1 0.2507 F I W - P L
56 TSI
0.21 12 FIFTY-PL 0.7477 PAClFlC
57 WESCAM
0.6285 TS TELEI 0.1 979 PRIMETE
58 EX1
0.7569 RESEARl0.0117 COMMUt
59 JEMTEC
1.O000 JEMTEC 1.0000
60 PAClFlC
1.0000 PAClFlC 1.0000
61 SIERRA
0.9963 RESEARl0.0052 TS TELE1
62 SNIF
1,0000 SNlF SEC 1.0000
63 STRATEGIC 0,5342 FIFTY-P10.6999 PAClFlC
64 SYMPLEX 0.5622 TS TELEI 0.0143 FIFTY-PL
65 WI-LAN
0.8732 RESARi 0.0359 TS TEL(
66 CIRCUIT
0.7525 GLOBAL 0.4312 AASTRA
67 PRIMETECH 1,0000 PRfMETE 1.0000
68 CELESTICA 1.0000 CELESTI 1.0000
69 CORECO
0-9834 FIFiY-Ft 0.8480 GENNUb
70 GENNUM
1.0000 GENNUk 1.0000
71 SHELLCASE 0.0847 FIFlY-PL 0.8555 PACIFIC
72 TUNDRA
0.9817 TS TELEt 0.t004 COMMUt
73 C-MAC
0.8171 CAMCO 1 0.2203 MITEL Cc
74 Efectrohome 0.6733 CAMCO 1 0.0752 TS TELE1
75 LAMBUS
0.7096 TURBO C 0.0660 FIFTY-PL
76 LEITCH
1.0000 LEITCH 11-0000
77 PLAiNTREE 0.8663 RESEARl0.0884 TS TELE1
78 TRI-VISION 0.7426 RESEM~
0.0850 TS TELE1
79 BALLARD
1,0000 BALLARC 1.0000
80 BLUESTAR 0.5548 CAMCO 1 0.0926 PRIMETE
81 COMPRES 0.31 05 FIFTY-PL 0.9716 PAClFlC
82 GSI Lumonic 0.9167 CAMCO I 0.0196 CELESTI
83 NEXTRON 0.5345 RESEARl0.0032 TS TELE1
84 ROYAL LASE 0.6837 FIFW-PL 0.8098 GENNUhr
0.9437 RESEARl0.0088 TS TELEf
85 WE X.L.
43
44
45
46
47
48
49
50
51

0.761 7 FIFIY-PL
0.1285 SNlF SEC
0.8379
O. 1523
0.0760 LEITCH 1
0.41 92 LEITCH 1
0,0012 FIFIY-PL
0.4233 FIFlY-PL
0.8796 LEITCH 1
0.71 53 PAClFlC
0.1203 SNlF SEC
0.7437 LEITCH 1
0.2240 FIFiY-PL

0.0775
0.4445
0.4871 JEMTEC Iii
0.5220 PRlMEtEC
0.0449
0.0339
0.1 320
0.0584
0.7599 LElTCH TE

0.2631 COMMUt

0.6358 LEITCH TE

0.0642 SNlF SE<

0.9054 PACJFIC
O. 1599 FIFIY-PL
0.4006 FIFlY-PL

0-2359
0.0802

0.8042
0.1531 INTERNAT

0.01 19 LEITCH 1
0.01 13 SNlF SE<
0.7976 PRIMETE
0.1 599 CELESTi
0.4257 PRIMETE
0.9340

0.1 332
0.0710 LEITCH TE
0.0588 LEITCH TE
0.4990

0.4958 FIFTY-PL
0.141 5 COMMUF
0.8695
0.0131
0.01 35
0.0452
0.1021
0.0060

LEITCH 1
SNlF SEC
LEITCH 1
FIFTY-PL
LEITCH 1
FIFTY-PL

VALUCNG
PRIVATE
COMPANIES:
A DEA APPROACH

147

ABL CANADA (SIC3600)

DMU
Score Market Cap Distanc Efficient Peen
#
$(OOOs)
32
42
18 ABL CAh 0.399
13.382
o.wm 0.5432 0.4001
13 TIR S Y S 0.656
1,710
o.4101 0.1198 0.8776
12 A W A T E 0.847
2,429
0.5753 0.0245 0.9525
22 DBATEL 0.64
1 . 9 ~ ~
0.3590 0.1446 0.8450
55 TRIPLE (: 0.526
4.979
0.1854 0.2507 0.7153
64 SYMPLE 0.562
5.879
0.5356 0.0143 0.9054
32 TS TELE
1
11,832
0.3720 1-0000 0-0000
42 FIFTY-PL
1
2,404
0.6581 O.OOOO
I.OOOO
60 PAClFlC
1
3,878
1.3450 0.0000 0.0000
Ranking, according to the distances that am less than 1 (for ABL CANADA):

TRIPLE (
DBATEL
TS TELE
TIR S Y S
SYMPLE
A W A TE
FIFlY-Pl

0.526
0.64
1
0.656
0.562
0.847
1

4.979
1.956
11.832
O
5,879
2,429
2.404

0.1854
0.3590
0.3720
0.4101
0.5356
0.5753
0.6581

Similarity of the efficiency s c o m (difference between ABC CANADA'S score ):

TRIPLE (
SYMPLE
DBATEL
TIR S Y S
A W A TE
TS TELE
FIFTY-Pl

0.526
0.562
0.64
0.656
0.847
1
1

4,979

0.1854

5,879
1.956
1.710
2,429
11,832
2.404

0.5356
0.3590
0.4101
0.5753
0.3720
0.6581

Similarity of the Market Cap.

TIR S Y S 0.656
1,710
DBATEL 0.64
1.956
FIFTY-PI
1
2.4W
2.429
A W A T E 0.847
4,979
TRIPLE ( 0.526
SYMPLE 0.562
5,879
TS TELE
1
11.832

(diffemnce between ABL CANADA's market cap.):

0.4101
0.3590
0.6581
0.5753
0.1854
0.5356
0.3720

Range for ABL CANADA's market cap. based on peer group is $1, T l 0,000 and $11,832.000
ABL is a little bit overvalued

Market Capitalization of ABL CANADA based on EFFICIENT PEERS

$7,609 thousand
- -

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

148

DMU

Score

33 ADVANTEDC
9 AVANTE TEC
3 INVERPOWE
30 RESEARCH I
32 TS TELECOb
42 FIFTY-PLUS.
45 KASTEN CHI
65 WI-LAN INC
77 PLAINTREE :
83 NEXTRON C
85 WE X.L. HOL

Market C. Diitance Efficient Pwrs

$(000s) (displayec
30
32

Distance
42 Formula

0.1 80379
0.582956
0.544648
1
1
1
0-994883
0.87321
0-866344
0-534451
0.94369

Ranking, according to the distances that are less than 1 (for ADVANTEDGE TECHNOLOGIES I
Nextron
0.00474
Avante
0.01 06s
WE X-L. H0Id
0.01432
Wt-Lan
0.01463
Fifty-plus
0-01823
Plaintree
0.40491
Similarity of the efkiency scores (difference between ADVANTEDGE score (0.18):
Nextron
-0.354072
Avante
-0.402577
WE X.L. Hotd -0.76331 1
Wi-Lan
-0.692832
Fifty-plus
4-849621
Plaintree
4-685966
ADVANTEDGE's efficiency score is not close to any of the others, the closest is Nextron at (
Similarity of the Market Cap. (diffemnce between ADVANTEDGE's market cap. ($19,92
Nextron
17,502 (19,928-2
2.426
Avante
17,625
2,302
WE X.L. HoId
2,844
17.083
Wi-Lan
-309.412
329.339
Fifty-plus
17,524
2.404
Plaintree
4,981
14,946
Range for ADVANTEDGE's market cap. based on peer group is $2,302,000 $329,339,000
ADVANTEDGE's real market cap of $19,928,000falls between this range.

Market Capitaliution of ADVANTEDGE based on EFFICIENT PEERS

[(O-012'585,308) + (0.089'1 1,832) + (0.902,404)
is $10,064 thousand

VALUING PRIVATE COMPANIES:

A DEA APPROACH

149

AVANTE TECHNOLOGIES INC (SIC3600)

DMU

Score

9 AVANTE T
3 INVERPO\
30 RESEARC
32 TS TELEC
33 ADVANTE
42 FIW-PLL
45 KASTEN C
65 WI-LAN IN
77 PLAINTRE
83 NEXTRON
85 WE X.L. HI

Market C. Distance Efficient Peem

S(000s) (displayec
30
32

42

0,5829556
OS446483
1
1
0-1803789
i
0.994883
0.8732104
0-8663444
0.5344511
0-9436897

Ranking, according to the distances that an, less than 1 (for AVANTE TECHNOLOGIES INC):
WE X.L_ HI
0.00034
Fifty-plus
0.001
Nextron
0.0012
Advantedg
0.01O 6
Wi-Lan
0.05
Plaintree
0.5468
Similarity of the efficiency scores (diffemnce between AVANTE's score (0.583):
WE X.L. H
t -0.360734
Fifty-plus -0.41 7044
Nextron
0.0485045
Advantedg 0-4025768
Wi-Lan
-0.290255
Plaintree
-0.283389
AVANTE's efficiency score is very close to Nextron's than Plaintree
Similarity of the Market Cap. (difference between AVANTES market cap. ($2,302,000):
WE X.L HI
-14,781 (2,302-17,083)
Fi@-plus
-101 (2,302-2,404)
Nextron
-124
Advantedg
-17,625
Wi-Lan
-327,037
Plaintree
-12,644
Range for AVANTE's market cap. based on peer group is between $2,404,000 and $329,339.000
AVANTEs real market cap of $2,302,000 falls below this range.

Market Capitalization of AVANTE based on EFFICIENT PEERS

[(O-0028*585,308)+ (0.021 1
' 1,832) + (0.989,404)
is $4, 21 thousand

VALUINGPRIVATE
COMPAN~ES:
A DEA APPROACH

150

A W A (SIC 3600)
DMU
#
12 A W A TECH
32 TS TELECO
18 ABL CANAD
55 TRIPLE CRC
22 DBA TELEC
13 TIR SYSTEh
64 SYMPLEX C
42 FlFTY-PLUE
60 PAClFlC E-L

Score

Maricet Ca1Distance Efficient Peefs:

$(OOOs)

32

42

60 Distance

0.84726
1
0.39914
0.52596
0-63987
0.65569
0.56217
1
1

Ranking, according to the distances that are Iess than 1 (for AWA):
A W A TECH 0.84726
2,429
o.oo00
FIFW-PLUS
1
2.404
0.0034
SYMPLfEX C 0.56217
5,879
0.0056
1.71O
0.01 51
TIR SYSTEh 0.65569
DBA TELEC 0.63987
1,956
0.0261
TRIPLE CRC 0.52596
4,979
o. 1076
ABL CANAD 0.39914
13,382
0.5753
Similarity of the efficiency scores (dhmnee between AWA's score ):
ABL CANAD 0.39914
13.382
0.5753
TRIPLE CRC 0.52596
4,979
o. 1076
SYMPLEX C 0.56217
5,879
0.0056
DBA TELEC 0-63987
1.956
0.0261
TIR SYSTEh 0.65569
1,ilO
0.0151
AWATECH 0.84726
2,429
o.oo00
FIFTY-PLUS
1
2.404
0.0034
Similarity of the Market Cap. (difhrenee between AWA's market cap.):
1.ilO
0.01 51
TIR SYSTEB 0.65569
DBA TELEC 0.63987
1.956
0.0261
1
2,404
0.0034
FIFTY-PLUS
o.oo00
A W A TECH 0.84726
2.429
TRIPLE CRC 0.52596
4,979
0.1 076
SYMPLEX C 0.56217
5,879
0.0056
ABL CANA 0.39914
13.382
0.5753
Range for AWA's market cap. based on pm group i.$1,710,000 and $13,382,000
A W A ' s mal market cap of $2,429,000 b right in this ange

Market Capitalization of A W A based on EFFICIENT PEERS

$2,669 thousand

DMU
Score MsrlcetCap
Efficient Peers (lambdas)
#
$(OOOs) b
42
60
62 distance
19 BCS TECk 0.676
1 . ~ 7 o.oo00 0-9156 0-0417 0.0428 0.0000
0.3210 0.4601 0-3397 0.2002 0.3210
9,700
31 TRADE WI 0.6762
0.0213 0.8016 0-1285 0.0699 0-0213
46 MARATHC 0.6045
623
56 TSI TELSI 0.21 12
12,193
0.0423 0.7477 0-1203 0.1320 0-0423
63 STRATEG 0.5342
1.720
0.0843 0.6999 0.0642 0.2359 0.0843
0.0127 0.8555 0.01 13 0.1332 0-0127
71 SHELLCA: 0.0847
is.696
81 COMPREE 0.3105
717
0.0047 0.9716 0.0131 0-0153 0-0047
42 FI+-PL1
1,0000
2,404
0.0107 1.0000 0.0000 0.0000 0-0107
60 PAClFlC E 1.0000
3,878
1.7584 0.0000 1.0000 0.0000 1.7584
62 SNlF SEC1 1-0000
307
1.7562 0.0000 0.0000 1-0000 1-7562
Ranking, according to the distances that are les8 than 1 (for BCS TECH):
BCS TECk 0.6766
1,027
0.0000
COMPREL 0-3105
717
0.0047
FIFTY-PLL 1-0000
2,404
0.0107
SHELLCA: 0.0847
is,sss
0.0127
MARATHC 0.6045
23
0.0213
TSI TELSI 0.21 12
12,193
0.0423
STRATEG 0.5342
1,720
0.0843
TRADE WI 0.6762
9,700
0.3210
Similarity of the eniciency scoms (diffemnce between BCS TECH'S score ):
SHELLCA: 0-0847
i5,696
0.0127
TSI TELS) 0.21 12
12,193
0.0423
COMPREE 0.31 OS
717
0.0047
STRATEG 0.5342
1,720
0.0843
MARATHC 0.6045
623
0.0213
TRADE WI 0.6762
9 . 7 ~ 0.3210
BCS TECk 0.6766
1,027
0.00
FIFP/-PLL 1.O000
2,404
0.0107
Similarity of the Market Cap. (difference between BCS TECH'S market cap.):
MAi3ATf-K 0.6045
623
0.0213
COMPRE 0.3105
717
0.0047
BCS TECk 0.6766
1.027
0.0000
STfWTEG 0.5342
1,720
0.0843
FIW - P U . 1.O000
2.404
0.0107
9,700
0,3210
TRADE Wi 0.6762
12.193
0.0423
TSI TELS) 0.21 12
I
5,698
0.01
27
SHELLCA: 0.0847
Range for BCS TECH'S market cap. b r s d on peer g i w p is $623.000 to $15,898,000

Market Capitalizaon of BCS TECH based on EFFICIENT PEERS

$2,376 thoussnd

VALUING PRIVATE COMPANIES: A DEA APPROACH

152

VALUATION ANALYSIS BELL CANADA INTERN (SIC3600)

DMU

Score

20
68
76
79

Market Cal Efficient Peem

ib(000s)
68
76

BELL CANADA 11 0.25406

CELESTICA INC
1
1
LEITCH TECHNC
1
BALLARD POWE

79 distance

1,371,120 0.14023 0.40904 0.45073

1 3.97E-16
O
O
1
O
O
O
1

2,854,867
1,031.5~9
s,soa,271

0.0000
1.1097
0.5721
0.4887

Bell is similar to Leitch and Ballard

Therefom, Market Value nnge of $l,O3l,SSS,OOO to $3,W8,27t,000
Upper bound
$2,448,639,000

BLUESTAR BAlTERY SYSTEMS INTER. (SIC3600)

DMU

Score

BLUESTAR BAT
SR TELECOM IN
COM DEV INTEF
TEKLOGIX INTE
7 CAMCO INC
67 PRIMETECH EU
76 LEITCH TECHNC
80
51
37
54

Market Cal Efficient Peem:

S(00Os)
7
67

0.55479
0.44695
0.86326
0.80991
1
1
1

Distance
BLUESTAR BAT 0.55479
TEKLOGIX INTE 0.80991
PRIMETECH EL1
1
SR TELECOM IN 0.44695

76 distance

9.26502 0.86946 3.79E-02 0.00000

112.897 0.13625 0.41924 0.44451 0.36994
252.315 8.24E-02 0.1 168 0.80078 1.14858
243.501 7.55E-02 0.87962 4.49E-02 0.00044
1
O
O 1.58077
1w.m
O
1
O 0.02705
143,365
O
O
1 1.69017
1,03i,s!i9
44.404

O
0.00044
143.365 0.02705
i12,897 0.36994
44,404

243.501

Market value range: $112,897,O to $243,Sl ,Mi0

It appears to be undervalued.
Upper bound
$173,006 thousand

VALUINGPRIVATECOMPANIES:
A DEA APPROACH

i 53

VALUATION ANALYSIS FOR CALlAN TECHNOLOGY LTD (34)

DMU
#

Score

34 CALIAN TECH1 0.80014

6 LG TECHNOLC 0.58659
1 HAMMOND W 0.75677
7 CAMCO INC
1
32 TS TELECOM I
1
67 PRIMETECH E
1
74 ELECTROHON 0.67333

Market Cal Distanu Efkient Peem

Distance
67 (formula)

0-00000
0-02846
18,426 0.04795
100,000 1.35618
11,832 0.28897
143,365 0.78275
11.398 0.05162
33.191
27,707

Accordingly, Calian is clorest to (in distance):

CALIAN TECHI 0.00000
LG TECHNOLC 0.02846
HAMMOND W 0.04795
ELECTROHON 0.05162
TS TELECOM l 0.28897
PRIMETECH E 0.78275
Similarity of eficiency scores (Distance of Calian. to):
LG tech
0.21356
Hamrnond
0.04337
Eiectrohorne
O.lZ68l
TS telecom
-0.19986
Primetech
4.19986
Similarity in Market Caps (Calian- M.C. minus):
LG tech
5,484
Hamrnond
14,765
Electrohorne
21,793
21,359
TS telecom
Pnmetech
-1 10,174
Range for Calian's market cap based on peer groups is $11,398,O to $27,70?,WO
By adding Pnmetech we can increase ttih ange to a m u of $143,365,000
But, Primetech is the moet diferent from Calian in terms of efficiency score, distance, and m.v.

