Ilias Lekkos [email protected] Haris Giannakidis [email protected] [email protected] Bloomberg Page: <PBGR> Piraeus Bank Amerikis 4, 105 64, Athens Tel: 210 328 8187, Fax: 210 328 8605
Jul 28, 2015
Ilias Lekkos [email protected] Haris Giannakidis [email protected]
Bloomberg Page: <PBGR>
Piraeus BankAmerikis 4, 105 64, AthensTel: 210 328 8187, Fax: 210 328 8605
2March 2015
A large amount of academic and empirical research has repeatedly shown the difficulties in estimating a long-
term equilibrium or a “fair” value of stock indices. All of these studies demonstrate the extremely limited
and controversial ability of predictive models to capture the high variability of stock returns.
The crux when it comes to determining a “fair” value for an equity market, is related to the fact that expected
stock returns depend crucially on the level of future dividends, interest rates and other economic variables.
Unfortunately, estimates of those values require correctly specified statistical projections over the long-term as
well as a large history of measurable data points. For example, variables such as the risk premium are not directly
observable and need to be estimated.
Finally, especially for short investment horizons, equity indices are heavily influenced by a large number of
unpredictable and intangible factors, such as the political climate.
3March 2015
The purpose of this study is to sidestep the pitfalls of the non-predictability of stock markets and to investigate the
evolution of the General Index of the Athens Stock Exchange during the period 1996-2014, using an alternative
approach which enables us:
Ø To deconstruct or segment the General Index returns into three underlying and directly observable factors,
namely the dividend price ratio, the rate of change in book value per share (BV) and the rate of change of price
to book value ratio (P/BV), thereby improving our understanding of the historic development of the Greek equity
market.
Ø To study the covariance of each underlying factor with the macroeconomic environment and the economic cycle.
Ø To diversify our expectations for the evolution of these factors with respect to the current macroeconomic state.
Ø To reconstruct our expectations regarding all the individual factors in a holistic estimate by normalizing our long-
term expectations for the General Index future movements relative to the economic cycle.
4March 2015
Starting from the apparent tautology that the total return on the General Index can be separated into the dividend
yield and capital gains (in the form of index price appreciation), we end up – see appendix A – with the following
three factors which determine equity returns:
It is important to note, that the above-mentioned relationship constitutes an accounting identity and not, strictly
speaking, a theoretical equation. This means that this relationship holds for all values in the domain of its
constituents.
General Index Return
Dividend Price RatioRate of Change in the Book Value
per Share
(BVPS growth)
Rate of Change in the Price to Book Valuation Ratio
(P/BV growth)
5March 2015
Beyond the dividend yield which is easily understood and defined by the dividend policy of listed
companies, the other two factors require further investigation. The rate of change of book value per share is
directly related to the performance of a firm such that securitization of non-distributed profits will lead to an
increase in book value while continuing losses will, sooner or later, lead to reductions in net worth and thus to a
reduction in the book value per share. Therefore the change in BV reflects the positive or negative developments in
the fundamentals of a firm.
In contrast, the rate of change in the price to book value ratio is determined by market forces and reflects the risk
tolerance profile of investors. In periods of high investment risk appetite or high positive expectations, investors are
willing to pay a higher price to own company fundamentals (book value). Conversely, in times of low risk tolerance,
investors rate the same fundamentals value much lower.
General Index Return
Dividend Price Ratio Change in Company Fundamentals
Change in Market Expectations
Source: Piraeus Bank Research, Bloomberg 6March 2015
Evolution of Dividends per Share
As a first step in this analysis, we proceed with the historical evolution of each factor, using data for the last 20 years. Specifically,
the 12-month moving sum of dividends for the General Index is characterized by an upward trend in the period 1996-2008,
recording a historic high in 2008 of €127* per share. After 2008 we observe a short stabilization period and an immediate drop
afterwards due to the unfavourable macroeconomic developments in the country. A historic low was recorded in December 2013
at €11.6 per share.
126,71
11,550,00
20,00
40,00
60,00
80,00
100,00
120,00
140,00
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
12 Month Dividend per Share
*Note that this number should not be taken literary as a per share figure, since we are talking about a value-weighted stock index. This means that all variables are adjusted accordingly by the index divisor and not the number of shares of the listed firms.
