Professional Documents
Culture Documents
Abstract: The purpose of this paper is to develop new financial models of BOT projects to
replace the SLR (self-liquidation ratio) since SLR cannot be used to analyze BOT projects. In
this paper, we use the financial cash flow concept and mathematical analytical method to
develop the PCCR (private construction cost ratio), GCCR (government construction cost
ratio), ongoing royalty, and GFRR (government finance recovery ratio) for the BOT finance
policy decision model. Also, we explore the relationships among the PCCR, GCCR, ongoing
royalty, and GFRR models.
1.INTRODUCTION
The BOT (Build, Operate, and Transfer) approach has been widely employed to implement
infrastructure projects by many developed and developing countries around the world (Walker
and Smith, 1995). The good financing project is one of the critical factors in BOT contract
delivery system. Only with good load bank can a BOT project be carried out (Chang and
Chen, 2001). In the process of the financing planning, there are many assessment methods
such as NPV (net present value), BCA (benefit cost analysis), IRR (internal rate of return),
and PBY (pay back year) can be used to evaluate the financing project of the BOT projects
(Finnerty, 1996). To meet this need, in 1999, the Ministry of Traffic and Communication
(MOTC) in Taiwan has developed a model for financing and evaluation of bidders’ proposal.
The SLR (self-liquidation ratio) index has been applied to assess financing project of BOT
projects based on the Act for Facilitation of Private Participation in Infrastructure Projects
(AFPPIP).
Prior studies show that the B/C analysis has been widely used to evaluate the economical or
financial effectiveness of an investment project (Daniel, 2002; Hanspeter, 1973; Asensio and
Roca, 2001; Xing and Wu, 2000). SLR and B/C are indifferent in their definitions. However,
can SLR be used to evaluate the financial project of the BOT project? This issue is seldom to
explore from the past studies (Chang & Chen, 2001; Lu, 2000; Wu, 2002). The purpose of this
Journal of the Eastern Asia Society for Transportation Studies, Vol.5, October, 2003
2921
paper is to introduce a new finance model in order to analyze the financial project of the BOT
project.
The paper is organized as follows: Section I introduces motivation, purpose of this research;
Section II details problem statement; Section III develops the new finance model; and Section
IV concludes this paper.
The financial cash flow of a BOT project is different from that of a non-BOT project.
Government can only carry out the non-BOT project. Figure 1 shows the investment fund
source of the non-BOT project including public budget, A or B bonds. The construction cost
fund source comes from public budget, A or B bonds. They are used to build facilities such as
high-speed rail, highway, or other facilities. The government receives the fees, tolls, or
operation revenue from users or passengers and use them to pay construction cost during the
construction period, operation cost, maintained cost, depreciation cost, and any other costs
during the operating period, also to pay back the A and B bond. This structure of net cash flow
mentioned above is conception of the self-financing project proposed by Musgrave (1959).
Concept of the self-financing project is shown in Figure 1.
user fee
infrastructure project passengers
government
A Bond B Bond
Now, considering that the government adopts the BOT approach to implement infrastructure
project. The original non-BOT project of Figure 1 becomes a private-public joint venture
project. It is because the private sectors’ investment costs joint during the construction and
operation period. Thus, the structure of net cash flow will be changed and the concept is
depicted in Figure 2.
Journal of the Eastern Asia Society for Transportation Studies, Vol.5, October, 2003
2922
operating income
public sector's investment cost
credtiors stockholders
From the concept of the Fig. 1 and Fig. 2, it is easy to compare the difference in the structures
of cash flow between the BOT project and the non-BOT project. Therefore, we can easily
understand that the SLR index can no longer be applied to analyze financial project of the
BOT project. From the concept of Fig. 2, it implies that the B/C method or SLR index, which
Xing and Wu (2000), Lu, et al., (2001), or Chang and Chen (2001) have adopted, was not a
good financial indicator to express the detail relationship between private sectors’ investment
and government investment in the structure of cash flow. To deal with this problem mentioned
above, we will develop new finance model of the BOT project to correct the drawbacks of
SLR index.
3.THE MODEL
The government can calculate royalty as lump sum fee or depending on the output, total
revenue of the BOT project (Wu, 2002). Whereas, the royalty as lump sum fee is easier used
than other methods for public sector (Kang, et al., 2002). The royalty depends on the total
revenue or the output of the BOT project in operation efficiency of the private sector is better
than other methods (Wu, 2002). In this section, we will develop the models which calculate
royalty depending on the total revenue and the output of the BOT project.