Calian's real market cap. of $33,191,000 falls just out of this range

UPPER BOUND CALCULATION:

Based on lambda relation of CALlAN to efficient p e r s
Looks like:[(0.0599'100,000) + (0.59'1 1,832)+(0.35'143,365)]
$62,705 thousand

CELL-LOC INC 8 ClRCA (SIC 3600)

DMU
#
36
40
2
30
42

Score

Market Cal Distance

S(0OOs)

Efficient Peers
2
30

42 Distance

CELL-LOC IN1 0,9803

DIMENSIONS 0.1 1029
TURBO GENd
1
RESEARCH It
1
FIFTY-PLUS.1
1
Distance
CELL-LOC IN1 0,9803
DIMENSIONS 0.11029
FIFIY-PLUS.1
1
Efficiency:
DIMENSIONS 0.1 1029
CELL-LOC IN( 0.9803
FIFTY-PLUS.1
1
Market Value
DIMENSIONS 0.1 1029
FIFTY-PLUS.!
1
CELL-LOC IN1 0.9803

Range for market value $1,970,000 to $2,404,000

Upper Range
$23,996 thousand

CIRCA
DMU
#

Score

21 CIRCA ENTEF 0.85947

17 AASTRA TEC
1
32 TS TELECOM
1
38 COMMUNfCA'
1
67 PRIMETECH I
1

Market Cap
$(OOOs)
6,601
46.091
1 1,832

3.855
143,365

Sirnilar to Aaara and Communication, therefore

range for market value of $3,855,000 to $46,091,000

Upper bound
$22,797 thousand
$22,797,000

VALUING
PRIVATE COMPANIES: A DEA APPROACH

155

CIRCUIT WORLD CORPORATION 8 C MAC (SIC3600)

DMU
#

Scom

Market Ca( Efficient Peers:

$(oOos)
5
17

42

0.4006
0.6210
0,0000
1.0000
0-0000
0.0000

0.1531
0.3099
0.0000
0.0000
1.0000
0.0000

66 CIRCUIT W( 0-75253
11 DANBEL IN1 0.75534
5 GLOBALTH
1
17 AASTRA TE
1
42 FlW-PLUS
1
44 INTERNATIC
1

0.4312
5.798 0.0556
13,222
1.0000
4s.0~1 0,0000
2,404
0.0000
3.893
0.0000
289

44 distance

0.0151
0.0135
0.0000
0.0000
0.0000
1.0000

0.0000
0.2142
0.5077
03689
1.0638
1-3399

In terms of distance, Circuit s clasest to Danbel, then Global, then Aastra

In terms of efficiency score it is closest to Danbel
Range for market cap. $5,798,000 to $46,091,000

Market Capitalization of Circuit Wodd based on EFFICIENT PEERS

$24,592 thousand
Thus it seems that Circuit World is undervalued

VALUATION ANALYSIS FOR C-MAC INDUSTRIES INC (SIC3600)

DMU

Score

73
7
27
68
76

Market Cap
$(OOOS)

C-MAC lNDl 0.81714

1
CAMCO INC
1
MITEL CORI
1
CELESTKA
1
LEITCH TEC

291.127

f00.000
1,257,378
2.854.867
l,osl.sss

27

68

0.2203
1.0000
0.0000
0.0000
0.0000

0.1 599
0.0000
1.0000
0.0000
0.0000

0.0588
0.0000
0.0000
1.0000
0.0000

76 distance

0.5610
0.0000
0.0000
0.0000
1.0000

0.0000
0.9518
1.0725
1.2747
0.2702

MV range $100 million to $1 billion, based on peem Leitch and Camco

Upper bound:
$969,673 thousand

VALUING PRNATE COMPANIES:

A DEA APPROACH

1 56

COM DEV INTERNATIONAL LTD 8 DANBEL (SIC3600)

Score

DMU
#
37
51
80
54
7
67
76

COM DEV lt
SR TELECO
BLUESTAR
TEKLOGIX 1
CAMCO INC
PRIMETECk
LEITCH TEC

Market Ca1 Effithrit Peers

$(OOOs)
7
67

0-86326
0,44695
0.55479
0.80991
1
1
1

0.0824
112,897 0.1363
44.404 0.0926
243.~01 0.0755
ioo,o00
1.0000
143,365 0.0000
1,031,559 0.0000
252.31 5

76 distance

O. 1168
0.4192
0.8695
0.8796
0.0000
1.O000
0.0000

Two peers: Leitch and SR Tel

Market value ange: $9 l2,8W,WO to $l,O3l,SS9,
Upper bound estimate:
$851.037 thousand

DANBEL INDUSTRIES CORPORATION (SIC3600)

DMU

Score

11 DANBEL IN[ 0.75534

5 GLOBAL TH
1
66 CIRCUIT W( 0.75253
1
17 AASTRA TE
42 FIFTY-PLUS
1
44 INTERNATK
1

Market Ca( Efficient Peen,

$(Ooos)
5
17
5.798
13.222
289
46.091

2.404
3.893

0.0556
1.0000
0.4312
0.0000
0.0000
0.0000

0.6210
0.0000
0-4006
1.0000
0.0000
0.0000

42

0-3099
0.0000
0.1531
0.0000
1.0000
0.0000

44 distance

0-0135
0.000
0.0151
0.0000
0.0000
1.0000

0.0000
1-3737
0.2142
0.2429
0.8652
1.4579

In terms of distance, Danbel b neariest Circuit, then Aastra, then Fifty

In terms of efkiency score it is closest to Circuit
Range for market cap. $289,000 to $46,091,000

Market Capitalization of DANBEL based on EFFICIENT PEERS

$30,156 thousand

VALUINGPRIVATE COMPANIES: A DEA APPROACH

157

COMPRESSION & ENCRYPTlON TECH (SIC3600)

DMU
#

Score

81 COMPRESE 0.31046
71 SHELLCASE8.47502
63 STRATEGIC 0.53425
56 TSI TELSYS 0.21119
46 MARATHOh 0.6045
31 TRADE WIN 0.67621
19 BCS TECHh 0.67662
42 FlFlY-PLUS
1
60 PACIFIC E-L
1
62 SNIF SECUl
1

Market Cal Distance Efficient Peers

$(OOOs)
o.oaoo

717
15,698
1,720
12,193

0.0274
0,1251
0.0752

623

0.0452

9,700
1,027

0.4024
0-0047

2,404
3.878
307

0.0012
1.9182
1-9138

42
0.9716
0-8555
0.6999
0.7477
0.8016
0,4601
0.9156
1.0000
0.0000
0.0000

60
62 diaance
0.0131 0.0153 0.0000
0-0113 0.1332 0.0274
0.0642 0.2359 0.1251
0.1203 0.1320 0.0752
0.1285 0.0699 0.0452
0.3397 0.2002 0.4024
0.0417 0.0428 0.0047
0.0000 0.0000 0.0012
1 .O000 . 0.0000 1.9182
0,0000 1 -0000 1 -9138

Ranking, according to the distances that are Iess than 1 (for COMPRESSION):
COMPRESE 0.31046
717
o.ooo0
FIFW-PLUE
1
2,404
0.0012
BCS TECHh 0.67662
1.027
0.0047
SHELLCASE8.47E-02
15,698
0.274
MARATHOh 0.6045
623
0.0452
TSI TELSYS 0.21 119
12.1 93
0.0752
STRATEGIC 0.53425
1,720
0.1251
TRADE M N 0.67621
9.700
0.4024
Similarity of th8 efliciency scores (dmereiice between COMPRESSION'S score ):
SHELLCASE8.47E-02
15,698
0.0274
TSl TELSYS 0.211 19
12,193
0-0752
COMPRESE 0.31046
717
o.oo00
STRATEGIC 0.53425
1,720
0.1 251
MARATHOh 0.6045
623
0,0452
TRADE WlN 0-67621
9,700
0.4024
BCS TECHh 0.6762
1,027
0.0~7
FIFlY-PLUE
1
2.404
O.OOI2
Similarity of the Market Cap. (difference between COMPRESSION'S market cap.):
MARATHOh 0.6045
623
0.6452
COMPRESI 0.31046
717
o.ooo0
BCS TECHh 0,67662
1,027
0.0047
STRATEGIC 0.53425
1,720
0.1251
FIFTY-PLUE
1
2,404
0.0012
TRADE WIN 0.67621
9,700
0.4024
TSl TELSYS 0.211 19
12,193
0.0752
SHELLCASE8.47E-02
15,698
0.0274
Range for COMPRESSION'Smarket cap. based on peer group is $623,000 to $15,698,000

Market Capitalization of COMPRESSION based on EFFICIENT PEERS

$2,391 thousand

VALUATION ANALYSE FOR CORECO INC (SIC3600)

DMU
#

Score

Market Cal Efncknt Peers

42
70
$(OOOs)
0.01 19
CORECO IN 0.98338
44,923 0.8480
ROYAL LAS 0.68375
s i , m 0.8098 0.1021
FIFTY-PLU3
1
2.404
1.0000 0.0000
1
5 7 6 . ~ 8 0.0000
1-0000
GENNUM C
l
0.0000
76 LEITCH TEC
1
1,031,559 0.0000
Coreco's p e r s are Royal Laser and F'My plus
MV range $2.4 to $51 million
Upper bound estimate ( h e d on efficient pers):
$153,477 thousand

69
84
42
70

76
0.1402
0.0881
0.0000
0.0000
1.O000

distance
0.0000
0.0123
0.0429
1.7 150
1.4585

ROYAL LASER TECH CORP (SIC3600)

DMU
Score
Market Cal Efficient Peers
#
S(00Os)
42
70
76 distance
84 ROYAL LAS 0.68375
5i.m
0-8098 0.1021 0.0881 0.0000
0.01 19 0.1402 0.0123
69 CORECO IN 0.98338
44.923 0.8480
1-0000 0.0000 0.0000 0.0544
2.404
42 FIFlY-PLUS
1
70 GENNUM Ci
1
576,578
0.0000 1.O000 0.0000 1-4697
1
I,WI,SS~ 0-0000 0.0000 1.0000 1.4977
76 LEITCH TEC
Royal's p e r s are Coreco and F i i MV nnge $2.4 to $45 million

upper bound estimate (based on efficient peerr):

$151,633 thousand

CYGNAL TECHNOLOGIES CORP (SIC3600)

DMU
Score Market Cal Efficient Peers:
#
$(OOOs)
30
32
38
39 CYGNAL TE 0.58919
8.637 0.0242
0.6881 0.2612
0.2631 0.6358
32,735 0.0052
61 SIERRA WIF 0.99635
585.308
1.O000 0.0000 0.0000
30 RESEARCH
1
11,832 0-0000 1.O000 0.0000
32 TS TELECO
1
0.0000 1.0000
3,855 0.0000
38 COMMUNIC
1
76 LEITCH TEC
1
.i.o31,ssg 0-0000 0.0000 0.0000
Cygnal is similar to Sierra and TS telecorn in terms of distance
Therefore, market value range: $1l,832,OOOto $32,735,000
UPPER BOUND:

Qgnal
76 distance
0.0265 0.0000
0.0959 0.3260
0.0000 1.4945
0.0000 O. 1668
0.0000 1.O205
1.O000 1.4899

Market Capitalization of CYGNAL based on EFFICIENT PEERS

$50,650 thousand

SIERRA WIRELESS. INC (SIC3600)

Sierra is similar to Comm, Cygnal and TS telecom
Market range $3,855,000 to $11,832,000

Siern
distance
0.32603
O
1-47229

UPPER BOUND:
0.9564
Market Capitalization of SIERRA based on EFFICIENT F 0.21112
$107,529 thousand
1.29087
-

---

VALU~NG
PRIVATE
COMPANIES:
A DEA APPROACH

i 59

DBA (SIC 3600)

DMU

Scom

22 DBA TELEC
18 ABL CANAD
13 TIR SYSTEh
12 A W A TECH
55 TRIPLE CR<
64 SYMPLEX C
32 TS TELECO
42 Flm-PLUS
60 PACIFIC E-L

0,63987
0.39914
0.65569
0.84726
0.52596
0.56217
1
1
1

Market Ca( Distan$(OOos)

1,956

13.382
1,710
2,429
4.979
5,879
11.832
2.404
3,878

Efficient Peers:
32
42
0 . m 0-14457 0.84501
0.3590 0,54321 0.40014
0 . ~ 1 70-11977 0.87756
0.0261 0.02450 0.95252
0.0286 0.25074 0-71533
0.0255 0.01430 0.90542
1 . ~ 5 91.00000 0.00000
0.04% 0.00000 1.OOOOO
1-7142 0.00000 0.00000

60
0,01040
0.05670
0.00267
0.02300
0.03390
0.08020
0.00000
0.00000
1.O0000

Distance
0.00000
0.35897
0.00173
0.0261 3
0.02864
0.02549
1.44591
0.04503
1-71425

Ranking, according to the distances that are les$ than 1 (for DBA):
DBA TELEC 0-63987
1.956
o.oo00
TIR SYSTEh 0.65569
1.ilO
0.0017
SYMPLEX C .OS6217
5.879
0.0255
A W A TECH 0.84726
2,429
0.0261
TRIPLE CR< 0,52596
4,979
0.0206
FIW-PLUS
1
2,404
0.0450
ABL CANAD 0.39914
13,382
0.3590
Similarity of the efficiency scores (difference between DBA's score ):
ABL CANAD 0.39914
13.382
0.3590
TRIPLE CRC 0.52596
4,979
0.0286
SYMPLEX C 0,56217
5.879
0.0255
DBA TELEC 0.63987
1,956
o.oo00
TIR SYSTEA 0.65569
1.71O
0.0017
AWATECH 0.84726
2,429
0.0261
FIFTY-PLUS
1
2,404
0.0450
Similarity of the Market Cap. (diflerence between DBA's market cap.):
TIR SYSTEA 0.65569
1,710
0.001 7
DBA TELEC 0.63987
1.956
o.oooo
FIFTY-PLUS
1
2.404
0.0450
A W A TECH 0.84726
2.429
0.0261
TRIPLE CRC 0.52596
4,979
0.0206
SYMPLEX C 0.5621 7
5,879
0.0255
ABL CANAD 0.39914
13.382
0.3540
Range for DBA's maricet cap. based on peer group is $1,710,000 and 513,382,000

Market Capitalization of DBA based on EFFICIENT PEERS

$3,782 thousand

~-

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

160

DIGITAL PROCESSING SYSTEMS INC (SIC3600)

DMU

Score

Market Cap
$(OOOs)

30

32

76

14 DIGITAL PR 0.88428
28.123 3.45E-03 0.85907 0-13748
1
585.308
1
O
O
1
11,832
O
1
O
76 LEITCH TEC
1
i,o~l.sss
O
O
1
Digital is only similar to TS telecorn, minimum of $1 1-8 million

30 RESEARCH
32 TS TELECO

0.00000
1.75002
0.03877
1.48196

UPPER BOUND:
Market Capitalization of DIGITAL based on EFFICIENT PEERS
$154,005 thousand

DIMENSIONS WEST ENERGY INC (SIC 3600)

DMU

Score

40 DIMENSION 0.1 1029

36 CELL-LOC Il 0.9803
2 TURBO GEt
1
30 RESEARCH
1
42 FIFIY-PLUE
1

Distance
DIMENSION
CELL-LOC Il
FIFW-PLUS
Eniciency:
DIMENSION
CELL-LOC I!
FIFTY-PLU3

0.1 1029
0.9803
1

Market Ca1Distance Efficiint Peenr

2
30
3,970

28.669
80,615
585.308

2.404

1,970

28,669
2.404

0.11029

1,970

0.9803
1

28,669
2.404

Market Value
DIMENSION 0.1 1029
FIFTY-PLUS
1
CELL-LOC Il 0.9803

42 Distance

1,970

2.404
28.669

Range for market value $2,404,000to $28,669,000

Upper Range
$12,241 thousand

VALUINGPRIVATE
COUPANIES:
A DEA APPROACH

161

DYNAMIC SIC 3600

DMU
#

Score Market Ca! Distance Efficient Peem

2
$(OOOs)