Source: Piraeus Bank Research, Bloomberg 7March 2015
Evolution of Book Value per Share
2.174
663
0
500
1000
1500
2000
2500
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
Book Value per Share
The change in the book value of the companies that constitute the General Index is in direct relation to the level of corporate profitability
and therefore the wider course of the Greek economy. Based on that, it is not surprising that during the period 2001-2010 the weighted
book value per share of the index moved upwards, reaching the level of €2,174 per share. Then from mid-2010, it fell dramatically to the
historic low of €663 per share in 2012, influenced both by the overall economic crisis that affected the results of all listed companies and
the banking sector balance sheet impairment from the PSI.
Source: Piraeus Bank Research, Bloomberg 8March 2015
Evolution of Price to Book Value Ratio
6,9
0,50
1
2
3
4
5
6
7
8
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
Price to Book
As mentioned earlier the price to book value ratio reflects investors’ appetite for risk-taking and simultaneously is a discount
mechanism of market expectations regarding future corporate profitability. Excluding the period 1999-2000 where an investor was
willing to pay up to 7 times more than the book value of the General Index members, the price to book value ratio hovered
between a historic low of 0.5 (late 2010) and 2.8.
9March 2015
Athens General Index Factors
Source: Piraeus Bank Research, Bloomberg
Looking at the annual returns of the General Index for the period1996-2014 and the evolution of the three individual factors wecan conclude the following:
Ø The period 1997-1999 is characterized by excessiveinvestment risk appetite and positive expectationsas stock market returns are mainly driven by changes inthe P/BV ratio and not by corporate fundamentals.
Ø In contrast, the next three years, 2000-2002, arecharacterized by a complete reversal and shrinkageof valuations and market expectations (negativechange in P/BV).
Ø The years 2008 and 2014 can be described asperiods of significant deterioration in expectations(de-rating) with regard to the evolution of thestock market as it recorded a large drop in the ratioP/BV. Of particular interest is the year 2014 as thedowngrading of expectations is not consistent withthe improvement in corporate data. In contrast,2012 was a period of significant improvement inexpectations (re-rating) despite the retreat ofcorporate data.
Ø The five-year period 2003-2007 is characterized byaverage annual returns of 20% which werebolstered by a relatively balanced relationshipbetween market expectations (P/BV growth) andfirm fundamentals (BV growth).
-120
-100
-80
-60
-40
-20
0
20
40
60
80
100
pe
rce
ntag
e p
oin
ts
DP BV growth PB growth Return
10March 2015
Cumulative Rates of Change for Stock Prices and Factors
Source: Piraeus Bank Research, Bloomberg
The previous findings are confirmed by examining the cumulativereturns of the General Index as well as the cumulative changes ofits individual components. More specifically:
Ø Analyzing the cumulative returns of the Athens StockExchange we note that an investor who placed $100 in late1996 in the Greek stock market, liquidated his/her position at€154 in 2014.
Ø It is striking that this performance gain is drivenexclusively via the dividend policy of the equitymarket’s listed companies.
Ø Throughout the decade from 1996 to 2005, companyfundamentals or else, the variation of BV, had no materialeffect on the performance of the General Index. On thecontrary, the big stock decline in 2011 and 2012 isinterpreted as the dramatic reduction of the book value oflisted firms which was a direct consequence of the profoundeconomic crisis that struck particularly the banking sector.
Ø During the period 1996-2003, fluctuations in theGeneral Index (either positive or negative) weredetermined almost exclusively by the changes inmarket expectations and the risk tolerance attitudesamong investors.
