13 September 2013

Portfolio liquidity risk

measures based on our
market impact model
In this note, we look at historical liquidity trends. We show that long-term
turnover evolution is driven mostly by price levels. We analyse volumes and
demonstrate that seasonality explains only part of the variations observed. We
examine bid-ask spreads across European small, mid and large caps and
highlight the strong link between bid-ask spreads and market volatility.
We then present our market impact model and show how it can be used in
practice to estimate expected trading costs. We illustrate the main properties
of market impact models. Our model gives practical results applicable within
normal trading conditions.
Based on our market impact model, we have built three liquidity risk measures
on trading duration, liquidation market impact and liquidation market
risk. We illustrate these three measures with small caps portfolios. Lastly, we
show how these measures can help risk managers meet the future ESMA
requirements for liquidity risk reporting.

Authors
Paul Besson
Head of Quantitative Research
pbesson@keplercheuvreux.com
+33 1 70 81 57 34

Stphane Galzin
Quant Research Analyst
sgalzin@keplercheuvreux.com
+33 1 70 81 57 33

Stphanie Pelin
Quant Research Analyst
spelin@keplercheuvreux.com
+33 1 70 81 57 32

Matthieu Lasnier
Quant Research Analyst
mlasnier@keplercheuvreux.com
+33 1 70 81 57 31

IMPORTANT. Please refer to the last page of this report for

Important disclosures and analyst(s) certifications

keplercheuvreux.com

Quant Research

Contents
Introduction

Liquidity and its two components ..................

Volumes and turnover: fundamental characteristics

Bid-ask spreads vary in line with market volatility

Market impact overview: theory & practice

Market impact models share common properties

Practical use and model validity

11

Measuring liquidity risk on a portfolio .........

12

Portfolio liquidity measures for different market contexts

12

Practical applications for future ESMA regulatory framework

15

Legal and disclosure information ....................

17

keplercheuvreux.com

Quant Research

Introduction
Before 2008, liquidity issues were often considered to be secondary in asset management.
Since then, liquidity risk has become a key element for risk managers. Today new standards
are emerging and regulators are specifying new obligations.
Liquidity risk has always been a key question for execution and market microstructure
academics. Liquidity risk measures are rooted in market impact models that estimate the
cost of liquidity. These models are extensively used for optimal execution and trading
algorithms.
In this note we show how key quantitative tools, initially developed for execution, can be
applied to liquidity risk analysis for portfolio managers. We first present the main
characteristics of liquidity: volumes and bid-ask spreads. We highlight the general
properties of market impact models. We then present three key measures of liquidity risk.
We conclude by discussing ESMAs new regulatory requirements.

keplercheuvreux.com

Liquidity risk has

become a key issue for
risk managers since
2008
We propose three
portfolio liquidity
measures based on our
market impact model

Quant Research

Liquidity and its two components

The liquidity risk of a portfolio can be defined as the costs incurred from the sale
of part of its assets in a given market environment. A portfolios liquidity
measures depend on its holdings but also on current market liquidity conditions.
In this section, we analyse market liquidity characteristics.
Market liquidity has two main components: volumes and bid-ask spreads. They
both characterise the capacity of the market to execute a transaction without
generating too large a market impact.

Volumes and turnover: fundamental characteristics

Long-term view on European stocks since 2007
The most well-known liquidity measure is market turnover (in cash), but this measure is
greatly influenced by price levels. For this reason volumes (in numbers of shares) are used
to measure trading activity independently of price effects.
Compared with the pre-crisis year of 2007, subsequent years experienced a drop in
turnover, as we can see in Chart 1 (dark blue bars). This drop was mainly driven by prices, as
shown by the Stoxx 600 Index (orange line). Volumes expressed in numbers of shares were
much less volatile. The best example of this situation is the year 2009, when turnover
declined -41% while volumes rose +9% compared with 2007 levels.
The year 2013 to date shows a +9% rise in turnover compared with 2012, accompanied
with a -5% drop in volumes in number of shares. The rise in turnover was mostly driven by
European markets recent performances.

Chart 1: Yearly median turnover and volume since 2007*

140
Turnover
Volume
DJ STOXX 600

125
120
109

109

110
101

100

96

91

74

73

59

60

40

80

79

80

2007*

2008

2009

2010

2011

2012

2013*

Source: Kepler Cheuvreux data from mid-2007* to August 2013* on Stoxx 600 Europe constituents, equal l weighted measure

keplercheuvreux.com

Turnover is greatly
influenced by price.
For this reason volume
is used to measure
trading activity
independently of price
effects

Quant Research

Volume dynamics across large, mid and small caps vary accordingly
From 2007 to 2013, the components of the Stoxx Small Cap Index have had a daily median
turnover of EUR5m, compared with EUR16m and EUR78m for mid and large caps, as
shown in Chart 2. Volumes evolve in similar ways across market capitalisations, as we can
see below. We note large traded volumes on large caps in 2009 (orange line) and strong
traded volumes on small caps since 2012 (dark blue line).