41 DYNAMI4 0.939

110,354

0.0000
0.0006
1.9377

1
~0.61s
2 TURBO (
sss,3as
1
30 RESEAR
Distance
DYNAMI4 0.939
110,354
o.oo00
TURBO (
1
80.615
0.0006
Minimum market value is $80,6lS,OOO
Upper Range
$89,291 thousand

0.98275
1.O0000
0.00000

30 Distance

0.01720
0.00000
'i.00000

0.00000
0.00059
1,93169

VALUATION ANALYSIS FOR EIGER TECHNOLOGY INC (SIC3600)

Score Market Ca! Efficient Peers
42
32
$(OOOs)
7
EIGER T 0.815
8.289
0.01070 O. 17981 0.55943
1.O0000 0.00000 0.00000
CAMCO
1
ioo.000
1.O0000 0.00000
TS TELE
1
11,832 0.00000
0.00000 0.00000 1.O0000
FIFlY-PL
1
2.404
PACIFIC
1
3,878
0.00000 0.00000 0.00000
Eiger is similar to Fifty-plus and Pacifie E-iink, therefore ts
market value is betwesn $2.4 to 3.9 million
Eiger seems overvalued at $8 million
Upper Bound
$5,512 thousand

DMU

60 Distance
0.25006 0.00000
0100000 1-38654
0-00000 1-04832
0.00000 0-28908
1-00000 0.90782

Valuation Analysis for EX1 TECHNOLOGIES INC (SIC3600)

Score Market Ca; Efficient Peers
DMU
38
#
$(OOOs)
30
58 EX1 TECI 0.757
9,422
0.01170 0.22403
1.O0000 0.00000
30 RESEAR
1
585.308
0.00000 1.O0000
38 COMMUI
1
3.855
0.00000
42 FIFTY-PL
1
2.404 0.00000
0.00000
76 LEITCH '
1
1,03i,559 0.00000
Minimum mv of $2,404,000 (one peer)

42
0.75986
0.00000
0.00000
1.O0000
0.00000

76
0.00443
0.00000
0.00000
0.00000
1.O0000

Distance
0.00000
1-60433
1-17967
O. 10802
1-61887

UPPER BOUND:
Market Capitaliution of EX1 bsed on EFFICIENT PEERS
$14,108 thousand

VALUINGPRNATECOMPANIES:
A DEA APPROACH

162

VALUATION ANALYSIS FOR ELECTROHOME LlMlTED (SIC3600)

DMU

Score

74 ELECTROHC 0.67333
6 LG TECHNO 0.586585
1 HAMMOND l 0.756773
7 CAMCO INC
1
32 TS TELECOI
1
34 CALIAN TEC 0.800142
67 PRIMETECH
1

Market Ca D i i n c e Efkient Peem

S(O00s) (displayed
7
32
ii,398

27,707
78.426

ioo,ooo
t 1,832

33,191
143.365

0.00000
0.15288
0.04998
1.28554
0.58451
0.05162
0.43784

0.0752
0.0060
0.2217
1.0000
0.0000
0.0599
0.0000

Distance
67 (formula)

0.4257
0.7302
0.4467
0.0000
1.O000
0.5935
0.0000

Similarity of eniciency scores (Distance of Electrohoome. to):

hammond
-0.083443
calian
4.126813
0.086745
LG Tech
prime tech
-0.32667
TS telecom -0.32667
similarity in Market Caps (Electrohome- M.C. minus):
hammond
-7,028
calian
-21,793
LG Tech
-16,309
prime tech
-131,967
TS telecorn
-435
Range for Electrohome market cap based on peer groups is $11,832,000 to $33,191,000

And Efectrohomes real market cap. of $1l,398,OOO falls below this range

UPPER BOUND:
Market Capitalization of ELECTROHOME based on EFFICIENT PEERS
Looks like:[(0.075'100,000) + (0.43'1 1,832)+(0-5'143,365)J

VALUING
PRIVATE
COMPANIES:
A DEA APPROACH

163

GEMSTAR COMMUNICATIONS INC (SIC 3600)

Score Maricet C Distance Efficient Peen
#
SCoooS)
2
42
43 CEMSTAR CC 0.2324
4.m
o.cxxm 0.0118 0.9882
48 SENSE TECH 0.3741
61.2.21
0.0452 0.1621 0,8379
75 LAMBUS E N 0.7096
1,729
o.s9 0.0660 0.9340
2 TURBO GEN!
1
80,615
1.9531 1,0000 0.oOOO
42 FIFTY-PLUS.1
1
2.404
0.0003
0.0000 1.0000
Distance
4.888
O-
CEMSTAR CC 0.2324
FIFP(-PLus.I
1
2.404
0.0003
WMBUS EIV 0.7096
1.729
0.0059
SENSE TECH 0.3741
61.221
0-0452
Efficiency:
CEMSTAR CC 0.2324
4.888
O-oo00
SENSE TECH 0.3741
61,221
0.0452
LAMBUS EW 0.7096
1,729
0.0059
FIm-PLus.I
1
2,404
0.0003
Market Value
LAMBUS EfV 0,7096
1,729
0.0059
FIFIY-PLUS.1
1
2.404
0-0003
CEMSTAR CC 0.2324
4.888
o.
SENSE TECH 0-3741
ai ,221
0.0452

DMU

~istance

0-0000
0.0452
0-0059
1-9531
0~0003

Range for market value S I ,729,000 ta $61,221,000

Upper Range

$3,327thousand

VALUATION ANALYSIS FOR GSI LUMONICS INC (SIC3600)

DMU

#
82
7
68
76

Scom Market Cap

SCoooS)

GSI LUMON14 0.91 67

CAMCO INC
1
CELESTICA II
1

2.8s.867

LEITCH TECk

1,031,559

i29.626

ioo,ooo

0.0196
1,0000
0.0000
0.ooOo

68

76 distance

0.01 35 0.9669 (3.0

0.0000 0.0000 f ,8962
1.0000 -0.0000 1.9084
0.0000 1. 0.0017

Minimum market value of $1,031,559,000, based on one peer

Upper bound:
$1,037,907 thousand

VALUATION ANALYSIS HAMMOND MANUFACTURING (SIC 3600)

Based on BCG1
DMU
#

Score

1 HAMMONC
6LGTECHN
7 CAMCO IN
32 TS TELECI
34 CALIAN TE
67 PRIMETEC
74 ELECTROI

0.7568
0.5866
1.O000
1.O000
0,8001
1.0000
0.6733

Market 4 Distance
Efficient Pmm
S(O00s) (displayed better)
7
32
18.4%
27,707
1oo.m
ii,832
33,191
143,365
11,398

0,0000
0.1315
0.91 52
0.4652
0.0480
0.6956
0.0500

0.221 7
O.O60
1.O000
0,0000
0.0599
0.0000
0.0752

0-4467
0.7302
0-0000
1-0000
0.5935
0.0000
0.4257

Distance
67 (formula)
0.331 5
0.2637
0.0000
0.0000
0.3466
1.O000
0.4990

0.0000
0.1315
0.91 52
0.4652
0.0480
0.6956
0.0500

Accordingly, Hammond k closest to (in distance):

CALIAN TE
0.0480
ELECTROI
0.0500
LGTECHN
0.1315
TS TELECf
0.4652
PRIMETEC
0.6956
CAMCO IN
0.9152
Similarity of efficiency scores (Harnrnond to):
Calian
-0.0434
Eiectrohorr
0.0834
LG Tech
0.1 702
TS
4.2432
4.2432
Primetech
Camco
-0.2432
Similarity in Market Caps (Hammond M.C. minus):
Calian
-14,765
Electrohorr
7,028
LG Tech
-9,281
TS
6,594
Primetech
-124,939
Camco
-81,574
Range for Hammonds market cap based on peer groups is $11,398,000 to $143,365,000
Ignon'ng the two biggest companies, wtiich are least sirnilar to Hammond in ternis of distance
and efficiency score, the range for Hammonds market cap is $11,398,000 to $33,191,000
Hamrnond's real market cap. of $18,426,000 falls nght between mis range

UPPER BOUND CALCULATION:

Looks like:[(O.22'100,000) + (0.45'1 l183S)+(O.33'I43,36S)]

$74,988 thousand
VALUING PRIVATE COMPANIES: A DEA APPROACH

165

Valuation Analysis of INVERPOWER (SIC 3600)

Score Market C.Distance Efficient Peem
#
$(000s) (displayec
30
32
better)
3 INVERPO\ 0.5446
9 AVANTE T 0.583
1
30 RESEARC
32 TS TELEC
1
33 ADVANTE 0.1804
42 FIFW-PLL
1
45 KASTEN C 0.9949
65 WI-LAN IN 0.8732
77 PLAINTRE 0.83
83 NUCTRON 0.5345
85 WE X.L. Hl 0.9437

DMU

Distance
42 Formula

Ranking, according to the distances that are Iess than 1 (for INVERPOWER):
TS Telecoi 0.011
Kasten
0.039
Plaintree
0.31
Similarity of the efficiency scores (diffemnce between INVERPOWER's score (0.545):
TS Telecoi -0.455
Kasten
-0.45
Plain4.322
INVERPOWER CONTROLS LTD's efficiency score is not close to the efficiency score of it's p e r s
Similarity of the Market Cap. (diffemnce between INVERPOWER's market cap.):
TS Telecoi 4,653 (7,179-1 1,832)
Kasten
-16,390 (7,179-23.569)
PIaintree -7,767 (7,179-14.946)
Range for INVERPOWER's market cap. based on peer group is $1 1,832,000 to $23,569,000
INVERPOWER CONTROLS LTD real market cap of $7,l79,OOO falls below this range

Market Capitalization of INVERPOWER based on EFFICIENT PEERS

[(0.0116*585,308)+ (0.92'1 1,832) + (0.0676'2,404)
is $q 7,84 thousand

VALUING
PRWATECOMPANIES: A DEA APPROACH

166

KASTEN CHASE APPLIED RESEARCH INC (SIC 3600)

ased on BCC-I for public company
DMU

Score

Market C.Distance Efficient Peem

$(000s) (displayec
30
32

Distance
42 Formula

45 KASTEN CHP
33 ADVANTEDG
9 AVANTE TEC
3 INVERPOWEl
30 RESEARCH II
32 TS TELECOM
42 FIFW-PLUSI
65 W-LAN INC
77 PLAlNTREE I
83 NUCTRON CC
85 WE XL, HOLi

Ranking, according to the distances that are Iess than 1 (for KASTEN):
lnverpower
0.0389
Ts Telecom
0.0852
Plaintree
0.1557
Wi-Lan
0.8240
Similarity of the eniciency scores (d'iflerence between KASTEN score (0.18):
lnverpower
0.4502
Ts Telecom
-0.0051
Plaintree
O. 1285
Wi-Lan
0.1217
KASTEN's efficiency score is very close to that of TS telecorn
Sirnilarity of the Matlcet Cap. (dfference between KASTENvsmarket cap. ($19,928,000):
lnverpower
16390.1 8
Ts Telecom
11736-78
Plaintree
8623.08
Wi-Lan
-305770.09
Range for KASTEN8smarket cap. based on peer group is between $7,179,000 and $329,339,000
KASTEN8sreal market cap of 523,569,000 falls between this range, lgnoring the biggest market
cap. company in the peer group (W-Lan), the range is between $7,179 and $14,946 million

KASTEN seems either slightly overvalued when one examines the market cap., however it is perforn
near the efficient frontier, with an efficiency score of 0.99, thus it may well be undervalued

Market Capitalization of KASTEN based on EFFICIENT PEERS

[(O. 11585,308) + (0.76'1 1.832) + (0.1 32,404)
is $75,454 thousand
-

VALUING
PRIVATE COMPANIES: A DEA APPROACH

167

Valuation Analysis of KING PRODUCTS INC (SIC3600)

DMU

Score

KING PRC 0.47195

PAClFlC E
1
SIMMONO 0.76631
SNlF SEC1
1

Market C. Efficient Peers:

)(ln9

Sirnrnond

0.214263 0.785737
O 0.8025
f 7-ISE-16 1.2348 0.0464
2.854 0.847691 0.152309 0.8025
O
307 1-62E-18
1 0.0918 1.4372
7,739
3,878

King is most similar to Simmonds and Snif security, therefore its

market value ranges from $307,000 to $2,854,000

UPPER BOUND:
Market Capitalization of KING based on EFFICIENT PEERS
$1,072 thousand
King is overvalued according to its peers

Valuation Analysis of SIMMONDS CAPITAL LlMlTED (SIC3600)

Simmond is closest to King and Pacific E-link
Market value range is $3.8to $7.7 million

Market Capitalization of SIMMOND based on EFFICIENT PEERS

$3,334 thousand
Sirnmond is undervalued according to its p e r s

MlCROCELL TELECOMMUNICATIONS INC (SIC3600)

DMU

Score

Market Cap
S(OO0s)

60 distance

26 MICROCE 6.64E-02 307.087 0.184519 0.815481

O
7 CAMCO Ib
1
100,000
1
O
1.33
60 PAClFlC E
1
3,878
O
1 0.0681

Microcell is at least $3,878,000, based on its one p r .

Upper bound
$21,615 thousand

VALUING PRlVATE COMPANIES:

A DEA APPROACH

DMU

Score

Market C.Distance Efficient Pwrs

(000s)

2
o.oo00 0.0660
0.0185 0.1621
o.wss 0.01 18
1.7448 1.0000
0.0087 0.0000

42 Distance

75 LAMBUS E 0.709557
1,729
0-9340 0.0000
SENSE TE 0.374065
61,221
0.8379 0.0185
GEMSTAF 0.232398
4,888
0.9882 0.0059
TURBO GI
1
80.61s
0.0000 1.7448
FIFTY-PL1
1
2,404
1.0000 0.0087
Distance
LAMBUS E 0.709557
1.729
o.oo00
GEMSTAF 0.232398
4.888
O-0059
FIFIY-P LC
I
2.404
0.0087
SENSE TE 0.374065
61.22t
0.01 85
Efficiency :
Market Value
GEMSTAG 0.232398
4,888
0.~59
LAMBU 0.7096
SENSE TE 0.374065
61,221
0.0185
FIFTY-F
1
LAMBUS E 0.709557
1,729
0.~00
GEMST 0.2324
Flm-PL1
1
2,404
0.0087
SENSE 0.3741
Range for market value $2,44,000 to $61,221,000
Upper Range
$7,566 thousand

48
43
2
42

1,729 0-0000
2.404 0.0087
4.888 0.0059
6 1 ~ 1 0.0185

FOR LG TECHNOLOGIES GROUP INC (6)

Market C.Distance
Efficient Peers
Distance
$(000s) (displayed better)
7
32
67 (formula)
6 LG TECHb 0.586585
27,707 0.00000
0.0060 0.7302 0.2637 1.4640
0.2217 0.4467 0.3315 0.8169
1 HAMMONI 0-756773
18,426 0.13148
1.0000 0.0000 0.0000 0.0044
7 CAMCO Ii+
ioo.ooo 1.59077
1
1
11.832 0.14237
0.0000 1.O000 0.0000 1.8724
32 TS TELEC
0.0599 0.5935 0.3466 1.2365
34 CAL~ANTi 0.800142
33.191 0.02846
0.0000 0.0000 1.O000 1.8724
143.365 1.07534
67 PRIMETE(
1
0.0752 0.4257 0.4990 1-1679
11,398 0.15288
74 ELECTRO (3.67333
Accordingly, LG TECH is closest to (in distance):
CALlAN TE 0.02846
HAMMONI 0.13148
TS TELEC 0.14237
ELECTRO 0.15288
PRIMETEC 1.07534 (VALUES greater than one are ignored the analysis)
CAMCO Ih 1.59077
Similarity in Market Cap:
Similarity of emciency scores (LG TECH. to):
Calian
4-213557
Calian -5,484
Electrohon -0.086745
Electrot 16,309
TS t e k 15,875
TS telecon -0.413415
Hammond -0.170187
Harnmo 9,281
Range for LG TECH'Smarket cap based on w r groups is $11,398,000to $33,191,000

DMU

Score

UPPER BOUND CALCULATION:

$47,055 thousand
-

VALUINGPRNATE COUPANIES: A DEA APPROACH

i 69

Market C Distance Efficient Peers:

S(OO0s)
42
60
62 distance
46 MARATHC 0,604497
623
o.oooo 0.801 57 0-128525 6.99E-02
31 TRADE W 0.676206
9,700
0.1781 0.46014 0,339707
0.2001 54
19 BCS TEC). 0-676619
1,027
0.0213 0-91557 4.1 TE-02 4.28502
0.0068 0.74767 0.1 20293 0.1 32032
12.193
56 TSI TELS\ 0.21 1188
0.0421 0.69985 6-42E-02 0.235936
1,MO
63 STRATEG 0.53425
1-13E-02 0.133211
0.0207 0.85545
15,698
71 SHELLCA! 8.47E-02
0.0452 0.971 58 1.31 E-02 1S3E-02
717
81 COMPREZ 0-310462
42 FIFIY-PLL
1
2,404
0.0608
1
O
O
60 PAClFlC E
1
3,878
1.4069
O
1 7.15E-16
62 SNlF SEC1
1
307
1.5241
O 1.62E-18
1
Ranking, according to the distances that are Iess than 1 (for MARATHON):
623
o.oo00
MARATHC 0.604497
12,193
0.0068
TSI TELS\ 0.21 1188
15.698
0.0207
SHELLCA! 8.47E-02
1,027
0.0213
BCS TECi- 0.67661 9
1,720
0.0421
STRATEG 0.53425
0.0452
717
COMPREZ 0.31 0462
FIFTY-PLL
2,404
0.0608
1
TRADE Wi 0-676206
9,fm
0.1781
Similarity of the efficiency scom (difference between MARATHON'S score ):
SHELLCA! 8.47E-02
15.698
0.0207
TSl TELS\ 0.21 1188
12,193
0.0068
COMPREZ 0.310462
717
0.0452
STRATEG 0-53425
1,720
0.0421
MARATHC 0.604497
623
o.oo00
TRADE Wl 0.676206
9,700
0.1781
BCS TECi- 0-676619
1.027
0.0213
FIFTY-PLL
1
2,404
0.0608
Similarity of the Market Cap. (diffsrence between MARATHON'S market cap.):
MARATHC 0.604497
623
o.oo00
COMPREI 0.310462
717
0.0452
BCS E C i - 0.67661 9
1,027
0.0213
STRATEG 0.53425
1,720
0.04~1
FIFPI-PLL
1
2,404
0.0608
TRADE Wl 0.676206
9,700
0,1781
TSl TELSI 0.21 1188
12,193
0.0068
SHELLCA: 8.47E-02
15.698
0.0207
Range for MARATHON'Smarket cap. based on peer group is $717,000 to $15,698,000
Marathon seems to be undefvalued

DMU

Score

Market Capitalization of MARATHON based on EFFICIENT PEERS

$2,447 thousand

VALUINGPRIVATE COMPANIES:
A DEA APPROACH

170

Valuation Analysis for MEMOTEC (SIC3600)

DMU
Score Market C. Efficient Pwrs
#
S(O00s)
32
67
76 distance
25 MEMOTEC 0,6492
4.953 0.8199
0.1655
0.0146
0.0000
YOSPECTRU 0-7867
38.637 0.8465
0.0760
0.0775
0-0127
57 WESCAM 0.6285
76,678 0.1979
0-7437
0-0584
0.7232
32 TS TELEC
1
11.832 1-0000
0.0000
0-0000
0,0601
67 PRIMETE(
1
143.365 0.0000
1-0000
0-0000
1-3688
76 LEITCH TE
1 1,031,559 0.0000
0-0000
1.0000
1-6706
Memotec's p e r s are: Spectnim, TS Tel, and Wescam
Market Value
Range $11,832,660 to $76,618,000
Upper bound
$48,488 thousand

Valuation Analysis for SPECTRUM (SIC3600)

Score Market C.Efficient Peem
S(00Os)
32
67
76 disbnce
50 SPECTRU 0.7867
38,637 0.8465 7.60E-02 7.75502
0,00000
0.01268
25 MEMOTEC 0.6492
4,953 0.8199 0.165499 1.46E-02
0.86696
76,678 0-1979 0-743748 5.84E-02
57 WESCAM 0.6285
O
0.03535
O
32 TS TELEC
1
11,832
1
O
1-57633
143,365
O
1
67 PRIMETEC
1
1.57333
O
1
76 LEITCH TE
1 1,031.ss9
O
Spectmm is nearest Memotec, Ts Tel and Wescam
Market Value
Range $4,953,000 to $76,678,000
Upper bound
$100,858 thousand

DMU
#

Valuation Analysis for WESCAM (SIC3600)

Score Market C Efficient Peers
67
76 distance
$(OOOs)
32
0.7437
0.0584
0.0000
57WESCAM 0.6285
76,678 0.1979
0.0760
0.0775
0.8670
50 SPECTRU .O-7867
38,637 0.8465
0.1655
0.0146
0.7232
25 MEMOTEC 0.6492
4,953 0.8199
0.0000
0.0000
1-2000
32 TS TELEC
1
11,832 1-0000
1.O000
0.0000
O. 1082
67 PRIMETEC
1 143,365 0.0000
0.0000
1.O000
1.4789
76 LEITCH TE
1 1,031,559 0.0000
Wescam is nearest Primetech, Memotec then Spectnim
Market Value
Range $4,953,000 to $143,365,000

DMU
#

Upper bound
$169,212 thousand
-

VALUING PRIVATE COMPANIES:

A DEA APPROACH

171

Score Market f Distant Efficient Pmws

S(O00s) (displayr
30
32
better)
0-0032
0.0452
83 N W R O 0.534
2.426 0 . 0 0 0
0.0884
0.4958
77 PLAlNTF 0.866 14,945 o.4gna
0.0359
0.1599
65 WI-LAN I 0.873 329.339 0.03594
0.0116
0.9208
3INVERPC 0.545
7,179 1.54824
0.0028
0.0210
9 AVANTE 0.583
2.302 0.00120
0.0000
30 RESEAR
1 sas.308 1 . ~ 1 1 2 1.O000
0.0000
1.0000
11.832
1.81716
32 TS TELE
1
O-0000
O.OOOO
2,404 O.Oo440
42 FIFTY-Pl
1
0.01 17
0.0891
33 ADVANT 0-18 19,928 0.00474
0.1
130
0.7617
1.20819
45 KASTEN 0.995
23.569
0.0088
0.0060
85WEX.L.l 0.944
17.083 0.00271

DMU
#

Distance
42 Formula

Ranking, according to the d i n e e s that are Iess than 1 (for NEXTRON):

NEXTRO 0.534
2,426 o.owoo
2.302 0.00120
AVANTE 0.583
WE X-L. I 0.944
17.083 0.00271
2.404 0 . O W
Fim - P L
1
ADVANT 0.18
19,928 0 . 0 ~ 7 4
WI-LAN I 0.873 329.339 0.03594
PLAINTF 0.866
14,946 0.49728
Similarity of the efficiency scores (diffemnce between NEXTRON's score (0.W):
ADVANT 0.1 8 19,928 0.00474
NEXTRC 0.534
2.426 0 - a m w
AVANTE 0.583
2,302 0.00120
PLAINTF 0.866
14,946 0 . 4 9 ~ 8
WI-LAN I 0.873 329,339 0.03594
WE X.L. 1 0.944
17,083 0.00271
FIFTY-Pl
1
2,404 0.0044O
Similarity of the Market Cap. (dinerence between NUCInON's market u p . ($2,426,000):
2.302 0.00120
AVANTE 0.583
FIFTY-Pl
1
2,404 O - O W
2,426 o.oooou
NEXTRC 0.534
1 4 . ~ 6 0.49728
PLAINTF; 0.866
WE X.L. I 0.944 17,083 0.00271
ADVANT 0.18
19.928 0.00474
WI-LAN I 0.873 329,339 0.03594
Range for NEXTRONr market cap. based on peer gmup is $2,302,000 and 5329,339,000
NUCTRON's real market cap of $2,426,000 k right in this range

Market Capitalization of NEXTRON based on EFFICIENT PEERS

$4,707 thousand

VAUJINGPRNATE COMPANIES:A DEA APPROACH

172

Valuation Analysis for NHC COMMUNICATIONS INC (SIC3600)

DMU

Score

28
30
42
44
76

Market C.Efficient Pwrs:

S(000s)
30
42

44

76 D h n c e

NHC COML 0.670855

7,317 1.32E-02 0.92567 0.0279
3.33E-02 0.0000
O
O 1.8325
1 6-18E-16
RESEARCF
1
585.308
O
O 0.0076
FIFTY-PLU:
1
2.404
O
1
O 1-3OE-17
1
O 1.8032
1
3.893
INTERNAT1
LEfTCH TE(
1 l,o3i.sss
O
O
O
1 1-7923
NHC is only similar to Fifty plus
therefore, minimum range is $2,404,000

UPPER BOUND:
Market Capitalization of NHC based on EFFICIENT PEERS
W,4ll thousand

Valuation Analysis for NORON TECH LTD (SIC3600)

DMU

Score

8 NORON TE 6.53E-02
42 FlFlY-PLU!
1
62 SNlF SECU
1

Market C. Efficient Peem

$(OOOs)
42
7,029

2.404
307

62

distancDisbnce
0.98995 1.Oq E-02 0.0000
O
1
O 0.0002 0.000203
O
1 1.9599 1.959903

Noron only has F i i plus close to it in terms of distance, therefore its market
value should be near $2.4 million
Upper Bound
$2,383 thousand
According to the upper bound Noron is ovenralued

VALUATlON ANALYSIS FOR SEMI-TECH CORPORATION (SIC 3600)

DMU

Score

Market Cap
$(OOOs)

68 Distance

10 SEMI-TECF 0.214508
10,026 0.?74623 0.22538
O
7 CAMCO IN<
1 100,006
1
O 0.1016
68 CELESTICE
1 2,854,867 2.81 E-16
1 1.2001
In terrns of distance only Camco is close, therefon, lower bound of $100 million

Upper bound:
$720,885 thousand

VALUING
PRIVATE
COMPANIES: A DEA APPROACH

1 73

PLAINTREE (SIC 3600)

Score Market CgD i n c e Efficient Peem
Distance
$(OOoS)
30
32
42 Formula
77 PLAINTRE 0-8663
14,946 o.oom 0.0884 0-4958 0.4159 0.0000
65 WI-LAN IN 0-8732 329,339
0.26639 0.0359 0.1599 0.8042 0.2664

DMU
#

3 INVERPO\ 0,5446
7.179
0.30786 0.0116 0.9208 0.0676 0.3079
9 AVANTE T 0.583
2 . 3 ~ 0.54681 0-0028 0.0210 0.9763 0.5468
30 RESEARC
1 585,308
1.24976 1.O000 0.0000 0.0000 1.2498
32 TS TELEC
1
11.832
0.43498 0.0000 1-0000 0.0000 0.4350
42 FIFTY-PLL
1
2,404
0.59484 0.0000 0.0000 1-0000 0.5948
33 ADVANTE 0.1 804
19,928
0 . 4 ~ 9 10.01 17 0.0891 0-8992 0.4049
45 KASTEN C 0.9949
23.569
0.15575 0.1 130 0.7617 0.1253 0.1557
83 NEXTRON 0.5345
2,426
0.49728 0.0032 0.0452 0.9516 0.4973
85 W E XL. Hi 0.9437
17,083
0.57049 0.0088 0.0060 0.9853 0.5705
Ranking, according to the distances that are Iess than 1 (for PLAINTREE):
PLAINTRE 0.8663
1 4 , s ~ O-owoo
KASTEN C 0.9949
23.569
0.15~75
WI-LAN IN 0.8732
329.339
0.26639
INVERPO\ 0.5446
7.179
0.30786
ADVANTE 0.1804
19.928
0.4~91
TS TELEC
1
11,832
0.43498
NEXTRON 0.5345
2,426
0.49728
AVANT T 0.583
2.302
0.54681
WE XL, Hl 0.9437
17,083
0.57049
FIm-PLL
1
2.404
0.59484
Similarity of the Market Cap.:
Similarity of the efficiency scores:
AVANTl 0.583
ADVANTE 0-1804
FIFPI-f
1
NEXTRON 0-5345
NEXTR! 0.5345
INVERPO\ 0.5446
l NVERF 0.5446
AVANTE T 0.583
TS TEU
1
PLAINTRE 0-8663
PLAINT 0.8663
WI-LAN IN 0.8732
WE X.L 0.9437
W E XL. Hl 0.9437
ADVAN
0-1804
KASTEN C 0.9949
KASTEl 0.9949
TS TELEC
1
WI-LAN 0.8732
FIFPI-PLL
1

Range for PLAlNTREE's market cap. based on peer gmup is $2,302,000 and $329,339,000

Market Capitalization of PLAINTREE based on EFFICIENT PEERS

$58,607 thousand

VALUI
NG PRIVATE
COMPANIES:
A DEA APPROACH

174

SENSE TECHNOLOGIES INC (SIC 3600)

Score Malket C.Distance Efficient Pwrs
2
42 Distance
SENSE TE 0.3741
o.oooo 0.1621 0.8379 0.0000
GEMSTAE 0.2324
0.0452 0.01 18 0-9882 0.0452
LAMBUS E 0.7096
0.0185 0.0660 0-9340 0.0185
TURBO Gi
1
1.4040 1.0000 0.0000 1.4040
FIFTY-PLC
1
0.0526 0.0000 1.O000 0.0526
Distance
SENSE TE 0.3741
LAMBUS E 0.7096
GEMSTAF 0.2324
1
FIFTY-PLI.
Eficiency:
GEMSTAF 0-2324
SENSE TE 0.3741
LAMBUS E 0.7096
FIm - P L 1
1
Market Value
LAMBUS E 0.7096
FIFTY-PLI.
1
GEMSTAF; 0.2324
SENSE TE 0.3741
Range for market value $1,729,000to $4,888,000

DMU
48
43
75
2
42

Upper Range
$15,085 thousand

Valuation Analysis for STRATEGIC TECHNOLOGIES INC (SIC3600)

Score Market C.Enicient Peers
#
$(OOOs)
5
7
52 STRATEG 0.9832
3,943
0.3983 0.0012
1
1 3 . 2 ~ 1.0000 0.0000
5 GLOBAL 1
1
i00,ooo 0.0000 1.0000
7 CAMCO ICI
42 FIFW-PLI.
1
2,404
0-0000 0.0000
59 JEMTEC II
1
330
0.0000 0.0000
DMU

42
0.4871
0.0000
0.0000
1.0000
0.0000

59 Distance
0.1 133 0.0000
0.0000 0.6122
0.0000 1.4064
0-0000 0.4345
1.0000 1.1822

In terms of distance, sirnilar to Global and Fifty plus

Range for market value $2,4U4,000 to $13,222,000

Upper bound
$6,597 thousand

VALUINGPRIVATE
COMPANIES:A DEA APPROACH

175

Based on BCC-I for public Company SHELLCASE (SIC 3600)

DMU
#

Scom

Market C.Distance Efficient Peem

$(OOOs)
42
60

62 distance

71 SHELLCA: 8.47E-02
is.698
0 . m 0.8555 0.01 13 0.1332 0-0000
63 STRATEG 0.53425
1,720
0.0376 0.6999 0.0642 0.2359 0-0376
56 TSI TELSI 0.21 1188
12,193
0.0235 0.7477 0-1203 0.1 320 0-0235
46 MARATHC 0.604497
623
0.0207 0.8016 0-1285 0.0699 0-0207
31 TRADE Wi 0.676206
9.700
0.2686 0.4601 0.3397 0.2002 0.2686
1.027
0-0127 0.9156 0.0417 0.0428 0-0127
19 BCS TECk 0.67661 9
81 COMPREI 0.310462
717
0.0274 0.9716 0.0131 0,0153 0-0274
42 FIFlY-PL!.
1
2.404
0.0388 1.0000 0.0000 0.0000 0.0388
1
3,878
1.7271 0.0000 1,0000 0.0000 1.7271
60 PAClFlC E
62 SNlF SEC1
1
307
1.4832 0.0000 0.0000 1.O000 1.4832
Ranking, according to the distances that are lem than 1 (for SHELLCASE):
O-oooo
SHELLCA: 8.47E-02
I5.698
1,027
0.0127
BCS TECk 0.676619
623
0.0207
MARATHC 0.604497
12,193
0.0235
TSI TELS) 0.21 1188
COMPREZ 0.310462
717
0.0274
STRATEG 0.53425
1,720
0.0376
FIm-PLl
1
2,404
0.0388
TRADE Wi 0.676206
9,700
0.2686
Similarity of the eficiency scores (diHerence between SHELLCASE's score ):
SHELLCA: 8.47E-02
is.698
o.oooo
TSI TELSt 0.21 1188
12.193
0.0235
COMPREZ 0.310462
717
0.0274
STRATEG 0.53425
1.720
0.0376
MARATHC 0.604497
623
0.0207
TRADE WI 0.676206
9.700
0.2686
BCS TECF 0.67661 9
1,027
0.0127
FIFW-PL1
1
2,404
0.0388
Similatity of the Market Cap. (difference between SHELLCASE's market cap.):
MARATHC 0.604497
623
0.0207
COMPREZ 0.310462
717
0.0274
8CS TECk 0.67661 9
1,027
0.0127
STRATEG 0.53425
1.720
0.0376
FIFTY-PL1
1
2,404
0.0388
9,700
0.2686
TRADE WI 0.676206
12,193
0.0235
TSI TELS) 0.21 1188
15,698
O-OO
SHELLCA! 8-47E-02