-100%
-50%
0%
50%
100%
150%
200%
250%
19
96
19
97
19
98
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09
20
10
20
11
20
12
20
13
20
14
DP BV growth PB growth Return
Source: Piraeus Bank Research, Bloomberg 11March 2015
Athens General Index Average Returns
The past 20 years of stock return data, confirm that the average annual return of the General Index of 2.3% was determined mainly by
the dividend price ratio of 2.82 percentage points. In contrast, the average effect of the changes in P/BV and BV was only 0.32% and
0.85% respectively. However, the decomposition of historically expected index returns tells us perhaps only half the story. It is obvious
that using two different reference periods, that before the crisis of 2008 and the period after the financial crisis, leads us to a different
interpretation. In particular, both in the post crisis period and in the period between 1996-2007, the rate of change of book value and
that of P/BV ratio were the main factors driving stock returns. In conclusion, we can say that the effect of each factor varies with time
and this time-varying feature is most likely dependent on the domestic and international macroeconomic environment.
2,82 3,61 2,36
-7,46
3,00
-18,76
11,45
-30
-25
-20
-15
-10
-5
0
5
10
15
20
1996-2014 2008-2014 1996-2007
perce
nt
PB growth BV growth DP
22,61
2,29
16,81
Source: Piraeus Bank Research, Bloomberg 12March 2015
The timely identification of economic cycles can giveus important evidence for the expected long-termequilibrium level of a stock index.
However, because of the delay in the announcement and ofthe data point frequency constraints of macroeconomicvariables, their use in determining the phase of economiccycles is considered quite difficult.
For this reason, we will use the seasonally adjustedEuropean Commission's Economic Climate Index for Greece(ESI). This indicator is constructed via the use of researchquestionnaires and covers areas such as construction,services, consumer confidence, industrial production andretail trade.
Economic Sentiment Index
60
70
80
90
100
110
120
130
ESI ESI sa
Source: Piraeus Bank Research, Bloomberg 13March 2015
Assuming that the phases of the economic cycle are thefollowing:
We classify the dividend price ratio and the rate of changeof P/B and BVPS depending on the phase of the cycle inwhich they are realized.
Expansion
Slowdown
Recession
Upturn
Economic Cycle Phase Tracker
January2009
February2015
-2,50
-2,00
-1,50
-1,00
-0,50
0,00
0,50
1,00
1,50
2,00
2,50
-0,3 -0,2 -0,2 -0,1 -0,1 0,0 0,1 0,1 0,2
Le
ve
l (i
n s
ta
nd
ard
d
ev
iatio
ns)
Monthly Change
Expansion
Recession Upturn
Slowdown
Source: Piraeus Bank Research, Bloomberg 14March 2015
General Index Monthly Returns per Economic Phase It is reasonable to assume that the economic cycle affectsthe rates of change of the three components, with thiseffect being different for each component.
The results from the classification of the General Indexmonthly returns for the four economic cycle phases, aresummarized as follows:
Ø The average monthly return of the General Index inthe phase of recovery and recession is different fromthe historical average of stock returns.
Ø We observe extreme performances (higher or lowerthan 3 standard deviations) and greater volatilityduring the phases of downturn and recession.
Ø It is quite interesting that there are high positivereturns in the phase of slowdown and relatively highnegative returns in the upturn phase.
-4
-3
-2
-1
0
1
2
3
4
EXPANSION
DOWNTU
RN
RECESSION
UPTURN
Stan
dard
Dev
iatio
ns
Source: Piraeus Bank Research, Bloomberg 15March 2015
Book Value Rate of Change per Economic Phase We can better understand stock index returns in each phase of the economic cycle by examining their individual components. Starting with the changes in book value per share (or the change of corporate fundamental) we observe that:
Ø In the recession phase, we do not identify theoccurrence of extreme negative changes exceptduring December 2011, where the book value pershare decreased from €1637 to €891.
Ø Outliers and higher variability are observed during thephases of upturn and expansion.
Ø Apart from the recession phase, in all other phasesthe average change in book value per share is veryclose to the historical average of the total number ofobservations.
-8
-6
-4
-2
0
2
4
EXPANSION
DOWNTU
RN
RECESSION
UPTURN
Stan
dard
Dev
iatio
ns
Source: Piraeus Bank Research, Bloomberg 16March 2015
Price to Book Rate of Change per Economic Phase From the monthly changes in market expectations we distinguish the following:
Ø Slow adjustment of valuations during the significantdrop in book value in December 2011, where thechange in the P/BV was about 5 standard deviationsgreater than the global mean change.