3.1 Notation
The notations of variables are defined as following.
Cgt: government investment cost at time t during the construction period;
Cpt: private investment cost at time t during the construction period;
C: the sum of present value of the construction cost which is discounted to the first year in
construction period of BOT project during the construction period (including land
acquisition cost, engineering cost, but not including capitalized interest);
Journal of the Eastern Asia Society for Transportation Studies, Vol.5, October, 2003
2923
Cg: the sum of present value of the construction cost by government investment, and the cost
is discounted to the first year during the construction period;
Cp: the sum of present value of the construction cost by private investment, and the cost is
discounted to the first year during the construction period;
Rt: nominal total revenue of the project at time t;
Kt: nominal operating cost (including asset maintained cost, asset replacement cost, salary,
change in net working capital, other cost, but not including land-use rent, royalty, interest
expense);
Bt: nominal land rent at time t;
Tc: marginal tax rate of the BOT project;
Dt: nominal tax expense of the private firm which is supposed to be un-levered at time t;
L: the sum of the present value of the part of construction cost which government agree to pay
in advance, L is included in Cg;
h: the first year for royalty-collection, n + 1 ≤ h ≤ N ;
θ : the royalty rate depends on the total revenue of the BOT project at hth year,
α:the annual growth rate of θ ;α>0 is increasing by growth ratio; α<0 is decreasing by
growth ratio; α=0 presents that the growth rate is constant. To keep the variable of royalty
is not negative value, variables of the α>-1, θ ≧0 and the other variables must be not
negative.
g : the multiplier of the total quantity of the BOT project at hth year,
φ : the annual growth rate of θ ; φ >0 is increasing by growth ratio; φ <0 is decreasing by
growth ratio; φ =0 presents that the growth rate is constant. Also to keep the variable of the
royalty is not negative value, we assume the variable of φ >-1,g≧0 and other variables
must not be negative.
Qt:quantity of the BOT project at time t;
Journal of the Eastern Asia Society for Transportation Studies, Vol.5, October, 2003
2924
Cgt
Government BOT project
Cpt Kt Rt
Bt+θ*(1+α)t*Rt + Dt
Concessionaire
Concept of the Figure 3 shows that the investment funds source of the BOT project contains
the government investment and the private sectors investment. The construction cost resource
comes from government investment (Cgt) and private investment (Cpt). Private sector pay the
land-used rent, royalty and tax (that is Bt+θ(1+α)t Rt + Dt) to the government by operating
revenue and affiliated business income.
Let GCCR (government construction cost ratio) and PCCR (private construction cost
ratio) are government investment ratio in construction cost and private investment in
construction cost, respectively.
Cg Cg
GCCR = = (1)
C Cg + C p
Cp Cp
PCCR= = (2)
C Cg + C p
It can be calculated as C g = GCCR × C , C p = PCCR × C by Equations (1) and (2). Since
C g + C p = C , both sides of this equation divide by C then get GCCR + PCCR = 1 .
Considered variable of Cg, Cg is part of the construction cost which government agrees to pay.
L
Therefore Cg≧L, and ≤ GCCR ≤ 1 . Since GCCR + PCCR = 1 , so the upper and lower
C
bound value of PCCR is:
L
0 ≤ PCCR ≤ 1 − (3)
C
Equation (3) implies that the upper and lower bound value of the PCCR are affected by
variable of Cg and L. Assumed private sector used the NPV method to evaluate the BOT
project and used the WACC method to determine the cost of capital. Based on the concept of
Fig. 3 and Equations (1) and (2). The NPV p1 was defined as:
NPV p1 = NI − θ ⋅ fac p1 − PCCR ⋅ C (4)
where NI = ∑
N Rt − K t − Bt − Dt N
; fac p1 = ∑
(1 + α ) Rt ; r is the risk-adjusted cost of
t −h
t =0 (1 + r )
t
t =h (1 + r )t
capital of the project of the concessionaire, and r > i .
The weighted-average-cost-of-capital with corporate tax is:
B S
r = rB × (1 − Tc ) × ( ) + rs × ( ) (5)
S+B S+B
where rB is the cost of long-term debt of the BOT project for the private firm; rS is the cost
of equity of the BOT project for the private firm; B is the market value of the debt of the BOT
Journal of the Eastern Asia Society for Transportation Studies, Vol.5, October, 2003
2925
project for the private firm; S is the market value of the equity of the BOT project for the
private firm.