Range for SHELLCASE's market cap. based on peer group is $623,000to $12.1 93,000

Market Capitaliution of SHELLCASE based on EFFICIENT PEERS

$2,141 thousand

VALUATION ANALYSIS FOR SR TELECOM INC (SIC3600)

DMU
#

Score Market C Efficient Peers:

$(OOOs)
7
67

51 SR TELEC 0.447
i12,897
44,404
80 BLUESTAI 0.5548
37 COM DEV 0.8633
252.31 5
54 TEKLOGI) 0.8099
243,501
7 CAMCO It+
1
iw,m
67 PRIMETE(
1
143,365
76 LEITCH TE
1 1.031,!~9

0.1363
0.0926
0.0824
0-0755
1.0000
0.0000
0.0000

Companies anked by Distance:

SR TELEC 0.447
112,897
O
COM DEV 0.8633
252,315 0.2213
BLUESTAI 0.5548
44,404 0.3699
TEKLOGI) 0.8099 -- 243,501 0.3753
LEITCH TE
1 1,03l,sss 0.5029
PRIMETEC
1
143,365 0.5534
Rank by Eniciency score
SR TELEC 0.447
112,897
O
BLUESTAI 0.5548
44,404 0.3699
TEKLOGI) 0.8099
243,501 0.3753
COM DEV 0.8633
252.315 0.2213
LElTCH TE
1 1,031,559 0.5029
PRIMETEC
1
143,365 0.5534
By market value
BLUESTAI 0.5548
44,404
SR TELEC 0.447
112,897
PRIMETE(:
1
143,365
TEKLOGI) 0.8099
243,501
COM D N 0.8633 . 252.315
LEITCH TE
1 1.031,!%9

Upper bound
$532,289 thousand

0.3699
O
0.5534
0-3753
0.2213
0.5029

0.4192
0-8695
0.1 168
0,8796
0.0000
1.0000
0.0000

76 distance
0.4445
0.0379
0.8008
0.0449
0.0000
0.0000
1.O000

0-0000
0-3699
0.2213
0-3753
1-1194
0,5534
OS029

STRATEGIC VISTA (SIC 3600)

DMU
Score
Market CeDistance Eflicient Peers:
#
$(OOOs)
42
60
62 distance
63 STRATEG 0.53425
1,720
o. 0.6999 0.0642 0.2359 0.0000
12.193
0.0162 0.7477 0.1 203 0.1 320 0.0162
56 TSI TELS) 0.21 1188
46 MARATHC 0.604497
623
0.6421 0.8016 0.1285 0.0699 0.0421
31 TRADE W 0.676206
9,700
0 . 1 0.4601
~
0.3397 0.2002 0.1 346
19 BCS TECF 0.676619
1.027
0.0843 0.9156 0.0417 0.0428 0.0843
71 SHELLCA! 8.47E-02
15,698
0.0376 0.8555 0.0113 0.1332 0.0376
81 COMPREL 0.310462
717
0.1251 0.9716 0.0131 0.0153 0.1251
42 FIFW-PLL
1
2,404
0.1499 1.0000 O-OOOO O.OOOO 0.1499
60 PAClFlC E
1
3,878
1.4212 0.0000 1.O000 0.0000 '1-4212
62 SNlF SEC1
1
307
i . o m 0.0000 0.0000 1.O000 1.0777
Ranking, according to the distances thrt are les8 than 1 (for STRATEGIC VISTA):
STRATEG 0.53425
1,720
o.w
12.193
0.0162
TSI TELS) 0.21 1188
i5,698
0.0376
SHELLCA! 8.47E-02
MARATHC 0.604497
623
0.0421
BCS TECk 0.67661 9
1,027
0.0843
COMPRE: 0.310462
717
0.1251
9 . 7 ~ 0.1346
TRADE Wl 0.676206
FIFTY-PU
2,404
0-1499
1
Similarity of the efficiency scores (difference between STRATEGIC VISTA'S score ):
SHELLCA! 8.47E-02
is.698
0.0376
TSlfELSi 0.211188
12,193
0.0162
717
0.1251
COMPREZ 0.310462
STRATEG 0.53425
1,720
o.~~
623
0.0421
MARATHC 0.604497
9,700
0.1346
TRADEWi 0.676206
1.027
0.0843
%CSTECk 0.67661 9
2,404
0.1499
FIITY-PLL
1
Similarity of the Market Cap. (difference between STRATEGlC VISTA'S market cap.):
MARATHC 0.604497
623
0.0421
COMPREd 0.310462
717
0.1251
BCS TECk 0.676619
1,027
0.0843
STRATEG 0.53425
t ,720
0.0000
FIFW-PLC
1
2,404
0.1499
TRADE Wl 0.6762%
9,700
0.1346
TSI TELS\ 0.21 1188
12.193
0.0162
SHELLCA: 8.47E-02
15,698
0.0376
Range for STRATEGIC VISTA% market cap. bued on peer gmup is $23,000 to $15,698,000

Market Capitalization of STRATEGlC VISTA based on EFFICIENT PEERS

$2,004 thousand

VALUING
PRIVATE
COMPANIES:
A DEA AFPROACH

178

DMU
#

Score Market 1 Distanc Etlicknt Peers

%(OOOs)
32
42

64 SYMPLE
55 TRIPLE (
22 DBA TEL
18 ABL CAh
13 TIR SYS
12 A W A TE
32 TS TELE
42 FIFTY-PI
60 PAClFlC

0.562
0.526
0.64
0.399
0.656
0.847
1
1
1

60 Distance

Ranking, according to the distancm that are less than 1 (for SYMPLEX):
A W A TE 0.847
2.429 0.0056
FIFTY-PI
1
2,404 0.0156
TIRSYS0.656
1,710
0.0179
DEA TEL 0.64
1.9s
0.02~5
TRIPLE(0.526
4.979
0.0942
ABL CAh 0.399 13,382 0 . 5 s
Similarty of the efticiency scores (difference between SYMPLEX's score ):
ABL CAh 0.399 13,382 0.5356
4.979
0.0942
TRIPLE(0.526

3
r

-<.--&fi'

,--

DBA TEL- 0.64

1.956 0.0255
TIR SYS 0.656
O
0.0179
A W A TE 0.847
2.429 0.0056
FIm-Pi
1
2,404
0.0156
Similarity of the Market Cap. (difference between SYMPLEX's market cap.):
TIR SYS 0.656
1,710
0.0179
DBA TEL 0.64
1.9%
0.02~~
FIFTY-PL
1
2.404 0.0156
A W A TE 0.847
2.429
0.0056
TRIPLE ( 0.526
4,979
0.0942
ABL CAh 0.399

13.382

0.5356

Range for SYMPLEX's market cap. based on peer group is $1,Tl 0,000 and $13,382,000

Market Capitalization of SYMPLEX based on EFFICIENT PEERS

$2,657 thousand
-

VMUING PRIVATE COMPANIES: A DEA APPROACH

179

SYSCAN INTERNATIONAL INC (SIC 3600)

DMU
#

Score Market Cap

$(OOOs)

53 SYSCAN Il 0.6574
1
17 AASTRA T
38 COMMUN1
1
42 FIFTY-PLL
1
67 PRIMETEC
1

17

38

42

0.052205 0.423308 0,521971

O
1
O
3,855
O
1 1.29E-18
1
O
2.404
O
O
143,365
O 8.70E-17
7.260

46.091

67 Distance
2-52E-03
O
O
O
1

0.0000
1.3500
0.6078
0.4104
1.4493

Range for mv: $2.4 to 3.8 million

Upper bound
$5,654 thousand

VALUATION ANALYSE FOR TEKLOGIX INTERNATIONAL INC (SIC3600)

DMU
#

Score Market C. Efficient Peers:

$(OOOS)
7

67

76 distance

54 TEKLOGI) 0.8099
243,501 7.55E-02 0-879623 4.49E-02
112.897 0.13625 0.419239 O.444511
51 SR TELEC 0.447
80 BLUESTAi 0.5548
44.404 9.26E-02 0.869458 3.79E-02
252.315 8.24E-02 0.1 16802 0.800779
37 COM DEV 0.8633
1
O
O
loo,ooo
1
7 CAMCO Ib
O
1
O
67 PRIMETEC
1
143.365
O
O
1
76 LEITCH TE
1 1,031,559

Three peers: Bluestar, Primetech, and SR Tel

Market value range: $44,404,000 to $143,365,000

Upper bound
$179,974 thousand

0.00000
0.37533
0.00044
1.15330
1.63045
0.02221
1.69165

Public Company TIR (SIC 3600)

DMU
Score Market C.Distance Efficient Peers
#
$(OOOs)
32
42
60 Distance
13 TIR SYSTE 0.6557
1,710
0.000 0,1198 0.8776 0.0027 0.0000
2,429
0.0151 0-0245 0,9525 0.0230 0.0151
12 A W A TEC 0.8473
18 ABL CANP 0.3991
13,382
0.4101 0-5432 0.4001 0.0567 0-4101
55 TRIPLE CF 0.526
4,979
0 . w 0.2507 0.7153 0.0339 0.0444
22 DBA TELE 0.6399
1.95s
0 . ~ 1 70.1446 0.8450 0.0104 0.0017
64 SYMPLEX 0.5622
5.879
0.0179 0-0143 0.9054 0.0802 0.0179
32 TS TELEC
1
11.832
1.5449 1-0000 0.0000 0.0000 1.5449
42 FIFW-PLL
1
2,404
0.0293 0-0000 1.0000 0.0000 0.0293
60 PAClFfC E
1
3,878
1.7791 0.0000 0.0000 1.0000 1.7791
Ranking, according ta the diaances that are I e s s than 1 (for TIR):
TiR SYSTE 0.6557
l.710
0.~00
DBA TELE 0.6399
1,956
0.0017
A W A TEC 0.8473
2.429
0.0151
SYMPLEX 0.5622
5.879
0.0179
FIFTY-PLL
1
2,404
0.0293
TRIPLE CF 0.526
4.979
0.0444
ABL CANP 0.3991
13,382
0.4101
Similarity of the efficiency scores (difTerence between l R s score ):
ABL CANP 0.3991
13,382
0.4101
TRIPLE CF 0.526
4.979
0.0444
SYMPLEX 0.5622
5.879
0.0179
DBA TELE 0.6399
1,956
o.0017
TIR SYSTt 0.6557
1.710
o.oow
AWA TEC 0-8473
2,429
0.0151
FIFTY-PLL
1
2.404
0.0293
FIFTY-PLL
1
2,404
0.0034
Similarity of the Market Cap. (difference between TIR'S market cap.):
TIR S Y S 0.6557
~
1.710
0.o000
1,956
0.0017
DBA TELE 0.6399
2,0.0293
FIFTY-PLL
1
FIFW-PLL
1
2,404
0.0034
2,429
0.0151
A W A TEC 0.8473
4,979
0.0444
TRIPLE CF 0.526
5.879
0.0179
SYMPLEX 0.5622
13,382
0.4101
ABL CANP 0,3991

Range for TIR'. market cap. based on p m group k S1,956,000 and S13,382,000
TIR'S mal market cap of $1,7t 0,000 falls just short of this amount

Market Capitalkation of TIR based on EFFICIENT PEERS

$3,537 thousand
VALU~NG
PRIVATE COUPANIES:
A DEA APPROACH

181

f RADE WlND COMMUNICATIONS LTD (SIC 3600)

DMU
#

Scom

Market C Distance Efficient Pmrs

$(OOOs)
42
60

62 distance

31 TRADEW
19 BCS TECk
46 MARATHC
56 TSI TELS)
63 STRATEG
71 SHELLCA.'
81 COMPREZ
42 FIFTY-PLL
60 PAClFlC E
62 SNlF SECI
Ranking, according to the distances mat are le- than 1 (for TRADE WIND):
TRADEW
0:67@tL
6700? -=P- - 1.720
0-1346
0.53L)25
STRATEG
TSITELSiI
0.21119
12.193
0-1355
623
0.1781
0.60450
MARATHC
is.698
0.2686
0.08470
SHELLCA.'
1,027
0.3210
0.67662
BCS TECi717
0.4024
0.31 046
COMPREE
FIFW-PLL
2.W
0.4469
1.OOOOO
PACIFIC E
1-00000
3,878
0.6878
SNlF SECI
1.00000
307
0.9669
Similarity of the Market Cap.:
Similarity of the efficiency scores:
SHELLCA:
0.08470
15.698
0.2686
SNIF Sf
1
307 0.9669
TSI TELS)
0.21 119
12.193
0.1355
MARAT 0.6045
623 0.1781
COMPREZ
0.31 046
717
0,4024
COMPF 0.31046
717 0.4024
SlTWTEG
0.53425
1,720
0.1346
BCS TE 0.67662
1,027 0.3210
STRATt 0.53425
1,720 0-1346
MARATHC
0.60450
623
0.1781
TRADE
s . o o ~ j ~ ~ ~ UFIF~ l Y - F
1
2,404 0.4469
0.676Zf
PACIFI(:
1
3,878- 0.6878
-..1,027
0.3210
BCS TECF - 0.67662
2.0.4469
T R A D ~0 - 6 7 6 2 f 5 9,h1&.-- O ~..-I O -O
1-00000
FIFTY-PL1
PAClFlC E
1-00000
3.878
0.6878
TSITEL 0.21119 1z.193 0.1355
SNIF SEC1 1.00000
307
0.9669
SHELLC 8.47E-02 1S,698 0.2-

7
-

# T < F -

"

<

Range for TRADE WND's market cap. b a s d on peer group is $307,000 to $15,698,000

Market Capitalization of TRADE WlND based on EFFICIENT PEERS

$2,485 thousand

For public Company TRIPLE CROWN (SIC 3600)

DMU
Score Market C Distance Etlicient Peem:
#
S(0OOs)
32
42
55 TRIPLE CF 0.526
4,979
o.oo00 0.2507 0.7153
22 DBA TELE 0.6399
1.956
0.0286 0.1446 0.8450
18 ABL CANP 0.3991
13.382
0.1851 0.5432 0.4001
13 TIR SYSTE 0.6557
1.710
0.0444 0.1198 0.8776
2.429
0.1076 0.0245 0-9525
12 AWA TEC 0.8473
64 SYMPLEX 0.5622
5.879
0.0942 0.0143 0.9054
32 TS TELEC
1
17,832
1.0742 1.0000 0.0000
42 FIFW-PL1
1
2,404
0-1451 O.OOOO 1.OOOO
60 PACIFIC E
1
3.878
1.5079 0.0000 0.0000

60
0.0339
0.0104
0.0567
0.0027
0.0230
0.0802

Distance

0.0000
0-0286
0.1854
0-0444
0-1076
0.0942
0-0000 1-0742
O.OOOO 0.1451
1.0000 1.5079

Ranking, according to the distances that am lemm than 1 (for TRIPLE CROWN):
TRIPLE CF 0.526
4,979
o.oo00
DEA TELE 0.6399
1.956
0.0286
TIR SYSTt 0.6557
1,770
0-.
SYMPLEX 0.5622
5.879
0.942
AWA TEC 0.8473
2,429
0.1076
FIFW-PL1
I
2,404
0.1451
ABL CANP 0.3991
13.382
0.1854
'

Similarity of the efficiency scores (difference between TRIPLE C R O W s scom ):

ABL CANP 0.3991
13.382
0-1854
TRIPLE Ci 0.526
4.979
o.oo00
SYMPLEX 0.5622
5.879
0-0942
DEA TELE 0.6399
1,956
0.0286
TIR SYSTt 0.6557
1.710
0.6444
A W A TEC 0.8473
2,429
0.1076
FIFTY-PL1
1
2.404
0-1451
Similarity of the Market Cap. (difference between TRIPLE CROWN's market crp.):
TIR SYSTt 0.6557
1.710
0.044
DBA TELE 0.6399
1,956
0.0286
FIW-PLL
1
2,404
0.1451
2,429
0.1076AWA TEC 0.8473
4,979
0.00oo
TRIPLE CF 0.526
5.879
0.0942
SYMPLEX 0.5622
13,382
0.1854
ABL CANP 0.3991
Range for TRIPLE CROWNUsmarket cap. based on peer g i w p k $1,lO,OOO

- $13,382,000

Market Capitalization of 1RlPLE CROWN based on EFFICIENT PEERS

$4818 thouiand

VALUINGPRNATE COMPANIES:A DEA APPROACH

183

Valuation Analysis for TRI-VISION INTERNATIONAL LTD (SIC3600)

DMU

Score Market C Enicient Pwm:

$(OOOs)
30
32
TRI-VISIOI 0.7426
29,930 0.0850 0-1415
RESEARC
1 !585,308 1.0000 0-0000
TS TELEC
1
11.832 0.0000 1.O000
COMMUN1
1
3,855 0.0000 0,0000
FIFPI-PLL
1
2.404 0.0000 0,0000

38
0.1094
0.0000
0.0000
f -0000
O.OOOO

42
0.6-1
0.0000
0.0000
0.0000
1.OOOO

distance
0.0000
1.3101
1.1973
1.2614
0.1521

Minimum mv $2.4 million, one peer

UPPER BOUND:
Market Capitalization of TRI based on EFFICIENT PEERS
$53,470 thousand

