The Validity of the CAPM in the Russian and Kazakhstan Capital Market with Conditional Meansemivariance and Higher-order Moments Specifications Tamara Teplova, Professor, Head of Master Program “Financial Markets”, State University - Higher School of Economics, Evgeniya Shutova, State University - Higher School of Economics, Russia. EUROPEAN FINANCIAL MANAGEMENT SYMPOSIUM Beijing ,2011
24
Embed
The Validity of the CAPM in the Russian and Kazakhstan Capital Market with Conditional Meansemivariance and Higher-order Moments Specifications
EUROPEAN FINANCIAL MANAGEMENT SYMPOSIUM. The Validity of the CAPM in the Russian and Kazakhstan Capital Market with Conditional Meansemivariance and Higher-order Moments Specifications . Tamara Teplova , Professor, Head of Master Program “Financial Markets”, - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
The Validity of the CAPM in the Russian and Kazakhstan Capital Market with Conditional
Tamara Teplova,Professor, Head of Master Program “Financial Markets”,
State University - Higher School of Economics, Evgeniya Shutova,
State University - Higher School of Economics, Russia.
EUROPEAN FINANCIAL MANAGEMENTSYMPOSIUM
Beijing ,2011
The Laboratory of Financial Markets Analysis http://fmlab.hse.ru
CAPM Modifications for Emerging Markets. Analysis of Alternative
CAPM Beta Specifications
Exchange Rate Risk and Stock Price Dynamics. Evidence from Top
Russian Companies
Measuring Liquidity in Stock and Bond Market. Is Liquidity
Risk Offset by Excess Return?
CDS Spreads as Indicators for the Government and Corporate Bond
Market Performance
The popularity of the The popularity of the CAPMCAPM in a in a period of crisis has not decreasedperiod of crisis has not decreased• The studies based on the surveys of over 11 thousand US The studies based on the surveys of over 11 thousand US
chief financial directors conducted by the Duke University and chief financial directors conducted by the Duke University and the CFO Magazine have shown that nearly 75% of the CFO Magazine have shown that nearly 75% of respondents used the CAPM framework to take decisions in respondents used the CAPM framework to take decisions in 2008 and 20092008 and 2009 . .
• Source: Graham, John; Campbell Harvey, Equity risk Source: Graham, John; Campbell Harvey, Equity risk premium amid a global financial crisis, Evidance from the premium amid a global financial crisis, Evidance from the Global CFO Outlook survey 2009. SSRN WP; Graham, J. R., Global CFO Outlook survey 2009. SSRN WP; Graham, J. R., C. R. Harvey, 2009, The CFO Global Business Outlook: 1996-C. R. Harvey, 2009, The CFO Global Business Outlook: 1996-2009. http://www.cfosurvey.org.2009. http://www.cfosurvey.org.
Whether the CAPM beta completely measures systematic risk??Critical assumptions of CAPM:Critical assumptions of CAPM: quadratic investor’s utility function quadratic investor’s utility function the the assumptionassumption of the normal distribution of returns of the normal distribution of returns
the mean and variance suffice to describe the distribution completely
Evidence from emerging markets:Evidence from emerging markets: no simultaneous symmetrical and normal distribution of no simultaneous symmetrical and normal distribution of
the expected returnthe expected return (high kurtosis,(high kurtosis, asymmetry) asymmetry) law liquidity of most equities serious problems such law liquidity of most equities serious problems such
as understating the beta which is calculated using the as understating the beta which is calculated using the regression methodregression method
Purpose of investigation
development of alternative CAPM specifications
The replacement the original beta by downside systematic risk measures.
Incorporating higher-order moments.
Using conditional CAPM instead of unconditional construction.
Methodology
Two-step procedure Fama and MacBeth (1973)
1.The estimation of risk factors of each individual stocks (time-series regressions).
2.The estimation of the cross-sectional relationship between the mean return of assets and estimated risk factors.