Chart 2: Monthly volumes across capitalisations

90

240

80

220

70

200

60
50
40
30
16

20
10
0

Small Mid Large

Median daily volume

(nb shares, base 100 in 2007*)

Median daily turnover (mEUR, 2007-2013*)

78

Small
Mid
Large

180
160
140
120
100
80
60 Ja Ju D Ju D
n -2 n- 2 ec- n- 2 ec- Jun- 2Dec- Jun- 2Dec- Jun- 2Dec- Jun- 2Dec- Jun- 2Dec00 00 200 00 200 00 200 01 201 01 201 01 201 01 201
7
7
7 8
8 9
9 0
0 1
1 2
2 3
3
Source: Kepler Cheuvreux data from mid-2007 to August 2013* on Stoxx 600 Europe constituents

Beyond their seasonal profile, volumes remain highly volatile

Volume seasonality is a very well-documented phenomenon. Chart 3 below shows the
weekly median volume pattern. The dark blue line shows the two typical characteristics of
volumes: a large drop usually occurs in the last two weeks of December (-46%, and 66%),
and another usually comes in the second week of August (-18%), compared with median
yearly levels.

keplercheuvreux.com

Quant Research

Chart 3: Volume seasonality on a weekly basis

Weekly volume / annual weekly volume

220
200

Despite predictable
seasonality, volumes
remain highly volatile

Second week of August (-18%)

180

To reflect this
volatility, it is essential
to take confidence
intervals into account

160
140
120
100
80
60
40

Decembers last two

weeks (-46 and -66%)

10% and 90% percentiles

Median

20
Jan

Feb

Mar

Apr

May

Jun

Jul

Aug Sep

Oct

Nov Dec

Source: Kepler Cheuvreux data from mid-2007 to August 2013 on Stoxx 600 Europe constituents.

In addition to median patterns, the light blue lines in Chart 3 above show the 10% and 90%
percentiles of weekly volumes. We can clearly see the large variations that occur for a
given week across different years. This shows that the variations in volumes are only partly
explained by yearly seasonality.

Chart 4: Daily volume distribution

Number of observations

150

We promote the use of

VaR metrics when
estimating liquidity
risk

100

10% percentile: -24%

50

50

100
150
200
Volume / yearly volume (%)

250

300

Source: Kepler Cheuvreux data from mid-2007 to August 2013 on Stoxx 600 Europe constituents

Another way to measure the variability of volumes is to observe the past volume
distribution. We can see in the histogram (Chart 4 above) that in the 10% worst cases, daily
volumes were 24% below their yearly median.

keplercheuvreux.com

Quant Research

The high variability of volumes should be taken into account in order to compute liquidity
risk measures. Long-term average volumes give only a snapshot of market reality. We
recommend considering the 10% lowest volume situations that can be experienced in
practice to account for adverse liquidity situations. That is why we promote the use of VaR
estimates for liquidity risk measures.

Bid-ask spreads vary in line with market volatility

The second main market liquidity characteristic is the bid-ask spread. It reflects the
elementary cost of a transaction. Therefore it is important to determine the drivers of bidask spreads to better understand liquidity risk.

Larger caps often associated with tighter bid-ask spreads

Bid-ask spreads vary with market capitalisation and turnover. In Chart 5, it may be noted
however that bid-ask spreads for small, mid and large caps evolve in a similar manner. In
fact, we see that small cap bid-ask spreads are on average 4bps wider than mid cap bid-ask
spreads, which in turn are 2.5bps larger than large cap bid-ask spreads.
The relationship between free-float adjusted market capitalisation and bid-ask spreads is
described in Chart 6. This shows that on average stocks whose rankings in the Stoxx 600
are larger (i.e. smaller capitalisations) will on average have a higher bid-ask spread (+1.4bps
every 100 rank). This property results from competition between market makers, which is
much more intense on more liquid stocks.

Chart 5: Bid-ask spread across capitalisations

Chart 6: Bid-ask spread and index rank

35

20
Mid

Median bid-ask spread (bps)

30

Large
25

Last Main Crisis

20
15
10
5
0

y = 0,014x + 4,4

18
16
14
12
10
8
6
4
2

Jan

O
Ju
A
Ja
A
M
D
N
F
-2 0 ct- 20 l- 20 pr- 20 n -2 0 ov-2 ug-2 ay- 2 eb- 2 ec-2
01
01
01
01
07
10
01
09
07 08
1
3
2
3
0

Source: Kepler Cheuvreux monthly data from mid-2007 to August 2013 on Stoxx 600
Europe constituents

Median bid-ask spread (2013, bps)

Small

keplercheuvreux.com

0
0

200

400

600

DJ STOXX 600 rank

Source: Kepler Cheuvreux data on Stoxx 600 Europe constituents
Each blue point represents one stock in the index.