Although the private sector uses NPV p1 ≥ 0 rule to evaluate financial project of the BOT
project, the government will let NPV p1 be equal to 0, i.e. NPV p1 = 0 , because government
wants to get more royalty from the private sector to eliminate the profit premium of the
concessionaire. Therefore, setting Equation (4) equals 0, that is
NI − θ ⋅ fac p1 − PCCR ⋅ C = 0 . Then get θ ⋅ fac p1 = NI − PCCR ⋅ C . So the value of θ is:
θ = ( NI − PCCR ⋅ C ) / fac p1 (6)
Equation (6) shows that θ and PCCR exist negative relationship, θ and GCCR exist
positive relationship. This relationship implies that the more private sectors investment cost in
construction cost, the lower private sector willing to pay royalty to public sector. Since the
value of royalty should not be negative value so the ( NI − PCCR ⋅ C ) must be not negative
value, i.e. ( NI − PCCR ⋅ C ) ≥ 0 . So we can obtain the upper bounded value of the PCCR is:
1
PCCR ≤ ( NI ) (7)
C
From Equations (3) and (7), the interval value of the PCCR can be obtained:
L 1
0 ≤ PCCR ≤ min[1 − , ( NI )] (8)
C C
To keep equation (8) exist a solution, the NI must not negative value, that is NI ≧0. If
NI =0, then PCCR=0 and get θ =0, so the BOT project becomes the OT model
(operating-transfer approach). If NI <0, it implies that the operating revenue is deficit for the
BOT project, so the financial model proposed by this paper can not use to evaluate financial
project for BOT projects.
Towards to realize the financial recovery ability for government according the concept of
the Figure 3. We define GFRR (Government Finance Recovery Ratio) as the proportion of the
government income that can recover government investment. The government income
includes royalty, land-used rent, and tax to cover its investment cost. We use the GFRR index
to evaluate government’s financial ability but not SLR. And the GFRR was defined as
Equation (9).
GFRR=
1
Cg
[
ren + θ × fac p1 ] (9)
where ren = ∑
N
Bt + Dt N
; fac p1 = ∑
(1 + α )t − h Rt ; i is the interest rate of the government
t =0 (1 + i )t t =h (1 + r )t
bond.
Equation (9) shows that the GFRR and (ren + θ × fac p1 ) exist positive relationship. It infers
that the more of collection in royalty, tax, and land-used rent the higher of the GFRR index.
The GFRR is increasing when royalty is increasing; GFRR is decreasing when GCCR is
increasing.
Comparing SLR and GFRR, it is easy to understand that the SLR was just only used to
analyze the non-BOT or non-joint venture project, and the SLR can not describe the
relationship of between SLR and PCCR and between SLR and GCCR. But the GFRR can be
used to represent the characters of the financial project of the BOT project. Also it can be
Journal of the Eastern Asia Society for Transportation Studies, Vol.5, October, 2003
2926
provided us to find the value of the PCCR, GCCR, and royalty and to get the relationships
among the variables of PCCR, GCCR, GFRR, and royalty.
Use equations of the (1), (2), (6), (9), and GCCR + PCCR = 1 to solve the value of GFRR,
and we obtain the value of GFRR* :
fac g1
GFRR*= 1
ren + ( )( NI − PCCR × C ) (10)
C × (1 − PCCR ) fac p1
whereas PCCR ≠ 1 of equation (10). If PCCR=1, the GFRR can’t be computed by equation (9)
because the government investment is zero. Moreover, considered the GFFR association with
the minimum value by the legislation-regulated case. Let the GFRR0 be the minimum value of
GFRR, and GFRR0 will not be greater than 1, that is GFRR0 ∈ [0,1], GFRR ≥ GFRR0.
Use Equation (10) and GFRR0 ∈ [0,1] to solve GFRR, and we obtain Equation (11).
1 fac g1
ren + ( )( NI − PCCR × C ) ≧GFRR0 (11)
C × (1 − PCCR ) fac p1
Re-arranged the equation (11), the upper value of the PCCR is:
facg1
PCCR≦ ren + ( fac p1 )( NI ) − GFRR0 × C = X1 (12)
facg1
C × − GFRR0
fac
p1
Also, we use the Equations (8) and (12) to find the interval value of the PCCR is:
L 1
0 ≤ PCCR ≤ min[1 − , ( NI ), X 1 ] (13)
C C
Equation (13) shows that the lowest solution of PCCR is 0, the upper value of PCCR is the
L 1
minimum value of [1 − , ( NI ), X 1 ] under three conditions including the royalty value is
C C
not negative, GFRR has lower bounded value, and the government agree to bear the
proportion of construction cost of the BOT project.