Downside measures of risk Beta of Harlow and Rao (1989) with benchmark
equal to mean equal to zero
Beta of Estrada (2002) with benchmark equal to mean equal to zero
Gain-Loss Spread (GLS) of Estrada (2008)
20,min
0,min(
mm
mmiiHRi RE
RRE
20,min
0,min(0,min
mm
mmiiEi RE
RRE
M
t
N
ittLG LGTEEGLS
1 1
)/1(
Higher-order moments as Higher-order moments as SSystematic ystematic RRisk isk FFactoractorss Systematic skewness (co-skewness, third moment,
gamma)
The corresponding measure of downside co-skewness risk (HR-gamma) to HR beta:
The measure of systematic downside co-skewness risk (E-gamma):
Conditional Capital Asset Pricing ModelsImpact evaluation of general market conditions on the adequacy of asset pricing models The conditional four-moment model САРМ:
imtimtimtimtimtimttit kkkkkkR )1()1()1( 6543210
1k0)( ftmt RR 0)( ftmt RR 0kwhere when and where
Literature review
Downside Higher-order Conditional framework moments construction
Bawa&Lindenberg(1977), Arditti (1971), Pettengill, Sundaram Harlow & Rao (1989), Francis (1975) , and Mathur (1995),Estrada (2002, 2007) Kraus&Litzenberger(1976), Galagedera&Maharaj (2004)
The object of research
?? KazakhstanKazakhstan Russia Russia
49 49 Russian companiesRussian companies 10 Kazakhstan companies 10 Kazakhstan companies Source of data: Source of data: MICEXMICEX and KASE and KASE Sample period: Sample period: January 2006 – December 2010January 2006 – December 2010 FrequencyFrequency ofof data: weekly returns.data: weekly returns. WeeklyWeekly returns are calculated as returns are calculated as::
100*5
5
t
tti P
PPr
ASIAASIA
Russian and Kazakhstan stock exchanges Russia Kazakhstan
RTS MICEX KASE
RTS MICEX KASE Currency USD Rubles Tenge
Number of listings >500 696 125
Market capitalization
USD 1072,332 bln USD 966 bln USD 60.7 bln
Trading volume USD 8601 bln USD 2110 bln
USD 206.5 bln
Dynamics of KZKAK and MICEX index for the period: 01/2008-12/2010
KZKAK indexKZKAK index
MICEX indexMICEX index
30-day volatility of KZKAK and MICEX during 01/2008-12/2010
Top 10 summary statistics of Russian companies: January 2010 –December 2010
BANK CENTERCREDIT CCBN KZ Equity 38 0,36 3,62 1,64 0,44
Testing hypothesis
1. Downside risk measures are better for explaining cross-sectional return variations than traditional beta especially during the crisis.
2. The inclusion of higher-order moments (the gamma coefficient of systematic asymmetry and the delta coefficient of systematic kurtosis) may contribute to the explanatory power of one- and-multi-factor models.
3. Co-skewness plays a more important role in explaining Russian returns while co-kurtosis is consistently influential for Kazakhstan stock returns due to the fact that Russian stocks are more skewed but less leptokurtic than Kazakhstan stocks.
Are traditional and downside beta good measures of risk?
1. The explanatory power does improve in terms of a higher coefficient of determination if the traditional CAPM beta coefficient is replaced by one-sided risk measures.
2. The zero rate of return benchmark, the models display better explanatory power.
3. The downside beta specification of Harlow and Rao (1989) proves to be more efficient in explaining cross-sectional return variations than that of Estrada (2007).
4. Gain-Loss Spread is the best measure of risk among analyzed factors in downside constructions for Kazakhstan market.
5. refutation the hypothesis that the inclusion of higher-order moments may better explain cross-sectional return variations
What next? Future InvestigationsWe postulate that the differences in results are related to the underlying firm characteristics of the companies in the two indices taking into account investors’ expectations and size of companies using portfolio formation via ranking by BV/MV and size of companies