Quant Research

Bid-ask spreads and historical volatility vary accordingly

As we can see in Chart 5, the largest bid-ask spreads were observed during the last major
crisis. Currently (Q3 2013), bid-ask spreads are narrower than pre-crisis levels. While large
cap bid-ask spreads are close to mid-2007 levels (from 7 to 5 bps), small cap bid-ask
spreads are down by 5bps (from 15 to 10bps).
Market volatility and market bid-ask spreads evolve with very similar patterns (see Chart 7
below). This well-known characteristic stems from the fact that market makers will ask for
a larger margin (bid-ask spread) when they face a greater market risk (for which volatility is
a good proxy). On the other hand, when the market risk is lower, market makers will be
more inclined to reduce their bid-ask spreads.
A more striking result is the connection between bid-ask spreads, volatility and market
capitalisation. Chart 7 below clearly shows that the smaller the stock, the greater the
impact of volatility on its bid-ask spread. For example, a rise from 20% to 40% in volatility
should increase the bid-ask spread by 4bps for large caps and by 8bps for small caps. The
strong link between volatilities and bid-ask spreads explains much of the bid-ask spread
variability.
Volumes as well as bid-ask spreads show a lot of variability. Market liquidity features
cannot therefore be described by long-term averages alone.

Chart 7: Monthly volumes and volatility

Median bid-ask spread (bps)

35
30

Small cap bid-ask

spreads widen earlier
in response to volatility
increases than is the
case for mid and large
cap bid-ask spreads

Small: spread = 0.4 x volatility + 3

Mid : spread = 0.3 x volatility + 3
Large: spread = 0.2 x volatility + 3

25
20
15
10
5
0
10

20

30
40
Average intraday volatility (%)

50

60

Source: Kepler Cheuvreux monthly data from mid-2007 to August 2013 on Stoxx 600 Europe index

To examine this issue further, see the article from Chordia, Sarkar, and Subrahmanyam The
joint dynamics of liquidity, returns, and volatility across small and large firms, Federal Reserve
Bank of New York, Staff Report, 2005.

keplercheuvreux.com

Quant Research

Market impact overview: theory & practice

Based on current liquidity characteristics, market impact models evaluate
expected trading costs. In this section, we first present the general properties of
market impact models. We then discuss when they are applicable and what to do
in practice when assessing the risk of a trading outside the domain of validity of
statistical models.

Market impact models share common properties

Model inputs and formula
Market impact models estimate the price impact of a given order. The price impact is
defined as the difference between the last traded share price, and the first traded share
price of the same order (for a buy order). The usual inputs for these models are:

the trading duration (total duration of the trade, expressed in minutes);

the participation rate (transaction size in shares / total number of traded shares
during the same period;

the bid-ask spread (stocks average bid-ask spread applicable during the
transaction, expressed in bps);

the volatility (some models use volatility although it is possible to work with bidask spreads only, since these two variables are highly correlated, as we saw
earlier).

Our market impact model can be written as follows:

(
(

)
)

Its parameters are estimated on our internal historical executed orders database from
2010.

Main properties
As expected, the estimated market impact increases with the trading duration, the
participation rate and the bid-ask spread. One interesting property of this model is that
trading two different stocks with the same duration and participation rate should lead to
the same market impact expressed in units of bid-ask spreads, as we will see in the
following example.
To get a good overview of the large academic research on this topic, we recommend
Almgren et al., Equity market impact, Risk Magazine, July 2005.

keplercheuvreux.com

Calibration with real

client orders gives us a
simple and robust
market impact model
based only on:
bid-ask spreads,
participation rate and
trading duration

Quant Research

Example on Deutsche Telekom and Portugal Telecom

Let us consider a standard percentage of volume order at 20% participation rate on a 30minute duration on Deutsche Telekom (DTE GY). Its average bid-ask spread is 3bps. The
expected market impact would be:
( )

)= 4.3 bps

= 1.43 DTE GY bid-ask spread

Now let us take a less liquid stock: Portugal Telecom (PTC PL), with an average bid-ask
spread of 7bps. For a standard percentage of volume order at 20% participation rate on a
30-minute duration, the expected market impact would be:
( )

)= 10.0 bps

= 1.43 PTC PL bid-ask spread

Trading with the same participation rate and duration on two different stocks can result in
two different market impacts. But these impacts relative to their stocks own bid-ask
spread will be the same, as we can see above.
Chart 8 (below) displays the estimated market impact for a given participation rate and
trading duration. The estimates are expressed in bid-ask spreads of the underlying stock.
Thus this single surface shows the estimated market impacts for any stock for a given
duration and participation rate.