TUNDRA SEMICONDUCTOR CORP (SIC3600)

Score Market C Effkient Peers
S(000~)
32
38
72 TUNDRA Z 0.9817 120,705 0.1OO4 0,7976
32 TS E L E C
1
11.832 1.O000 0.0000
38 COMMUN1
1
3.855 0.0000 1.0000
67 PRIMETEC
i 143,365 0.0000 0.0000
76 LEITCH TE
1 1,031,559 0.0000 0.0000

DMU
#

67
0.0710
0.0000
0.0000
1.0000
0-0000

76 distance
0,031 1 0.0000
0,0000 1.4515
0.0000 0.0571
0.0000 1.5102
1.O000 1.5900

Minimum rnv $3.8 million

UPPER B9UND:
Market Capitalization of TUNDRA based on EFfIClENT PEERS
$46,523 thousand

Valuation Analysis for UNITEC INTERNATIONAL CONTROLS (SIC3600)

DMU

Score Market C. Distanc Efficient Peers

5
17
#
$(OOOs)
4 UNITEC 1b 0.4349
5.571 o.woo 0.0217 0.0472
5 GLOBAL T
1
13.222 1.8215 1.0000 0.0000
17 AASTRA 1
1
46.091
1.7705 0.0000 1.O000
42 FIFPII-PL1
1
2.404 0.0078 O-0000 O-0000
67 PRIMETEC
1 143.365 1.8598 0.0000 0.0000

42
0-9285
0.0000
0,0000
I,OOOO
0.0000

67 distance
0.0025 0.0000
0.0000 1.8215
0.0000 1.7705
O.OOOO 0.0078
1.O000 1.8598

Fifty-Plus is only simibr to Unitec in terms of distance, therefore it should have a market cap
near $2.4 miIlione

Market Capitalization of UNITEC based on EFFICIENT PEERS

UPPER BOUND
$5,057 thousand

TSI TEL (SIC 3600)

DMU
Scom
Market C.Distance Efficient Peem
#
$(OOOs)
42
60
56 TSI TELS)
0.21 119
12,193
ct.0000 0.7477 0.1 203
46 MARATHC
0.60450
623
0.0068 0,8016 0-1285
0.67621
9,700
0.1355 0.4601 0.3397
31 TRADE W
19 BCS TECk
0.67662
1,027
0.0423 0.9156 0.0417
63 STRATEG
0.53425
1,720
0.0162 0.6999 0.0642
71 SHELLCA!
0.08470
is.698
0.0235 0.8555 0.01 13
81 COMPREZ
0.31046
717
0.0752 0.9716 0.0131
42 FIFTY-Pt1
1.00000
2.404
0-0956 1.0000 0.0000
60PAClFlCE
1.00000
3,878
1.3503 0.0000 1.0000
307
4.3269 0.0000 0.0000
62 SNIF SEC1
1-00000

62 distance
0.1 320 0.0000
0.0699 0.0068
0.2002 0-1355
0.0428 0-0423
0.2359 0.0162
0.1332 0.0235
0.0153 0.0752
0.0000 0.0956
0.0000 1-3503
1-0000 1-3269
Ranking, according to the distances that are less than 1 (for TSI TEL):
TSI TELS)
a211191 ~ 1--=_-~-~?ZO:mo
MARATHC
0.60450
623
0.0068
STRATEG
0.53425
1,720
0.0162
SHELLCA!
0.08470
15,698
0.0235
BCS TECk
0.67662
1,027
0,0423
COMPREZ
0.31046
717
0.0752
FIFTY-PLL
1.O0000
2,404
0-0956
TRADE W
0.67621
9,700
0.1355
-%O

Similarity of the efficiency scores (difference between TSI TEL'S score ):

SHELLCA!
0.08470
15,698
0.0235
0.21 119
-F:121
9
3
:
; - - -.
- O-oooo.
TSI TELS)
COMPREZ
0.31 046
717
0.0752
STRATEG
0.53425
1.720
0.0162
MARATHC
0.60450
623
0.0068
TRADE Wi
0.67621
9,700
0.1355
BCS TECk
0.67662
1,027
0.0423
FIFTY-PLL
1.O0000
2,404
0.0956

Similarity of the Market Cap. (difference between TSI TEL'S market cap.):
MARATHC
0.60450
623
0-0068
COMPREZ
0.31 046
717
0-0752
BCS f Eck
0.67662
1.027
0-0423
STRATEG
0.53425
1,720
0.0162
FIFTY-PL1
1-OOOOO
2.404
0.0956
TRADE WI
0.67621
9,700
0.1355
TSl TELSY
O 2 1119 -- 12,19%z:
---2! O.0
SH ELLCA!
0.08470
15,698
0.0235
Range for TSI TEL'S market cap. based on peer group b $623,000 to $15,698,000

Market Capitalization of TSI TEL based on EFFICIENT PEERS

$2,304 thousand
VALUINGPRIVATECOMPANIES: A DEA APPROACH

185

WE X.L (SIC 3600)

DMU

Market C. Dktanc8 Enicient Peers

$(OOOS)
30
32
0,94369
ir,m o.ooooo 0.00879 0.00595
0.53445
2,426
0.00271
0.00322 0.04520
0-86634
14,946
0 . ~ 0 4 9 0.08840 0.49579
0.87321
329,339
0.05122
0.03590 0.15989
0.54465
7,179
1.67913
0.01 160 0,92084
0.00276 0-02100
0.00034
2,302
0.58296
1.00000 0-00000
1.95326
58s.x18
1.00000
0.00000 1-00000
1.95894
i1,832
1-00000
0.00000 0.00000
O.O33
2.404
1.00000
0.01 170 0.0891 0
19,928
0.01432
0-18038
0.1 1300 0.76171
23,569
1.32157
0.99488

Score

#
85 WE X-L. Hl
83 NEXTRON

77 PLAl NTRE
65 WI-LAN IN
3 INVERPO\
9 AVANTE T
30 RESEARC
32 TS TELEC
42 FlFlY-PLL
33 ADVANTE
45 KASTEN C

Distance
42 Formula
0.98526 0.00000
0,951 58 0.00271
0.41 586 0.57049
0.80420 0.05722
0.06760 1-67913
0.97626 0.00034
0.00000 1.95326
0,00000 1.95894
3 .O0000 0.00033
0.89922 0.01432
0.12529 1.32157

Ranking, according to the distances that are less than 1 (for WE XL):
WE X-L. Hl 0.94369
17,083
0.OOO
FIFW-PLL 1.OOOOO
2.404
0.00033
AVANTE T 0.58296
2,302
0.000%
NEXTRON 0.53445
2,426
0-00271
ADVANTE 0.1 8038
19,928
0.01432
WI-LAN IN 0.87321
329.339
0-05722
PLAINTRE 0.86634
f 4.946
0.51049
Similarity of the efkiency scores (difference betweeri WE X.L's score (0.94):
ADVANTE 0.1 8038
19,928
0.01432
N W R O N 0.53445
2,426
0.00271
AVANTE T 0.58296
2,302
0.00034
PLAINTRE 0.86634
14,946
0.51049
WI-LAN IN 0.87321
329,339
0.05722
WE X. L. Hl 0.94369
17,083
0.00000
FIFTY-PLL 1.O0000
2,404
0.00033
Similarity of the Market Cap. (dwerence between WE X.L's market cap. ($17,083,000):
AVANTE T 0.58296
2.302
0.00034
FlFTY-PLl. 1.00000
2,404
0.00033
2,426
0-00271
NEXTRON 0.53445
14.946
0.57049
0.86634
PLAINTRE
17,083
0.00000
WE X.L. Hf 0.94369
19,928
0.01432
ADVANTE 0.1 8038
329,339
0.05722
WI-LAN IN 0.87321
Range for WE X.L's macap. based on peer group is $2,302,000 and $329,339,000
WE X.L's real market cap of $11,083,000 b rigtit in this ange
Especially after ignoring the effect of Wi-lan, the range is betwee $2.3 and $20 million

Market Capitaliution of WE X.L based on EFFICIENT PEERS

$7,584 thousand
VALUING PRWATECOUPANIES:
A DEA APPROACH

186

Valuation Analysis for WI-LAN (SIC 3600)

DMU
Score
Market C.Distance Efficient Peem
Distance
#
$(OOOs)
30
32
42 Formula
65 WI-LAN IN 0.87321
329,339
0.00000 0-0359 0-1599 0.8042 0.0000
3 INVERPOI 0.54465
7.179
1.12221 010116 0.9208 0.0676 1.1222
9 AVANTE 1 0.58296
2.302
0-05000 0.0028 0.0210 0.9763 0-0500
30 RESEARC 1-00000 sas.308
1.60178 1.0000 0.0000 0.0000 1-6018
1i,832
1.35380 0.0000 1-0000 0-0000 1-3538
32 TS TELEC 1-00000
19,928
0.01463 0.01 17 0-0891 0-8992 0.0146
33 ADVANTE 0.1 8038
42 FIFTY-PLL 1.O0000
2.404
0.06519 0.0000 0.0000 1-0000 0.0652
45 KASTEN C 0.99488
2 3 . ~ 9 0.82904 0.1130 0.7617 0.1253 0.8290
77 PLAINTRE 0.86634
14,946
0.26639 0.0884 0.4958 0,4159 0.2664
83 NEXTRON 0.53445
2,426
0.03594 0-0032 0-0452 0.9516 0.0359
85 WE X.L. Hf 0.94369
17,083
0.05722 0.0088 0.060 0.9853 0.0572
Ranking, according to the distances that are les$ than 1 (for WI-UN):
W-LAN IN--- 0.87321 . 329.339 - -000000'
ADVANTE 0.1 8038
i9.928
0.01463
N M R O N 0.53445
2,426
0.03594
AVANTE T 0.58296
2.302
0.05000
WE X. L. Hi 0.94369
17.083
0.05722
FIFTY-PLL 1-00000
2,404
0.06519
PLAINmE 0.86634
14,946
0.26639
23,569
0.82904
KASTEN C 0.99488
Similarity of the efficiency scores (difference between W-UN'S score (0.873):
ADVANTE 0.1 8038
19,928
0.01463
r \ l M R O N 0.53445
2,426
0.03594
AVANTE T 0.58296
2.302
0.05000
PLAINTRE 0.86634
14.996
0.26639
W
U IN, 0.87321
329,339
EW*
WE X.L. Hl 0-94369
17.083
0.05722
KASTEN C 0.99488
23,569
0.82904
fIF-!Y-PLL ~.ooooo
2.404
O.519
Similarity of the Market Cap. (diffemnce between W U N b market cap. ($329,339,000):
AVANTE T 0.58296
2.302
0.05000
FIFP/-PLL 1.ooooo
2.404
0.06519
NEXTRON 0.53445
2,426
0.03594
PLAINTRE 0.86634
14,946
026639
WE X.L. Hf 0.94369
17.083
0.05722
ADVANTE 0.18038
19,928
0.01463
KASTEN C 0.99488
23.563
0.82904
WI-LAN IN'- 0.87321
329,339 --10.00606
Range for W I - ~ N ' Smarket cap. based'on peer gmup is $2,302,000 and $23,569.000
WI-LAN CONTROLS LTD real market cap of $329,339,000 is far above this range

Market Capitalization of Wi-LAN based on EFFICIENT PEERS

$24,838 thousand
-

VALUING PRIVATE
COMPANIES:
A DEA APPROACH

187

APPENDIX C
Efficient Companies-SIC 3 600
Market Value Analysis

Based on BCC-1
For Company Aastm (17)
Score

DMU
#
17 AASlRA
4
53
21
66
11

TE

Market Cap
$(OOOs)
1
46,091

UNITECINT 0.43488
SYSCAN IN' 0.657407
ClRCA ENTl 0.859468
CIRCUIT W 0.75253
DANBEL IN[ 0.755337

17
1 nonnalized
4.72E-02
0-052205
0-278
0.400605
0-621007
1.399018

5.571
7.260
6.601
289
5.798

0.033738
0.037316
O. 198711
0.286347
0-443888
1

REAL MARKET CAP of Aastra Tech

(as can be seen it is really similar to
Therefore, very small estimated

187.9442594
270.8941374
1311-689405
82.67916377
2573.473118
$4,427

46,091

some really small companies)

market cap

Based on BCC-I
For Company Research in Motion (30)
Score

DMU
#
30 RESEARCH

Market Cap
$(OOOs)
1
585,308

79 BALLARD P.
1
40 DIMENSION 0-110285

3,608.271

9 AVANTE TE 0.582956
83 NEXTRON C 0.534451

2.302

14 DIGITAL PR 0.884281

28.123

SIERRA W l F
CELL-LOC Il
WE X.L. HO1
INVERPOW
33 ADVANTEDI
58 XI TECHNt
28 NHC COMM
41 DYNAMIC D
39 CYGNAL TE
65 Wl-LAN INC
78 TRI-VISION
77 PLAINTREE
45 KASTEN Ch

32.735

61
36
85
3

0.996348
0.980304
0.94369
0.544648
0.180379
0.756899
0.670855
0-939374
0.589195
0.87321
0.742628
0.866344
0.994883

1.970
2.426

28.669
17.083
7.179
19.928
9.422
7.317
110.3~4
8.637
329.339
29.930
14.946
23.569

30
1 normalued
3.12E-16
3.26E-04
2.76E-03
3.22E-03

7.07E-16
7.39E-04
6-25E-03
7.30E-03

3.45E-03
5. i9E-03
5.58E-03
8.79E-03
1.16E-02
1.17E-02
1.17E-02
1.32E-O2
1-72E-02
0.024202
3.59E-02
0.085046
8.84E-02
0.1 12999
4.41E-01

7.82E-03
1-18E-02
1-26E-02
1-99E-02
2.63E-02
2.65E-02
2.65E-02
2-99E-02
3.90E-02
5.48E-02
8.14E-02
7 -93E-07
2.00E-01
2.56E-01
1.OOE+OO

Real Market Cap of Reseamh in Motion

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

i 89

Lower Bound for CELESTICA(68)

BCC-1, for public companies only and SIC code of 3600s
Lambda
68

Market Cap
$(OOOs)

Score

DMU
#

68 CELESTIC

2.850.867

10 SEMI-TEC
20 BELL CAN
73 C-MAC IN[
82 GSI LUMO
47 NEWBRIDl

0.214508
0.254057
0.817145
0.916721
1

10,026
1,371,120
291.T 27
129.626
8,125.481

TOTAL-

1 Nom
0.225377
O. 140231
5.88E-02
1-35E-02
4.70E-18
0.437908

0.514668
0.320229
0.134275
0-030828
1-07E-17
1

MCCALC
5160.04695
439072.5952
39090.97203
3996.146357
8.72095E-11
$487,320 (000s)

Lower Bound for COMMUNICATION SYSTEMS INTERNAT (#38)

BASED on BCC-I
DMU

W a r k a Cap
$(OOOs)

Score

38 COMMUN1

CIRCAEN'
CYGNAL 7
SYSCAN 11
EX1 TECHI
61 SIERRAW
62 SNlF SECI
72 TUNDRA 5
78 TRI-VISIOI

0.859468
0.589195
0.657407
0.756899
0.996348
1
0.98174
0-742628

21
39
53
58

38
1 NORM

3,855

MCCALC

Comm appars undewalued bard on its pm.

Lower Bound for BALLARD POWER SYSTEMS INC (79)

Based on BCC-l
Score

DML:

Market Cap
$(OOOs)

79 BALLARD

20 BELL CAN 0.254057

1
47 NEWBRIDl
-

3,608,271
1,371.120
8,125,481

0.450728
1
1.88E-17 4.17E-17
0.450728
1

1371119.948
3.38916E-1 0
$1,371,120

VALUING PRIVATE
COMPANIES:
A DEA APFROACH

190

Lower Bound for LElTCH TECHNOLOGY CORPORATlON (76)

Market Cap
$(OOOs)

Score

DMU
#

20 BELL CAN
25 MEMOTEC
28 NHC COM
35 CANADIAF
37 COM D W
39 CYGNAL T
50 SPECTRU
51 SR TELEC
54 TEKLOGIIC
57 WESCAM
58 XI TECHI
61 SIERRA*
68 CELESTIC
69 CORECO 1
70 GENNUM 1
72 TUNDRA I
73 C-MAC IN[
80 BLUESTAF
82 GSI LUMO
84 ROYAL LA

0.254057
0.649191
0.670855
1
0.863264
0.589195
0.78669
0.446954
0-809908
0-628467
0-756899
0.996348
1
0-983382
1
0.98174
0.817145
0-55479
0.916721
0.683747

TOTAL

76

0.409041
1.46E-02
3.33E-02
4.53E-17
0.800779
2.65E-02
7.75E-02
0.444511
4.49E-02
5.84E-02
4.43503
9.59E-02
3.97E-16
0.140155
1.21018
3.11E-02
0.56102
3.79E-02
0.966892
8.81 4 2
3-972509

1-41+O5
1.O3E-Ol
1.82E+01
3.68E-03
6.13E+Ol
8.38E-03
1.14E-17
4.87E-12
5-09E+04
2.02E-01
5.76+01
6.67E-03
7.54+02
1-95E-02
1.26+04
1-12E-01
2-75E+03
1-13E-02
1-47E-02
l.l3E+O3
1.05E+01
1.12E-03
7.90+02
2.41E-02
2.85E-10
9.99E-17
3.53E-02
1.58+03
3.05E-19
1.76E-13
9.45+02
7.836-03
4.1 1+O4
1.41E-01
4.24+02
9.54E-03
3.16+04
2.43E-01
l.l4E+O3
2.22E-02
$287,982 Lower bound
1
$t,O3l,SS9
(000s)
REAL

Lower Bound For Efficient Company CANADIAN MARCONI(35)

Based on BCC-I (for public companies only)
Score
DMU
#
theta
35 CANADIAE

Market Cap

YOOfW
426808.3836

CANADIAN MARCONI is a self evaluator

VALUINGPRIVATECOMPANIES:
A DEA APPROACH

191

Valuation Analysis for Efficient Company SNlF SECURIM INC (SIC3600)

Efficient
Market Cap Peer
$(OOOs)
62 Nom

Score

DMU
#

62 SNlF SECURi'

307

5.68E-03
2.41 4 2
4.42E-01
1-13 E 6 l
3.93E-02
8-57E-02
7.43E-O2
4.02E-16
1-33E-07
7.49E-02
8-61E-03
1.00E+00

8 NORON TECI19 BCS TECHNO

24 KING PRODUi
31 TRADEWlND
46 MARATHON F
49 SIMMONDS C
56 TSI TELSYS C
60 PAClFlC E-LIF
63 STRATEGIC L
71 SHELLCASE L
81 COMPRESSIC

-.-p.