Ø Extreme adverse changes in valuations (de-ratings)occurred in expansion periods in December 1999 andsix months later while the highest positive change(re-rating) occurred in January 1999.
Ø Of particular interest is the fact that by accounting forthe extreme values during the four phases of theeconomic cycle, the uncertainty in the valuation ofmarkets is greater in the phases of downturn andupturn and not in the expansions and recessions.
-4
-3
-2
-1
0
1
2
3
4
5
EXPANSION
DOWNTU
RN
RECESSION
UPTURN
Stan
dard
Dev
iatio
ns
17March 2015
Source: Piraeus Bank Research, Bloomberg
0
5
10
15
20
25
30
35
DP PB growth BV growth Est Return
pe
rce
nt
Expansion
-15
-10
-5
0
5
10
DP PB growth BV growth Est Return
perc
ent
Downturn
-50
-40
-30
-20
-10
0
10
DP PB growth BV growth Est Return
pe
rce
nt
Recession
-5
0
5
10
15
20
25
DP PB growth BV growth Est Return
pe
rce
nt
Upturn
18March 2015
Source: Piraeus Bank Research, Bloomberg
The main feature of all phases is the significant impact of the change in P/BV, i.e. the expectations of future
developments in corporate profitability. Unsurprisingly, expectations are negative during the downturn and recession
phases and positive during upturns and phases of expansion.
Although this observation may be considered trivial, a closer look reveals that in the long term, negative expectations
during adverse cyclical conditions offset positive expectations in the most favourable phases.
In contrast, corporate strategy does not seem to follow the same course as market expectations. Specifically, while the
dividend yield is almost independent of the business cycle, the rate of change in the book value of firms seems to be
correlated with the economic environment. In the phases of expansion and downturn the book value per share
increased by an average of 4-5 percentage points per annum. On the contrary, during the recession phase a dramatic
decrease of 16.5% is experienced while, in the upturn phase, the rate of change is almost imperceptible.
Source: Piraeus Bank Research, Bloomberg 19March 2015
By considering each economic phase separately, weobserve that during the expansion stage in the economiccycle:
Ø Market expectations, company fundamentals anddividends contribute positively to the overallannualized General Index performance that goesbeyond (on average) 30%.
Ø The contribution of Price to Book changes reach onaverage three quarters of total return with theremaining performance coming mainly from thechange in Net Worth of the firms in the stock index.
0
5
10
15
20
25
30
35
DP PB growth BV growth Est Return
pe
rce
nt
Expansion
It is important to add that according to the ESI index the average duration of the expansion phase is estimated to be 14 months.
Source: Piraeus Bank Research, Bloomberg 20March 2015
-15
-10
-5
0
5
10
DP PB growth BV growth Est Return
pe
rce
nt
DownturnThe downturn phases are characterized by:
Ø Negative average change in market expectations at anannualized rate of 14%.
Ø Positive change in company fundamentals atapproximately the same level as in the expansionphase.
Ø Faster and simultaneously stronger market reaction tothe deterioration of economic conditions, which dragsdown the average annual return of the General StockIndex to -7%.
Source: Piraeus Bank Research, Bloomberg 21March 2015
Similarly, in periods of recession, total returns for the index are characterized by:
Ø Significant changes in the valuation of markets for stocks which on average surpass -30%.
Ø Strong negative impact on the balance sheets of companies. The average change in the equity return is on average -17% per annum.
-50
-40
-30
-20
-10
0
10
DP PB growth BV growth Est Return
pe
rce
nt
Recession
We should also note that the recession phase is shorter than the other phases of the cycle, with an average duration estimated tobe 7 months.
Source: Piraeus Bank Research, Bloomberg 22March 2015
-5
0
5
10
15
20
25
DP PB growth BV growth Est Return
pe
rce
nt
UpturnFinally, when the economy moves into the upturn phase:
Ø Market expectations and risk appetite return to positive levels, consequently boosting total returns.
Ø The book value per share remains relatively stable, signalling greater uncertainty about the economic environment for decision making at the corporate level.