Chart 8: Participation rate, duration and estimated market impact

Estimated market impact

(20% participation rate,
30 min duration) =
1.43 bid-ask spreads

Source: Kepler Cheuvreux

10

keplercheuvreux.com

Quant Research

Practical use and model validity

Reality-based model designed for usual trading conditions
Our model is estimated with real executions from 2010. Therefore our forecasts will be
much more accurate upon normal trading durations and participation rates. Beyond
standard values, statistical inference lacks relevance. In practice, this means that we
recommend the use of our model with a participation rate between 0% and 30%, and an
execution duration time up to one day execution. Here we can assess the quality of our
model. For unusual trading conditions (participation rates above 30% or trading durations
exceeding one day) a proper statistical estimation of special market impacts is not
achievable since very few transactions occur in such conditions.

Market impact for large participation rates

When participation rates are very large (above 30-35%) trading impacts are particularly
hazardous, as traders well know. In such situations, the ongoing trading is likely to be
discovered by other market participants. Academics call this phenomenon information
leakage. Its main consequence is that the order book anticipates aggressive orders and
moves accordingly. This may result in a huge market impact, which all traders want to
avoid.
For this reason, block trading can be particularly useful in such situations. From a
quantitative point of view we recommend not to use a participation rate beyond 30% even
if the trading duration should lengthen. This of course generates a higher market risk for
the trader but this is usually better than running the risk of market discovery. In such
situations, market risk measures should be preferred to market impact forecasts.

11

keplercheuvreux.com

Quant research

Measuring liquidity risk on a portfolio

Liquidity risk monitoring is a developing field of risk management. It comprises
several measures: the first measure introduced is trading duration; the second
is market impact, which estimates the cost at which the portfolio can be
liquidated; and the third is market risk, which computes the risk incurred by the
portfolio during trading.

Portfolio liquidity measures for different market contexts

Trading duration analysis
In order to control the liquidity risk of portfolios, most risk managers establish basic risk
guidelines. The most frequent liquidity risk constraint is to limit the size of each portfolios
position to a fixed number of trading days given a reasonable participation rate. For
example, a risk limit may state that no position in a portfolio should represent more than 5
normal trading days with a given participation rate of 20%.
At portfolio level, the most common way to analyse liquidity is to study the trading duration
curve. This means, for a given trading duration, representing the percentage of the fund
that can be liquidated within this duration. In the examples below we considered an index
portfolio that replicates the Stoxx Small Caps Europe with EUR100m AUM. This is
illustrated by the dark blue line in the example below (Charts 9 and 10).

Trading duration
measures, based on
real historical volumes
distributions, provide
robust estimates

To take into account the varying volume contexts that can be encountered, it is useful to
represent not only the average duration curves, but also curves that correspond to adverse
liquidity contexts. For this reason, we used shades of blue to show confidence intervals
around the median trading duration curve in Charts 9 and 10 below.

Chart 9: Trading duration (20% participation rate)

Chart 10: Trading duration (zoom)

Median: +0.6 days
Liquidation of 90% of the portfolio

10% worst cases: +0.8 days

Source: Kepler Cheuvreux data from September 2011 to August 2013 on Stoxx Small
Caps Europe index constituents

12

keplercheuvreux.com

Source: Kepler Cheuvreux data from September 2011 to August 2013 on Stoxx Small
Caps Europe index constituents

Quant research

The curve on the left enables a risk manager to assess the liquidity risk of a portfolio. If we
zoom into the curve, as shown in Chart 10 on the right, we can see that, for example, 90% of
the portfolio would take on average 0.6 days of trading to be liquidated. However, in
adverse situations it might take up to 0.8 days (with a 90% confidence interval
corresponding to the lower part of the shaded blue).

Market impact analysis

Market impact analysis estimates the cost at which the portfolio can be liquidated. This
computation is performed through our market impact model. As we saw earlier in this
study, this estimate is only reliable in normal trading conditions. Thus, when trading
duration remains below one day and participation rate below 30%, the market impact
analysis remains valid.
Market impact models take into account trading durations, which are subject to changing

Market impact
measures are
estimated with a VaR
approach to account
for different liquidity
contexts

liquidity conditions. This explains why, given the same portfolio, the market impact analysis
may lead to very different results in different contexts. To get a fair estimate of the adverse
liquidity situations one may face, we plotted the histogram of liquidation costs. We can see
in Chart 11 below the historical liquidation costs of a small cap portfolio during the past
two years, with a 20% participation rate. In Chart 12, the corresponding distribution is
plotted with a VaR (10%).