MC Calc.

3.99E+01
2.47E+Ol
3.42E+03
1.OgE+O3
2.4SE+01
2.45E+02
9.06E+02
1S6E-12
2.28E+02
1.18E+03
6.17+00
$7,164 thouiand
Lower Sound
Undervalued

VALUINGPRIVATECOMPANIES:
A DEA APPROACH

192

For Efficient Company Primetech Elect. (67)

Based on BCC-I (for public companies only, SIC 3600s)
Score
theta

DMU
#
67 PRIMETECH ELECTRO

Market Cap Efficient

$(OOOs)
Peers
1

143,365

Normalized Market Cap Calculation

1

53 SYSCAN INTERNATIOh
4 UNITEC INTERNATIONJ
21 ClRCA ENTERPRISES I
72 TUNDRA SEMICONDUC
50 SPECTRUM SIGNAL PF
37 COM DEV INTERNAT10
25 MEMOTEC COMMUNIC
6 LG TECHNOLOGIES GF
1 HAMMONO MANUFACT
34 CALIAN TECHNOLOGY
51 SR TELECOM INC
74 ELECTROHOME LIMITE
57 WESCAM INC
80 BLUESTAR BATTERY S
54 TEKLOG lX INTERNATIC

REAL CAP

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

193

Lower Bound for PAClFlC E-LINK CORPORATION (60)

Based on BCC-1
Market Cap

Score

DMU
#

60 PAClFlC E

12 A W A TEC
13 TIR SYSTE
18 ABL CANA
19 BCS TECF

0.847256
0.655687
0.399144
0.676619
0.639866
0.815085
0.471945
0.0664
0.676206
1
0.604497
0.766307
0.525957
0-211188
1
0.53425
0.562167
0.0847
0.310462

22
23
24
26
31
32
46
49
55
56
62
63
64
71
81

DBA TELE
EIGER TE4
KING PRO
MlCROCEl
TRADE WI
TS TELEC
MARATHC
SIMMOND
TRIPLE CF
TSI TELSY
SNlF SEC[
STRATEGI
SYMPLEX
SHELLCA:
COMPREE

S(0OOs)

Lambda
60

TOTAL:

3.053191

$86,216 (000s)
3,878 (000s)

1
REAL

VALUINGPRIVATE
COMPAN~ES:
A DEA APPROACH

194

For Company SONY (15)

Score

DMU
#
15 SONY CORI

Market Cap Efficient Peers

$(OOOS)
15
1 52,935,385
1

1 HAMMOND 0.756773
2 TURBO GEt
1
3 INVERPOW 0.544648
4 UNITECINT 0.43488

18,426

80,615

7,179

5.571

Sony is a Self Evaluator

For Company Cinram (16)
16 CINRAM IN7
1
1 HAMMOND 0.756773
2 TURBO GEt
1
3 INVERPOW 0.544648

1,075,691
18.426
80.61 5

7.179

1
O
O
O

Cinram is a self evaluator

For Company MlTEL (27)
27 MlTEL COR

1,257,378

15 SONY CORI
1
73 C-MAC lNDl 0.817145

52.935.385

1.38E-15
291.127 0.159901
0.1 59901

SELF evaluator

MlTEL Corp is a

Based on 8CC-l
For Company Nortel(29)
Score

DMU
#

29 NORTEL NE

Market Cap
$(OOOs)
1 50,495,406

80 BLUESTAR
0.55479
81 COMPRESI 0.310462
82 GSI LUMON 0.916721

44.404

717
129.626

NORTEL is a Self Evaluator

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

195

Lower Bound for NEWBRIDGE NETWORKS CORPORATION (47)

based on BCC4
DMU
Score
#
47 NEWBRIDGE
15 SONY CORPC
29 NORTEL NET

MarketCap
$(OOOs)
1
1

52,935,385

50,495,406

8.125.481

47
1
4.25E-16 5.58E-01
2.96E+07
3.36E-16 4.42E-01
2.23E+07
7.61E-16 1.OOE+00 $51,858,075
51,858,075

Lower Bound for efficient JEMTEC INC (59)

Based on BCC-I
DMU
Score
Market Cap
#
$(OOOs)
59
59 JEMTEC INC
1
330
1
52 STRATEGIC T 0.983221
3.943 0.1 13323
0.1 13323

3,943,000 lower bound

undervalued

Lower Bound for INTERNATIONAL DATACASTING CORP

based on BCC-1
DMU
#

Score

Market Cap
$(OOOs)
1

44 INTERNATIOb

3.893

44

1 norm

rnc calc

5 GLOBAL THE1
1 13221-8604 9.54E-18 1 -696-16 2.2342E-12
11 OANBEL INDL 0.755337
5797.576
0-0135 0.239124 1386.34114
28 NHC COMMUI 0.670855 7317.08222 0.027856 0.493411 3610.32741
66 CIRCUIT WOF 0.75253 288.73734
0.0151 0.267465 77.2271112
$5,074 (000s)
5.65E-02 1.00E+00

Lower Bound for GENNUM CORPORATION (70)

Based on BCC-I
DMU
#
70 GENNUM COI

Market Cap
$(OOOs)

Score

70

576.578

16 CINRAM INTE
1
27 MITEL CORPC
1
47 NEWBRIDGE
1
69 CORECO INC 0.983382
84 ROYAL LASEf 0.683747

1,075,691

220E-17
5.55E-17
1.20E-16
1.19E-02
0.102096
1.f 4E-01

1,257,378
8,125.481

44,923
51,Si

1.SE-16

4.87E-16
1.OSE-15
1.ME-01
8.96E-01
t.00+00

VALUINGPRIVATE COMPANIES:
A DEA APPROACH

2.08E-1 O
6.12E-10
8.5SE-09
4.69+03
4.62+04
$50,868 fower bound

196

FIFTY-PLUS.NET INTERNATIONAL INC (42) (SIC 3600s)

DMU

Market Gag Eftcient Peers

Score

(lambda)

FIFTY-PLUS.NET Il

Normalizc Market Caps

lr

NORON TECH LTD 6.53~-02'

GEMSTAR COMMC 0332398

WE X.L. HOLDING: 0.94369
AVANTE TECHNOI 0.582956
COMPRESSION & E 0.3 10462
BCS TECHNOLOG1 0.6766 19
A W A TECHNOLO 0.847256
NEXTRON CORPOF 0.53445 1
LAMBUS ENTERPR 0.709557
UNITEC INTERNA1 0.43488
NHC COMMUNICA 0.670855
SYMPLEX COMMU 0.562 167
ADVANTEDGE iNT 0.1 80379
TIR SYSTEMS LTD 0.655687

MARATHON FOOD 0.604497

DMENSIONS WES' 0.1 10285
SHELLCASE LTD 8.47E-02
CORECO iNC
0.983382
DBA TELECOM CO 0.639866

SENSE TECHNOLO 0374065

ROYAL LASER TEC 0.683747
WI-LAN CNC

0.8732 1

KING PRODUCTSn 0.47 I945

CELL-LOC lNC
0.980304
EXI TECHNOLOGIE 0.756899
TSI TELSYS CORK 0.2 1 1 188
TRIPLE CROWN EL 0.525957
STRATEGlC VISTA 0.53425
TRI-VISION MTERl0.742628
TRADE WiND COM 0.676206
EIGER TECHNOLO 0.8 15085
SYSCAN INTERNA' 0.657407
STRATEGIC TECHP 0.98322 1
PLAMTREE SYSTE 0.8663U
ABL CANADA INC 0.399 144
DANBEL iNDUSTR 0.755337
CIRCUIT WORLD C 0.75253
SIMMONDS CAPIT, 0.766307
KASTEN CHASE AI 0.994883
W R P O W E R COh 0.544648

Real
-

%2,404 (000s)

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

197

FOR TS TELECOM(32) (SIC3600)

DMU

Score

85

-ricet

Car Efficient Peers

W E XL. H
c 0.94369

64 SYMPLEX 0.562 167

9 AVANTE 1 0.582956
12 A W A TEC 0.847256
83 NEXTRON 0.53445 1
2 1 CIRCA EIV 0.859468
33 ADVANTE O. 1 80379
72 TUNDRA E 0.98 174
13 TIR SYSTE 0.655687
78 TRI-VIS101 0.742628
22 DBA TELE 0.639866
65 WI-LAN IE 0.8732 1
23 EIGER T E C 0.8 15085
57 WESCAM : 0.628467
55 TRIPLE CE 0.525957
6 1 SIERRA W 0.996348
74 ELECTROI 0.67333
1 HAMMON: 0.756773
77 PLAINTJE 0.866344
18 ABL CANJ 0.399 144
34 CALIAN T: 0.800 142
39 CYGNAL 7 0.589 195
6 LG TECHE; 0.586585
45 KASTEN C 0.994883
25 MEMOTEC 0.649 19 1
50 SPECTRUE 0.78669
14 DIGITAL P 0.88428 1
3 JMERPOK 0.544648

FOR CAMCO LNC (7)

6 LG TECHh; 0.586585
23 EIGER T E C 0.8 15085
82 GSI LUMO 0.9 16721
34 CALIAN T: 0.800 152
74 ELECTROI 0.67333
54 TEKLOGD 0.809908
37 COM DEV 0.863264
80 BLUESTAi 0.55479
5 1 SR TELECr 0.446954
26 MICROCE1 0.0664
73 C-MAC IM 0.8 17145
1 HAMMON- 0-756773
1O SEMI-TECI0 3 14508

VALUINGPR~VATE
COMPANIES:
A DEA APPROACH

85.2 18
45.239
1295.904
10 14.082
437.190
9377.226
10604.658
1097.297
7845.925
28902.052
32707.433
2083.925
396 1.346
$100,460 REAL MC

$100,000

198

APPENDIX D
Analysis of Efficient and Inefficient
Companies SIC 3700

BUDD CANADA INC (3) (SIC 3700s)

Based on BCC-I for public companies
Swm

DMU
#

~arket
PRInUf Capiblbiblon
sic (ooor)1998
116,772

Efficient Peer

3
1.00000

Nom
1 GENERALM(
1
2 AE.VENTUR 0.8
4 DECOMA IN1 0.58
5 DEVTEK COF 0-67
6 LINAMARC0
1
7 MAGNAINTE
1

8 MERCH PERI 0.65

9 NORTHSIDE 0.9
10 ROSE CORPI 0.71
11 SPECTRA PF
1
12 STACKPOLE 0.63
13 TESMA INTEl
1
14 WESCAST IN
1
15 TRAlLMOBlLt 0.97

16 GLENDALEIt
17 AVCORP IN0
18 HEROUX INC
19 MAGELLAN P
20 SERENA REE
21 TEMPLAR RE
22 BOMBARDlEl

0.52
0.5
0.59
0.63
0.73
1
1

Sum:

3.75

MC Calc
0-0000
0.6006
68469.6354
23467,0717
0.0000
0.0000
0.0000
339.8526
0.0000
0.0000
8571.3271
61425.8952
0.0000
1i75.8158
5549.4522
1604.6554
4541.1077
68970.6190
0.0000
0.0000
0.0000
$24,116 Caltulated
1
$116,772 Real

This is a lower bound for the efficient Company, BUDD Canada, based on the
inefficient companies k i n g compand to it

Thus it seems that BUDD CANADA b rather undewaiued

- --

VALUING
PRIVATECOMPANIES:
A DEA APPROACH

200

LlNAMAR CORPORATION (6) (SIC 3700s)

Based on BCC4 for public companies
Score

DMU

HWIW
EfkientPeer
PWYI Cpiblizaa
6

SIC

6 LINAMAR CC
3 BUDD CANAI:
1 GENERAL Mt
2 AE-VENTUR
4 DECOMA lNT
5 OEVTEK COF
7 MAGNA INTE
8 MERCH PERl
9 NORTHSIDE
10 ROSE CORPt
11 SPECTRA PF
12 STACKPOLE
13 TESMA INTEf
14 WESCAST IN
15TRAkMOBlLf
16 GLENDALE lt
17 AVCORP IND
18 HEROUX INC
19 MAGELLANP
20 SERENA RES
21 TEMPLAR RE
22 BOMBARDIEI

1
1
1
0.79684
0.57823
0.66674
1
0.65439
0.89812
0.70626
1
0.63128
0.99948
1
0.97208
0.51504
0.50231
0-5871
0.62997
0.72956
1
1

3714

(ooos) 11,?9%5it

N o m MC Calc.
O
0.0000
3714
iis.nz 0.0000
0.0000
O
0.0000
3711 86,392,582
0.0000
3714
627
0.0000
O
3714
560.656
0.4793 0.386 216331-3983
3714
99.709
O. 1181 0.095
9481-6534
0.0000
O
0.0000
3714 7,374,103
0,0000
3714
2.869
0.0000
O
256.8636
37:4
14.056
0.0227 0.018
3714
25,150
0.0000
O
0-0000
0.0000
3714
525,280
0.0000
O
0.0000
O
0-0000
3714
131,497
3714
570.124
0.4697 0.378 215595-31O3
0.0000
3714
574,300 0.0000 7E-17
0-0000
3715
14.342
0.0000
0
0.0000
O
0.0000
3716
47,948
0-0000
3n8
#,no 0.0000
O
0.0000
O
0.0000
3728
62,462
60868.6475
3728
496,389
0.1 523 0.123
0.0000
3731
m 0.0000
O
0.0000
3732
4,576
0.0000 ".;
O
3743 7 . ~ 2 . 0 4 9
O
0.0000
Sum:
1.242f 4
1
$502,534 Calculated
$1,792,517 Real

This is a lower bound for the efficient Company, LlNAMAR CORP;, brsed on the
inefficient companies k i n g compared ta it

VALUING

PRIVATECOMPANIES:
A DEA APPROACH

20 i

AVCORP INDUSTRIES INC (SIC 3700)

Based on BCC-I for public company
DMU
#

Scon,

hrkut

PRIM ~apitai:
Dbtince

Efficient Peem
3
11

21 Distance

sic (ooos)1998
# AVCORP 0.502312 3728 45.730 0.0000
627 0.4565
2 AE. VEN' 0.796841 3714
3 BUDD CA
1 3714 116,772 1.1322
1 3714 525.280 0.5643
# SPECTR/
# TEMPLAF
1 3732 4,576 0.4889

0.131
0.004
1.000
0.000
0.000

0.415
0.015
0,000
1.000
0.000

0-453
0-982
0.000
0-000
1-000

0.000
0-456
1.132
0.564
0.489

AVCORP only has 4 other peers, 3 of which are efficient

As can be seen from the distances, AVCORP is closest to AE Venture's (d=0.456)
Similarity of the efficiency scores (difference between AVCORP score (O.=):
A Ventur -0.294529
Templar -0.497688
Spectra -0.497688
AVCORP efficiency score is not close to any of the others, the closest is AE Venture at 0.797
Similarity of the Market Cap. (difference between AVCORP market cap. ($45,730,000):
AE Ventur
45.1O3
Templar
41,154
Spectra
479,550
AVCORP market cap is not close to any of the othersRange for AVCORP market cap. based on peer group is between $627,000 and $525,280,000
AVCORP real market cap- of $45,730,000 falls into such a great range with ease

Market Capitalization of AVCORP based-onEFFICIENT PEERS

[(OA 31'1 16772) + (0.415'525280) + (0.45*4576)

is $235,632 million

--

VALUING
PRIVATE
COMPANIES:
A DEA APPROACH

202

APPENDIX E
Analysis of Efficient and Inefficient
Companies SIC 2000

Valuation Analysis for A. L. VAN HOUlTE LTEE (SIC2000)

DMU
#

Score

IY-

PRIMARICapiblitlti
SIC
(Os)1998

43.0000 Nom.

MC Calc.