To keep the equation (13) hold, value of the X1 of equation (13) must be not negative value,
fac g1
that is [ren + ( )( NI ) − GFRR0 × C ] ≧0。Use equation (13) and we obtain the value of
fac p1
GFRR0 is:
fac g1
GFRR0 ≦ (ren + ( )( NI )) / C =X2 (14)
fac p1
Equation (14) shows that the GFRR0 was affected by variables of the operating cost, project
revenue, land-used rent, tax, development fee, royalty, and construction cost. We obtain the
interval value of GFRR0 by equation (14) and GFRR0 ∈ [0,1], GFRR0 is:
0≦GFRR0≦min[1,X2] (15)
Equation (15) represents that the government finance recovery ratio, which was set by
government, must be considered the factors of the private sectors’ investment and minimum
lower bounded value of GFRR. Also, we get the PCCR* by equation (13), the PCCR* is:
L 1
PCCR*= min[1 − , ( NI ), X 1 ] (16)
C C
Since GCCR * + PCCR * = 1 , we use GCCR * + PCCR * = 1 and equation (16) can be easy to
obtain the GCCR * value; similarly, substituted the PCCR* value into the equations (10) and
Journal of the Eastern Asia Society for Transportation Studies, Vol.5, October, 2003
2927
(6), so we get the GFRR* and θ* value, respectively. The values of GFRR* and θ* are
following as equation (17) and (18), respectively.
fac g1
GFRR*= 1
ren + ( )( NI − PCCR * × C ) (17)
*
(
C × 1 − PCCR )fac p1
θ * = ( NI − PCCR * × C ) / fac p1 (18)
Equations (18) and (17) show that the GFRR* value was affected by variables of the royalty,
land-used rent, PCCR, and α. Also it infers that there exist the negative relationship between
θ* and private operating income despite the case of α>0, α<0, or α=0. The result of
equations (17) and (18) imply that the higher of the GFRR*, the more collection in the royalty,
land-used rent, and tax.
The result of model I shows that the government can easy to compute the PCCR*, GFRR*,
and annual royalty value under government has pre-set the lower value of government
financial recovery ratio for the BOT project.
Suppose nominal annual royalty depends on the multiplier of the output of the BOT project.
Concept of the cash flow is shown in Figure 4. The difference between Figure 4 and Figure 3
is the value of the Bt+ g (1+ φ )tQt + Dt and the Bt+θ(1+α)t Rt + Dt. The other assumptions
for model II are the same with model I.
Cgt
Government BOT project
Cpt Kt Rt
t
Bt+ g (1+φ ) Qt + Dt
Concessionaire
We follow the conception of model I to construct the model II. Similarly, use the NPV method
to evaluate the financial project of the BOT project by private sector investment. Then
Equation (4) can be modified as equation (19):
NPV p 2 = NI − g × fac p 2 − PCCR × C (19)
where, fac p 2 = ∑ (1 + φ )
t −h
N
× Qt .
t =h (1 + r ) t
Let NPV p 2 =0, i.e. NI − g × fac p 2 − PCCR × C = 0 . Re-arranged it and we get
g * × fac p 2 = ( NI − PCCR × C ) , and the value of g * was shown as equation (20).
g * = ( NI − PCCR × C ) / fac p 2 (20)
Journal of the Eastern Asia Society for Transportation Studies, Vol.5, October, 2003
2928
Equation (20) shows that there exist the negative relationship between value of g* and Qt ,
and the positive relationship between value of g* and φ . Also we obtain the value of GFRR*,
GFRR0, PCCR*, and g* according to the procedure of developed model I, respectively. The
result of GFRR* was shown as equation (21),
1
GFRR*= [ren + ratio× (NI − PCCR× C)] (21)
C × (1 − PCCR)
The interval value of GFRR0 was shown as equation (22):
0≦GFRR0≦min[ 1, 1 [ren + ratio × NI ] ] (22)
C
N (1 + φ )t − h × Qt
∑
where ratio = t = h (1 + i )t . And we get the interval value of PCCR was shown as
N
(1 + φ )t − h × Qt
∑
t =h (1 + r )t
equation (23):
L 1
0 ≤ PCCR ≤ min[1 − , ( NI ), X 3 ] (23)
C C
ren + ratio × NI − GFRR0 × C
where, X3= .
C [ratio − GFRR0 ]
From equation (23) and PCCR*+GCCR*=1, therefore, we get the value of PCCR* was shown
as equation (24).