Chart 11: Liquidation market impact over time

Source: Kepler Cheuvreux data from September 2011 to August 2013 on Stoxx Small
Caps Europe index constituents

Chart 12: Liquidation market impact distribution

Source: Kepler Cheuvreux data from September 2011 to August 2013 on Stoxx Small
Caps Europe index constituents

When portfolios are not liquid enough for market impact models to be relevant, eg because
trading durations are too large, then the theoretical results of the expected market impact
should be treated with caution. In such cases, considering the liquidation market risk is a
helpful alternative as it reflects more accurately the risk incurred by the portfolio while it is
being liquidated.

13

keplercheuvreux.com

Quant research

Liquidation market risk

The liquidation market risk of a portfolio is the market risk incurred during trading.

The longer the trading duration, the higher the market risk.

The more volatile the market, the higher the market risk, since price variations are
stronger.

Our market risk

estimates the potential
portfolios NAV
fluctuations incurred
during liquidation

To compute precisely the liquidation market risk, one has to estimate each portfolio
components trading duration as well as its volatility, and then to compute the result for the
portfolio as a whole.

Chart 13: Liquidation market risk over time

Source: Kepler Cheuvreux data from September 2011 to August 2013 on Stoxx Small
Caps index constituents

Chart 14: Liquidation market risk distribution

Source: Kepler Cheuvreux data from September 2011 to August 2013 on Stoxx Small
Caps index constituents

The results are displayed in Charts 13 and 14. The left side displays the historical
liquidation market risk estimation. The right side shows the histogram of its historical
distribution. As earlier, the VaR (10%) is represented by a vertical line delimiting the righthand side of the distribution tail in Chart 14, the corresponding level is plotted with a
horizontal line in Chart 13. This shows that in 10% of cases, liquidation market risk was
above 1.73%.

AUM and liquidation trading durations

The three different analyses we have just studied are applicable to any portfolio.
Nevertheless, liquidity issues increase with AUM. For example we can see in Chart 15
below that an index fund that replicates the Stoxx Small Caps could liquidate 80% of its
positions (with a 20% participation rate) within 3h25min, 1 day or 2 days for different AUM
(EUR100m, EUR250m and EUR500m).

14

keplercheuvreux.com

Quant research

Chart 15: Liquidation trading duration of Stoxx small caps according to AUM

Different Trading Durations for different AUM

Source: Kepler Cheuvreux data from September 2011 to August 2013 on Stoxx Small Caps index constituents

Practical applications for future ESMA regulatory framework

New ESMA recommendations on liquidity monitoring
In May 2013 the ESMA (European Securities and Markets Authority) issued a consultation
paper on reporting obligations for alternative funds (under Articles 3 and 24 of the AIFMD,
Alternative Investment Fund Managers Directive). These guidelines should be come into
force in 2014. It is very likely that similar liquidity monitoring will be required for UCITS
funds in the next regulation review.
The mandatory liquidity reporting will need to: report the percentage of the funds portfolio
that is capable of being liquidated within each of the liquidity periods specified. Each investment
should be assigned to one period only and such assignment should be based on the shortest period
during which such a position could reasonably be liquidated at or near its carrying value. The
reporting frequency is quarterly, biannual or annual depending on the funds AUM.
To assess the amount of the portfolio that can be liquidated at or near its carrying value, the
use of our market impact model can provide the expected answer. Then in order to
evaluate the percentage of the funds portfolio that is capable of being liquidated within each of
the liquidity periods specified the duration analysis presented earlier will be very helpful to
build and corroborate the new liquidity report required by the ESMA.

15

keplercheuvreux.com

Our different measures

help respond to the
ESMAs new liquidity
risk regulations

Quant research

Practical recommendations
First, liquidity analysis should be conducted using a VaR approach to take into account the
variety of liquidity situations faced in reality. Only such an approach can give a fair idea of
what would happen in adverse situations. VaR computations are widespread for market
risk analysis; they should be equally popular for liquidity risk analysis. Therefore we
recommend the use of a liquidation market impact VaR, as well as a liquidation trading
duration VaR.
Second, in order to put in place the future reporting obligations issued by the ESMA for
alternative funds, we propose rigorous liquidity measures: trading duration, market impact,
and market risk that can help risk managers build the new reports required by future
regulations.
We would be delighted to explain our approach to clients who want us to help them
monitor and estimate their liquidity risk based on our execution experience.