7 SAPUTq 0.787
15 RlDLEL 0.9894
10 LASSO 0.9632
31 VINCOf 0.7498
25 SLEEM 0.8323
35 CLEAR 0.9535
28 ANDRE 0.7103
27 UNIBRC 0.9878
44 ARCTIC 0.6402
22 BIG RO 0.6985
23 BRICK 0.7797
29 MAGNC 0.6915
8 DELICI4 0.5398
39 URBAN 0.8089
11 GLOBA 0.5749
20 AFRICF 0.8332
34 ARROV 0.8705
32 HlGHW 0.71O3
26 TREE E 0.8365
1
24 FOSTE
1
17 CANAD

LOWER BOUND

VAUJINGPRIYATE
COMPANIES:
A DEA APPROACH

204

Valuation Analysis of AMT Fine Food (SIC2000)

DMU
#

4 AMT FINE FI

Efficient
capiti~h;Peer
(ooor)is! AMT

nom.

mc calc.

ARROWHEP 0.8705
TML FOODS 0.2571
MAGNOTTA 0.6915
BETA BRAN 0.3039
BRICK BREC 0.7797
LEADING BF 0.4515
UNIBROUE 1 0-9878
39 UR8AN JUlC 0.8089
14 AGRO PACll 0.7703
42 LEF MCLEAI 0.7971
3 SEPP'S GO1 0.5941
25 SLEEMAN B 0,8323
22 BIG ROCK B 0-6985
19 SMALL FRY 0.7801
28 ANDRES WI 0.7103
32 HIGHWOOD 0.7103
24 FOSTERS E
1
12 DOVER INDI
1
1
9 SUN-RYPE F

34
6
29
18
23
37
27

Score

0.082
0.078
0,078
0.078
0.077
0.077
0.073
0.071
0.071
0-067
0.065
0-057
0-056
0.029
0.024
0.015
0.000
0.000
0.000
1-000

579
$139
$1,037
$3,732
$1,192
$659
$2.478
3917
S774
$228
91,778
$7,230
$1,098
$786
$1,917
$36
$0
$0
$0
$24,079 thouiand
lower bound

VALUING PRNATECOMPANIES: A DEA APPROACH

205

Valuation Analysis for AGRO & LEADING SIC2000

Score
14 AGRO PAC 0.77'034
37 LEADING E0.45355
2 SCHNEIDE 1
4 AMT FINE f 1
5 FLETCHER 1

Efficient Peer
SCHNEICAMT FIN FLETCHI Diaance
10,874 6.91E-02 0.85666 7.43E-02
0.0000 0.0094
8.~10
1-23E-03 0,92561 7.32E-02 0.0094 0.0000
1.6060 1-8597
139.628
t

MC
S(000s)

11,173

5,936

1
1

0.0308
1.5956

0.0109
1,7158

Agro is similar to Leading and AMT

Market Value range: $8,570,000 to $11,173,OOO
Upper Bound
$19,654 thousand
Leading is similar to Agro and AMT
Market Value: $10,874,000 to $11,173,000

Upper Bound
$ 10,947 thousand

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

206

APPENDIX F
Analysis Utilizing Al1
452 Companies

HAMMOND compared to al1 452 Companies

DMU
#

Score

Market Value SIC

Distance

Enicient Pwrs
73
322

0.0000
18,426,080
0.667 0.187
3612
175 HAMMOt 0.51431
0.0268
0.764 0.064
45,927,284
2331
83 CHATEA 0.82904
0.751 0-088
66,664,560
3523 0.0248
59 BUHLER 0.87389
0.0366
0.565 0.341
2834
296 PATHEO 0.52526 142,814,041
OS43 0.156
0.0463
17,824,705
2034
226 LASSON 0.63848
0.440 0.547
2092 0.1897
3,380,635
231 LEFMCL 0.58608
0.383 0.305
141 EXCO TE 0-73515 138,405,738
3544
0.1483
0.4791
0.249 0.732
5,177,589
2875
99 CONSOL 0.44399
0.245 0-675
2821
0.4206
302 PFB COF 0.94048 24,080,255
0.226 0.707
3571
0,4685
284 NTS COB 0.60877 48,091,341
0.5248
0.164 0.708
3716
163 GLENDA 0.31597 47,948,068
71,058,787
0.4547
0.158 0.625
2851
360 SIC0 INC 0.5653
0.6923
0.147 0.831
13,310,001
2084
242 MAGNO1 0.39093
80,755,275
3081 0.5509
0.1 10 0.658
210 IPL INC 0.69724
0,075 0.899
3661
0.8635
3 AASTRA 0.69365 46,090,505
0.7992
0.054 0.838
3086
308 POLYAIF 0.41828 27,260,000
0.029 0.944
0.9865
7,109,353
2621
93 CML IND 0.50334
0.013 0.937
0.9949
3585
261 MOBILE90,52422 33,407,325
1.000 0.000
9,240,498
3949
0.1564
73 CARBITE
1
0.000 1.000
3841
1.1160
322 PROMA1
1 10,374,300
0.000 0.000
1-3514
2522
1 87,643,098
372 SMED IN
0.000 0.000
1-3404
3714
443 VVESCAS
1 574,299,775
$57,473,975 Upper bound h e d on Enicient Pwrs
Initially, Hammond was found to have 21 peers

RANKED BY DISTANCETO HAMMONO

HAMMOl 0.51431
18,426,080
3612
BUHLER 0-87389 66,664,560
3523
CHATEA 0.82904 45,927,284
2331
PATHEO 0.52526 142,814,041
2834
LASSON 0.63848
17,824,705
2034
3544
EXCO TE 0.7351 5 138,405,738
CARBITE
1
9,240,498
3949
LEF MCL 0.58608
3,380,635
2092
PFB COF 0.94048 24,080,255
2821
2851
SIC0 INC 0.5653 71,058,787
NTS COB 0.60877
48,091,341
3571
CONSOL 0.44399
5,177,589
2875
GLENDA 0.31597 47,948,068
3716
IPL INC 0.69724
80,755,275
3081

VALUING
PRIVATE
COMPANIES:
A DEA APPROACH

208

MAGNO1
POLYAlF
AASTRA
CML IND
MOBILE

0.39093
0.41828
0.69365
0.50334
0.52422

13.310,001
27,260,000
46,090,505
7.109,353
33,407,325

RANKED BY EFFlClENCY
GLENDA 0.31597 47,948,068
MAGNO1 0.39093 13.310,001
POLYAlF 0.41828 27,260,000
CONSOL 0.44399
5,177,589
CML IND 0.50334
7,109,353
HAMMOt 0.5 1431 18,426,080
MOBILE 0.52422 33,407,325
PATHEO 0.52526 142,814,041
SIC0 INC 0.5653
71,058,787
LEF MCL 0.58608
3,380,635
NTS COIi 0.60877 48,091,341
LASSON 0.63848
17,824,705
AASTRA 0.69365
46,090,505
IPL INC 0.69724 80,755,275
EXCO TE 0.73515 138,405,738
CHATEA 0.82904 45,927,284
CHATEA 0.87389 66,664,560
CHATEA 0.94048 24,080,255
CHATEA
1
9,240,498
RANKED BY MARKET CAPITALKATION
LEF MCL 0.58608
CONSOL 0.44399
CML IND 0.50334
CARBITE
1
MAGNO1 0.39093
LASSON 0.63848
HAMMOt 0.51431
PFB COF 0.94048
POLYAlF 0.41828
MOBILE 0.52422
CHATEA 0.82904
AASTRA 0.69365
GLENDA 0.31597
NTS COB 0.60877
BUHLER 0.87389
SIC0 INC 0.5653
IPL INC 0.69724
=CO TE 0.73515
PATHEO 0.52526

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

209

Cornpared to Al1452 Cornpanies

DMU
Score SIC
443 WESCAST IN01
1
33 ATS AUTOMATIC 0.7513
0.746
66 CANADIAN MARC 0.8897
91 C-MAC INDUSTR 0-5323
402 TESMA INTERNA 0.8329
127 DORE1 INDUSTE 0.6254
114 DECOMA INTERt 0.5014
344 SAMUEL MANU-1 0.5054
251 MDC CORPORAl 0.3267
410 TOROMONT INDi 0.5549
204 INTERTAPE POL' 0-4225
239 MAGELLAN AER( 0.5323
447 VVlNPAK LTD
0.6517
432 VELAN INC
0-7829
124 DOMCO TARKET 0.7649
67 CANAM MANAC ( 0.4373
416 TRITECH PRECIZ 0.4673
63 CANADA BREAD 0.8189
1 A- L. VAN HOUT7 0.9167
32 AT PLASTICS IN( 0,4201
136 ENERFLM SYSl 0.8131
232 LEITCH TECHNO 0.9981
406 TIMBERWEST FC 0.761
294 PAPERBOARD IPI 0.3156
0.5976
334 RIOLEY INC
433 VENTRA GROUP 0.4732
141 EXCO TECHNOLl 0.7351
0.5274
337 ROLLAND INC
345 SAPUTO GROUP 0.5408
357 SHAW INDUSTRI 0.8167
226 LASSONDE INDL 0.6385
116 D M E K CORPO 0.4756
210 IPL INC
0.6972
12 AINSWORTH LUF 0.2531
159 GENNUM CORPC 0.9264
439 WAJAX LlMlTED 0.4234
382 STACKPOLE L M 0.6015
253 MERCER INTERP 0.5056
318 PRIMETECH ELE 0.943
59 BUHLER INDUST 0.8739
144 FANTOM TECHN 0.7381
435 VINCOR INTERU 0.5921
324 PRUDEMIAL STI 0.9378
360 SIC0 INC
0.5653
0.4545
60 CAE INC
238 MAAX INC
0.5946
162 GILDAN ACTIVRi 0.4694
83 CHATEAUSTORI 0.829

300 PERKINS PAPER

MC

443
1NORM

MCCALC

62 CAMCO INC

367 SLATER STEEL 1

335 RIVERSIDE FORE
281 NORWALL GROL
70 CANGENE CORP
175 HAMMOND MAN1
296 PATHEON INC
302 PFB CORPORATi
283 NQL DRlLLlNG Tc
346 SCHNEIDER COF
173 HALEY INDUSTR
285 NU-GR0 CORPO
308 POLYAIR INTER I
284 IUTS COMPUTER
257 MILLTRONICS L I
243 MALETTE QUEBE
163 GLENDALE INTE
261 MOBILE CLIMAT
369 SMALL FRY SNAI
56 BRAMPTON BRIC
347 SC1 INCOME TRL
1 96 1NTERNATlONAL
242 MAGNOlTA WlNl
209 IPEC LTD
58 BUDD CANADA 11
368 SLEEMAN BREW
93 CML INDUSTRIE!
99 CONSOLIOATED
389 SUN-RYPE PROC
421 TUNDRA SEMICC
3 AASTRA TECHN(
309 POLYCORP INC
165 GLOBAL THERMi
131 EAGLE PRECISIC
21 9 KELMAN TECHNI
135 ENERCHEM INTE
106 CPI PLASTICS GI
89 CIRCUIT WORLD
384 STELLA-JONES 1
167 GROUPE BOCEN
1 O7 CRS ROBOTICS 1
231 LEF MCLEAN BR
88 CIRCA ENTERPF;
100 CONSUMERS PA
120 DlVERSlFlED lm
430 UPLAND GLOBAl
289 OTATCO INC
47 BIOCHEM PHARI
234 LINAMAR CORPC

0.6324
0.6055
0.432
0.4815
0.9171
0.5143
0.5253
0.9405
0.7518
0.5659
0.5094
0.5592
0.4183
0.6088

0.967
0.5702
0.316
0.5242

0.4344
o.ni9
0.9085
0.4766
0.3909
0.5671
0-8532
0.5592
0.5033
0.444
0.6215
0.8818
0.6936
0.7287
0.8089
0.5777
0.6784
0.6288
0.4861
0.5216
0.6417
0.5427
0.55
0.5861
0.6802
0.2496
0.9812
0.7845
0.7962
1
1

Real Market Cap

VALUING PRIVATE
COMPANIES: A DEA APPROACH

$408,330 thousand
574,300 thousand

211

APPENDIX G
SIC 2800

DMU na1Score MC

min

max

INEFFICIENT COMPANIES
50 ACETUi 0.486
$71,059 one peer
15 ALTARE 0.858
$2,931
$57,159
1 AMR TE 0.697
$24,080 $153,632
5 AT P M 0.516
$24,080 $974,676
17 AXCAN 0.954
$24,080 $974,676
39 BESTAF 0.754
$3,725
$44,269
18 BIOTEC 0.554
$530
$32,776
51 CANHO10.385
$6,577 one peer
44 CANTO10.902
$3,725
$6,577
45 CCL INC 0.877
$767,740 one peer
46 CEAPRC 0.991
$936
$2,996
61 CHEMBt 0.752
$2,846
$3,207
47 CONS010.515
$2,846
$10,546
57 CONS010.629
$6,577 one peer
34 CV TEC10.778
$530
$32,776
35 DIAGNC 0.630
$2,846 $140,743
66 DIVERS:0.748
$2,846 one peer
12 FYTOKE 0.883
$936
$10,546
36 GENOM 0.823
$2,846 one peer
20 HAEMA( 0.474
$19,046
$77,292
58 HEDLnt 0.753
$530
$32,776
21 HILLES10.81 1
$2,846 one peer
37 IBEX TE 0.871
$2,931
$57,159
38 IMI INTE 0.960
$530
$32,776
22 INEX Pk 0.937
$2,931 $140,743
62 KRYSTP 0.745
$530
$32.776
23 LABOPF 0.886
$1,488 $140,743
4 LITHOS 0.489
$530
$32,776
7 MAGNIF 0.821
$3,725
$78,451
13 MARINE 0.621
$2.931
$77,292
- -.
59 NAPIER 0.880
$32,776
$530 $140,743
63 NORDE 0.727
$530
$936
$12,657
8 NOVA C 0.679 $1,838,447 $4,O81,7iO one peer
60 NU-GRC 0.866
$41,449
$24,080
$71,059
25 ORALIF10.985
$2,931
$6,895
$57,159
26 PATHEC 0.684
S142,814
$24,080 one peer
27 PHARMI 0.177
$4,651
$2,846 one peer
14 PROME 0.591
$78,451
$20,940 $140,743
29 SALlX P 0.861
$14,803
$530
$32,776
41 SIGNAL10.766
$17,484
$2,931
$57,159
42 STRES5 0.825
$44,932
$140,743 one peer
43 SYNSOF 0.489
$77,292
$2,931 $140,743
$107,756
30 TECHNI 0.871
$24,080 $153,632
31 THERAI 0.903
$57,159
$2,931 $140,743
32 UPLANC 0.868
$1 1,221
$3,725
$13,328

uppew boum (real-min)

Ireal

$80,431
$51,195
$15,545
$140,881
$72,892
$259,036
$87,578
$9,949

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

(mal-max) (mal-upper)
ireal
/mal

--

73.31 %
-79.40%
-2.52%
96.42%
-121-42% -245.85%
83.24Oh
-226.92%-213.23% #VALUE!
-213.54%
96-21%
-82.09%
42.57%
-140.39%
77.65%
-146.23%
-53.22%
94.87%
66.80%
-18.78%
28134 within range
33/45 upper
than mal

213

Real
2 BORDE1
3 DUPON'
9 PFB CO
10 POLY-P,
1IACTA NI
16 ANGlOT
19 BlOVAlL
24 NATRAC
28 QLT PH(
33 MOCHE
40 CANGE1
48 POLYME
49 SIC0 IN1
52 LANDO\
53 METHAI
54 AGRIUN
55 POTASF
56 TERRA I
64 AQUA-P
65 DIVERS
67 SHAW If

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

Lowe?Boun real-lowerlreal
$3,725
$17,264 -363-51%
$4,081,770 $1,541,128
62.24%
$24,080
$102,332 -324M%
$1,251 self evaluatoi WALUE!
$6,577
$2,622
60.14%
$140,743
$36,862
73-81%
$1,404,647 self evaluatoi #VALUE!
$8,108 self evaluatoi #VALUE!
$974,676
$139,816
8566%
$4,708,260
$146,983
9688%
$221,802
$66,915
69.83%
$3,207
$3,600
-12.27%
$71,059
$73,206
-3.02%
$2,846
$9,042 -21 7-67%
51,370,334 self evaluatoi #VALUE!
$1,541,000 self evaluatoi #VALUE!
$5,281,848 self evaluatoi #VALUE!
$716,918 self evaluatoi #VALUE!
$6,895
$25,847 -274,87%
$13,328 self evaluatoi #VALUE!
$767,740
$539,758
29.70%
1/43 lower than mal

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

VALUINGPRIVATE
COMPANIES:
A DEA APPROACH

214