23March 2015
We rely entirely on the use of the accounting identities of the supply-side valuation model for stock returns. More specifically, this methodology decomposes the return performance of a stock or an equity index in the individual components of the dividend yield and capital gains:
Where,
R (t): Equity Index Simple Return at t
P (t): Price at time t
D (t): Annual sum of dividends
B (t): Book value per share
DP (t): Index Dividend-Price Ratio
g (bvps): Rate of Change in Book Value per Share
g (p / b): Rate of Change in Price to Book Value Ratio
( )
( ) ( )( )
( ) ( ) ( )
)/()()()(/)(
)(/)()(
)()(
)()()(
)()(
)()(
)()(
)()(
)()(
)(_)(_)(
bpgbvpsgtDPtBtPtBtP
tBtBtDP
tPtPtDP
tPtP
tPtD
tPtP
tPtD
tPtP
tYieldDividendtGainCapitaltR
+´+´+=--
´-
´+=
-´+=
-´+
-=
-+
-=
++=+
úúû
ù
êêë
é
111111
1
11
11
11
11
24March 2015
Source: Piraeus Bank Research, Bloomberg
-100%
-50%
0%
50%
100%
150%
200%
6,45,64,83,93,12,31,40,6
Years
Expansion
DP BV growth PB growth
-120%
-100%
-80%
-60%
-40%
-20%
0%
20%
40%
60%
80%
6,45,64,83,93,12,31,40,6
Years
Downturn
DP BV growth PB growth
-200%
-150%
-100%
-50%
0%
50%
6,45,64,83,93,12,31,40,6
Years
Recession
DP BV growth PB growth
-60%
-40%
-20%
0%
20%
40%
60%
80%
100%
6,45,64,83,93,12,31,40,6
Years
Upturn
DP BV growth PB growth
Disclaimer: This note constitute an investment advertisement, is intended solely for information purposes and it cannot in any way be consideredinvestment advice, offer or recommendation to enter into any transaction. The information included in this note may not be construed as suitableinvestment for the holder, nor may it be considered as an instrument to accomplish specific investment goals or relevant financial needs of the holderand may neither be reckoned as a substitute to relevant contractual agreements between the Bank and the holder. Before entering into any transactioneach individual investor should evaluate the information contained in this note and not base his/her decision solely on the information provided. Thisnote cannot be considered investment research and consequently it was not compiled by Piraeus Bank according to the requirements of the law that areintended to ensure independence in the sector of investment research. Information comprised in this note is based on publically available sources thatare considered to be reliable. Piraeus Bank cannot be held accountable for the accuracy or completeness of the information contained in this note.Views and estimates brought forward in this note represent domestic and international market trends on the date indicated in the note and they aresubject to alteration without previous warning. Piraeus Bank may also include in this note investment research done by third parties. This informationis not modified in any way, consequently the Bank cannot be held accountable for the content. The Piraeus Bank Group is and organization with aconsiderable domestic and international presence, and provides a great variety of investment services. In cases where conflicts of interest issues shouldarise while Piraeus Bank or the rest of the companies of the group provide investment services in relation to the information provided in this note,Piraeus Bank and the companies of the Group should be underlined that (the list is not exhaustive): a) No restrictions apply in dealing for own account,or with regards to trading in relation to portfolios managed by Piraeus Bank or companies of its group before the publication of this note, or withregards to trading before an initial public offer. b) It is possible that investment or additional services are provided to the issuers included in this noteagainst a fee. c) It is possible that Bank or any of its subsidiaries participate in the share capital of any of the issuers included in this note or may attractother interests financial or not from them. d) The Bank or any of its subsidiaries may act a market maker or an underwriter for any of the issuersincluded in the note. e) Piraeus Bank may have issued similar notes with different or incompatible content with the content of this note. It should beexplicitly noted that: a) Figures refer to past performances and past performances do not constitute a safe indication for future performances. b) Figuresconstitute simulation of past performances and they are not a safe indication of future performances. c) Any projections or other estimates are not safeindications for future performances. d) Taxation treatment of information provided in this note may differ according to the rules that govern eachindividual investor. Therefore the holder should seek independent advice in relation to taxation rules that may affect him/her. e) Piraeus Bank is notunder any obligation to keep data and information provided herein updated.
25March 2015