16

keplercheuvreux.com

We can help clients put

in place the ESMAs
new reporting using
our liquidity risk
measures.

Quant research

Notes

17

keplercheuvreux.com

Quant Research

Legal and disclosure information

Copyright Kepler Cheuvreux, 2013. All rights reserved
All prices are those current at the end of the previous trading session unless otherwise indicated. Prices are sourced from local
exchanges via ThomsonReuters or Bloomberg unless otherwise indicated. Data is sourced from Kepler Cheuvreux and subject
companies.
This research report or summary ("Research") has been prepared by Kepler Capital Markets SA or one of its affiliates or branches
(collectively referred to as Kepler Cheuvreux). Kepler Cheuvreux refers to Kepler Capital Markets, Socit anonyme (S.A.) (Kepler
Capital Markets SA) and its affiliates, including CA Cheuvreux, Socit Anonyme (S.A.). This research report has been prepared from
information believed to be reliable. The opinions and projections expressed in this document are entirely those of Kepler Cheuvreux
and are given as part of its normal research activity. Kepler Cheuvreux has not independently verified the information given in this
document. Accordingly, no representation, guarantee or warranty, express or implied, is made as to the accuracy, completeness,
correctness or fairness of this information and opinions contained in this document or the research or analysis upon which such
information and opinions are based. Any opinions or estimates expressed herein reflect the judgment of Kepler Cheuvreux as of the
date the Research was prepared and are subject to change at any time without notice. Unless otherwise stated, the information or
opinions presented, or the research or analysis upon which they are based, are updated as necessary and at least annually.
Not all investment strategies are appropriate at all times, and past performance is not necessarily a guide to future performance. Kepler
Cheuvreux recommends that independent advice should be sought, and that investors should make their own independent decisions as
to whether an investment or instrument is proper or appropriate based on their own individual judgment, their risk-tolerance, and after
consulting their own investment advisers.
Kepler Cheuvreux, its parent companies or its affiliates may effect transactions in the securities described herein for their own account
or for the account of others, may have positions with the issuer thereof, or any of its affiliates, or may perform or seek to perform
securities, investment banking or other services for such issuer or its affiliates. The organizational and administrative arrangements
established by Kepler Cheuvreux for the management of conflicts of interest with respect to investment research are consistent with
rules, regulations or codes applicable to the securities industry. These arrangements can be found in Kepler Cheuvreuxs policy for
managing conflicts of interest, available at www.keplercheuvreux.com. Current research disclosures regarding companies mentioned in
this Research are also available at www.keplercheuvreux.com.
This Research is provided for information purposes only. It is not intended as an offer, invitation or solicitation to buy or sell any of the
securities described or discussed herein and is intended for use only by those Professional Clients to whom it is made available by
Kepler Cheuvreux. This Research is not for distribution to Retail Clients. The recipient acknowledges that, to the extent permitted by
applicable securities laws and regulations, Kepler Cheuvreux disclaims all liability for providing this Research, and accepts no liability
whatsoever for any direct, indirect or consequential loss arising from the use of this document or its contents.
KEPLER CHEUVREUX RESEARCH AND DISTRIBUTION
1. United States: This communication is only intended for, and will only be distributed to, persons residing in any jurisdictions where
such distribution or availability would not be contrary to local law or regulation. This communication must not be acted upon or relied
on by persons in any jurisdiction other than in accordance with local law or regulation and where such person is an investment
professional with the requisite sophistication to understand an investment in such securities of the type communicated and assume the
risks associated therewith.
This communication is confidential and is intended solely for the addressee. It is not to be forwarded to any other person or copied
without the permission of the sender. This communication is provided for information only. It is not a personal recommendation or an
offer to sell or a solicitation to buy the securities mentioned. Investors should obtain independent professional advice before making an
investment.
Notice to U.S. Investors: This material is not for distribution in the United States, except to major US institutional investors as defined
in SEC Rule 15a-6 ("Rule 15a-6"). Kepler Cheuvreux refers to Kepler Capital Markets, Socit anonyme (S.A.) (Kepler Capital Markets
SA) and its affiliates, including CA Cheuvreux, Socit Anonyme (S.A.). Kepler Capital Markets SA has entered into arrangements with
Crdit Agricole Cheuvreux North America, Inc. (CA Cheuvreux North America) and Kepler Capital Markets, Inc. ("KCM, Inc.")
(collectively US Broker-Dealers), which enables this report to be furnished to certain U.S. recipients in reliance on Rule 15a-6 through
US Broker-Dealers registered under the U.S. Securities Exchange Act of 1934, as amended.
Each U.S. recipient of this report represents and agrees, by virtue of its acceptance thereof, that it is a "major U.S. institutional investor"
(as such term is defined in Rule 15a-6) and that it understands the risks involved in executing transactions in such securities. Any U.S.
recipient of this report that wishes to discuss or receive additional information regarding any security or issuer mentioned herein, or
engage in any transaction to purchase or sell or solicit or offer the purchase or sale of such securities, should contact a registered
representative of the US Broker-Dealer that provided the report. Under no circumstance should you direct orders to CA Cheuvreux
North America as payment for products or services provided by KCM, Inc. In turn, under no circumstance should you direct orders to
KCM, Inc. as payment for products or services provided by CA Cheuvreux North America.
CA Cheuvreux North America is a broker-dealer registered with the Securities and Exchange Commission ("SEC"), a member of the
Financial Industry Regulatory Authority ("FINRA"), the New York Stock Exchange ("NYSE"), and the Securities Investor Protection
Corporation ("SIPC"). You can reach CA Cheuvreux North America at 1301 Avenue of the Americas, 15th floor, New York, NY 10019,
phone (212) 492-8800; Equity trading: (212) 492-8825. Further information is also available at www.cheuvreux.com.
KCM, Inc.is a broker-dealer registered with the SEC, a member of FINRA and SIPC. You can reach KCM, Inc. at 600 Lexington Avenue,
New York, NY 10022, phone (212) 710-7606; Equity trading: 212-710-7602. Further information is also available at
www.keplercapitalmarkets.com.
You may obtain information about SIPC, including the SIPC brochure, by contacting SIPC directly at 202-371-8300; website:
http://www.sipc.org.