PCCR*= min[ 1- L , NI / C , X3] (24)
C
The value of GFRR* and g*can be obtained according to the value of PCCR*. The result of the
GFRR* and g* were shown as equation (25) and (26), respectively:
GFRR* =
1
(
C 1 − PCCR*
)
[
ren+ ratio(NI − PCCR* × C) ] (25)
Journal of the Eastern Asia Society for Transportation Studies, Vol.5, October, 2003
2929
PCCR* L 1
min[1 − , ( NI ), X 1 ]
C C
GFRR* 1 fac g1
*
ren + ( )( NI − PCCR * × C )
(
C 1 − PCCR ) fac p1
Annual nominal royalty value
(1+α)t−h ×θ* ×Rt ( ( NI − PCCR * × C ) / fac p1 ) (1+α) Rt
t −h
GFRR* 1
[
ren+ ratio(NI − PCCR* ×C) ]
(
C 1− PCCR*
)
Annual nominal royalty value (( NI − PCCR * × C ) / fac p 2 )(1 + φ ) t − h Qt
(1+φ)t−h × g* ×Qt
The results of these two models show that the annual royalty value and GFRR* value of
model I and II are different in variables Qt × g * × (1 + φ )t − h and Rt × θ * × (1 + α )t − h . Also it shows
that there exists relationship among the PCCR, GCCR, ongoing royalty, and GFRR. The new
finance model for of the BOT project in this paper we have developed could be applied to the
BOT project practice.
Some issues of this paper can be explored for future study, (1) to develop finance model for
PCCR, GCCR, subsidy, and GFRR for BOT projects. Because the model in this paper cannot
be used to analyze operating revenue associate with deficit of the BOT project. (2) To
construct finance model of the BOT project according to the private sector’s viewpoint, and (3)
to develop bargain model of the royalty and the proportion of the construction cost for BOT
projects by using game theory or mathematical programming. To illustrate the model I and II,
the empirically case should be explored for future study.
Journal of the Eastern Asia Society for Transportation Studies, Vol.5, October, 2003
2930
ACKNOWLEDGMENTS
The authors would like to thank the National Science Council of the Republic of China for
financial supporting this research under Contract No. NSC90-2416-H-126-011.
REFERENCES
Shan-Ying Wu, (2002) Royalty Models for the BOT Projects of Transportation Infrastructure,
Master thesis of the Institute of Traffic and Transportation, National Chiao Tung University.
Chao-Chung Kang, Cheng-Min Feng, and Szu-Chi Huang, (2002) “Develop of the Royalty
Model of BOT Projects: The Lump sum case”, Proceedings of the 17th annual conference
for the Chinese Institute of Transportation, Taiwan, 66-74 (in Chinese).
Chao-Chung Kang, Cheng-Min Feng, and Szu-Chi Huang, (2003) “Develop of the BOT
Finance Model- the GFRR, PCCR, GCCR, and royalty Model”, Transportation Planning
Journal Quarterly, (in Chinese, forthcoming).
Chang, L. M. and Chen, P. H., (2001) “BOT Financial Model: Taiwan High Speed Rail Case”,
Journal of Construction Engineering and Management, Vol. 127, No. 3, 214-222.
Daniel, J. I., (2002) “Benefit-Cost Analysis of Airport Infrastructure: the Case of Taxiways”,
Journal of Air Transportation Management, Vol. 8, 149-164.
Finnerty, J. D., (1996) Project Financing: Asset-Based Financial Engineering. New York:
John Wiley & Sons, INC.
Lu, Y. C., Wu, Soushan, Chen, D. H., and Lin, Y.Y., (2000) “BOT Projects in Taiwan:
Financial Modeling Risk, Term Structure of Net Cash Flows, and Project at Risk Analysis”,
The Journal of Project Finance, 53-63.
Musgrave, R. A., (1959) The Theory of Public Financing- A Study in Public Economy.
London.
Walker, C. and Smith, A. J., (1995) Privatized Infrastructure: the Build Operate Transfer
Approach. Thomas Telford Publication, London.
Journal of the Eastern Asia Society for Transportation Studies, Vol.5, October, 2003
2931
Windsperger, J., (2001) “The Fee Structure in Franchising: A Property Rights View”,
Economic Letters, Vol. 73, 219-226.
Xing, W. and Wu, F. F., (2000, January) “Cost-Benefit Analysis of BOT Power Plants”,
Proceedings of the IEEE PES Winter Meeting, Singapore, 23-27.
Journal of the Eastern Asia Society for Transportation Studies, Vol.5, October, 2003