CAPM models with risk-neutrality and loss-aversion, consistency and predictability on emerging stock markets of BRIC (Brazil, Russia, India and China)
CAPM models with risk-neutrality and loss-aversion, consistency and predictability on emerging stock markets of BRIC (Brazil,
Russia, India and China)
Contents
Introduction Theoretical aspects of research• Classical CAPM model• Betting against beta model
Empirical tests. Results• Correlation tests• Beta-based portfolios
Conclusion
Introduction
Idea of market risk factor • Fama-French 3-factor model• FCF-models• WACC model etc.
CAPM deals with 2 factors• Idiosyncratic risk (alpha)• Market risk (beta)
Microstructure of developing markets differs from those of developed ones
Introduction
Aim of the thesis: analysis of CAPM model consistency and predictability in application to emerging markets stock returns
Subject of the study: application of CAPM model (and/or inverse CAPM model) as a tool for making successful and efficient investment decisions
Object of the study: stocks traded on emerging markets of BRIC
Introdcution
Hypothesis 1 - beta estimations have descriptive power for stock returns on emerging stock markets of BRIC
Hypothesis 2 - CAPM model implications are consistent on stock markets of BRIC-countries
Hypothesis 3 - Low beta stocks provide highest returns across stock market
Theoretical aspects
Classical CAPM model By Sharpe, Lintner and Mossin in 1960s General formula:
𝐸(𝑟𝑖)=𝑟𝑓+𝛽𝑖( (𝐸 𝑟𝑚)−𝑟𝑓)
Higher beta coefficient - greater market risk to which the asset is subjected
During market gains, high beta assets provide higher returns, in comparison to low-beta assets
Theoretical aspects
Betting against beta model Frazzini & Pedersen (2010) and Baker,
Bradley & Wurgler (2011) Deals with behavioral biases and agent
investment confines Leverage constraints and margin
requirements => overinvestment in high-beta stocks, lowering their returns
Stocks with lower CAPM market risk estimations provide higher returns
Empirical tests
Data analyzed A decade of years starting from 2001 (covering
“dot.com” bubble burst & recent financial crisis) Data acquired via Thomson Reuters DataStream Weekly stock returns (to avoid non-synchronous
trading) Basis period equals one year (52 weeks) Risk free rates: 90 day interbank rate (Russia),
overnight financial rate (Brazil), 91 day treasury bill (India) and 3 months relending rate (China)
Benchmark indices: RTS &MICEX (Russia), MSCI stock indices (Brazil, China, India)
Empirical tests
Data analyzed China - 793 stocks on A-shares
market & 53 stocks on B-shares market, India – 255 stocks, Russia – 133 and Brazil – 134
General formula applied:
𝑟𝑡−𝑟𝑡𝑓= + (𝛼 𝛽 𝑟𝑡𝑚−𝑟𝑡𝑓)+𝜀
OLS & GLS regressions, beta-based portfolios
Empirical testsCorrelation tests. China (A-shares)
Empirical testsCorrelation tests. China (A-shares). Positive outcomes
Empirical testsCorrelation tests. Brazil
Empirical testsCorrelation tests. Brazil. Negative outcomes
Empirical testsBeta-based portfolios. China
Empirical testsBeta-based portfolios. Brazil
Conclusions
Logic of original CAPM model didn’t hold true for every BRIC country
Russian and Indian stock markets showed mixed results Chinese market almost perfectly fit the logic of original
CAPM model (remark: model kept true only for a specific part of the market –A-shares )=> original CAPM model could be successfully applied when making investment decisions in regard to Chinese A-shares
BAB model to be consistent when applied to shares traded on Brazilian stock market
Low-beta stocks overperform on Brazilian stock market BAB model can be applied in constructing investment
strategy for Brazilian market
Questions are welcome