18

keplercheuvreux.com

Quant Research
CA Cheuvreux North America is a wholly owned subsidiary of Crdit Agricole Cheuvreux S.A. which is, in turn, owned by Kepler Capital
Markets SA. As part of the ownership of Crdit Agricole Cheuvreux S.A., Kepler Capital Markets SA is entitled to distribute certain
financial research produced by Crdit Agricole Cheuvreux S.A.
any insured depository institution, may lose value and are not guaranteed by the entity that published the research as disclosed on the
front page and are not guaranteed by US Broker-Dealers.
Investing in non-U.S. Securities may entail certain risks. The securities referred to in this report and non-U.S. issuers may not be
registered under the U.S. Securities Act of 1933, as amended, and the issuer of such securities may not be subject to U.S. reporting
and/or other requirements. Rule 144A securities may be offered or sold only to persons in the U.S. who are Qualified Institutional
Buyers within the meaning of Rule 144A under the Securities Act. The information available about non-U.S. companies may be limited,
and non-U.S. companies are generally not subject to the same uniform auditing and reporting standards as U.S. companies. Securities of
some non-U.S. companies may not be as liquid as securities of comparable U.S. companies. Securities discussed herein may be rated
below investment grade and should therefore only be considered for inclusion in accounts qualified for speculative investment.
Analysts employed by Kepler Capital Markets SA, a non-U.S. broker-dealer, are not required to take the FINRA analyst exam. The
information contained in this report is intended solely for certain "major U.S. institutional investors" and may not be used or relied upon
by any other person for any purpose. Such information is provided for informational purposes only and does not constitute a solicitation
to buy or an offer to sell any securities under the Securities Act of 1933, as amended, or under any other U.S. federal or state securities
laws, rules or regulations. The investment opportunities discussed in this report may be unsuitable for certain investors depending on
their specific investment objectives, risk tolerance and financial position.
In jurisdictions where US Broker-Dealers are not registered or licensed to trade in securities, or other financial products, transactions
may be executed only in accordance with applicable law and legislation, which may vary from jurisdiction to jurisdiction and which may
require that a transaction be made in accordance with applicable exemptions from registration or licensing requirements.
The information in this publication is based on sources believed to be reliable, but US Broker-Dealers do not make any representation
with respect to its completeness or accuracy. All opinions expressed herein reflect the author's judgment at the original time of
publication, without regard to the date on which you may receive such information, and are subject to change without notice.
US Broker-Dealers or their affiliates may have issued other reports that are inconsistent with, and reach different conclusions from, the
information presented in this report. These publications reflect the different assumptions, views and analytical methods of the analysts
who prepared them. Past performance should not be taken as an indication or guarantee of future performance, and no representation
or warranty, express or implied, is provided in relation to future performance.
US Broker-Dealers and any company affiliated with it may, with respect to any securities discussed herein: (a) take a long or short
position and buy or sell such securities; (b) act as investment and/or commercial bankers for issuers of such securities; (c) act as market
makers for such securities; (d) serve on the board of any issuer of such securities; and (e) act as paid consultant or advisor to any issuer.
The information contained herein may include forward-looking statements within the meaning of U.S. federal securities laws that are
subject to risks and uncertainties. Factors that could cause a company's actual results and financial condition to differ from expectations
include, without limitation: political uncertainty, changes in general economic conditions that adversely affect the level of demand for
the company's products or services, changes in foreign exchange markets, changes in international and domestic financial markets and
in the competitive environment, and other factors relating to the foregoing. All forward-looking statements contained in this report are
qualified in their entirety by this cautionary statement.
2. In the United Kingdom, this report is approved and/or distributed by Kepler Capital Markets SA which is authorized by the Financial
Conduct Authority (FCA).
3. In Italy, this Research is approved and/or distributed by Kepler Capital Markets SA and is not intended for circulation or distribution
either to the public at large or to any other parties other than professional clients and/or qualified counterparties, as defined in
Legislative Decree No.58 of 24 February 1998 as amended, and implementing Consob regulations.
4. In Germany this report is approved and/or distributed by Kepler Capital Markets SA.
THE DISTRIBUTION OF THIS DOCUMENT BY KEPLER CAPITAL MARKETS IN OTHER JURISDICTIONS MAY BE RESTRICTED BY
LAW, AND PERSONS INTO WHOSE POSSESSION THIS DOCUMENT COMES SHOULD INFORM THEMSELVES ABOUT, AND
OBSERVE, ANY SUCH RESTRICTIONS. BY ACCEPTING THIS REPORT YOU AGREE TO BE BOUND BY THE FOREGOING
INSTRUCTIONS.
The MSCI indexes are the exclusive property of MSCI Inc. (MSCI). MSCI and the MSCI index names are service mark(s) of MSCI or its
affiliates and have been licensed for use for certain purposes by KEPLER CHEUVREUX. The financial securities referred to herein are
not sponsored, endorsed, or promoted by MSCI, and MSCI bears no liability with respect to any such financial securities. The
Prospectus contains a more detailed description of the limited relationship MSCI has with KEPLER CHEUVREUX and any related
financial securities. No purchaser, seller or holder of this product, or any other person or entity, should use or refer to any MSCI trade
name, trademark or service mark to sponsor, endorse, market or promote this product without first contacting MSCI to determine
whether MSCIs permission is required. Under no circumstances may any person or entity claim any affiliation with MSCI without the
prior written permission of MSCI.
The Dow Jones GCC IndexSM, or other applicable index, are calculated, distributed and marketed by Dow Jones & Company, Inc., a
licensed trademark of CME Group Index Services LLC ("CME"), and have been licensed for use. All content of the Dow Jones IndexesSM
2012 Dow Jones & Company, Inc.
No part of this report may be reproduced in any manner or redistributed without the prior written permission of Kepler Cheuvreux.

19

keplercheuvreux.com

Local insight, European scale

Amsterdam
Kepler Cheuvreux Benelux
Johannes Vermeerstraat 9
1071 DK Amsterdam
+31 20 573 06 66
Frankfurt
Kepler Cheuvreux Germany
Taunusanlage 18
60325 Frankfurt
+49 69 756960
Geneva
Kepler Cheuvreux SA
Route de Crassier 11
1262 - Eysins
Switzerland
+41 22361 5151
London
Kepler Cheuvreux UK
12th Floor, Moorhouse
120 London Wall
London EC2Y 5ET
+44 20 7621 5100
Madrid
Kepler Cheuvreux Espana
Alcala 95
28009 Madrid
+3491 4365100
Milan
Kepler Cheuvreux Italia
Corso Europa 2
20122 Milano
+39 02 855 07 1
Paris
Kepler Cheuvreux France
112 Avenue Kleber
75016 Paris
+33 1 53653500
Stockholm
Kepler Cheuvreux Nordic
Regeringsgatan 38
10393 Stockholm
+468 723 5100
Vienna
Kepler Cheuvreux Vienna
Schottenring 16/2
Vienna 1010
+43 1 537 124 147
Zurich
Kepler Cheuvreux Switzerland
Stadelhoferstrasse 22
Postfach
8024 Zurich
+41 433336666

20

keplercheuvreux.com

Kepler Cheuvreux has exclusive international distribution rights for UniCredits CEE product.

North America
Boston
Kepler Capital Markets, Inc
225 Franklin Street, Floor 26
Boston, MA 02110
+1 617-295-01100

San Francisco
Kepler Capital Markets, Inc
50 California Street, Suite 1500
San Francisco, CA 94111
+1 415-255-9802

New York
CA Cheuvreux North America
1301 Avenue of the Americas, Floor 15
New York, NY 10019
+1 212-492-8800

New York
Kepler Capital Markets, Inc.
600 Lexington Avenue, Floor 28
10022 New York, NY USA
+1 212-710-7600