Momentum trading and Business Cycle Risk: Evidence from BRIC Countries

1. Introduction. BRIC (Brazil, Russia, India and China) countries are growing at an alarming rate. This growth can be attributed to a number of factors including globalisation, financial liberalisation which has led to an increase in cross-border capital flows, technological developments and the internet. These countries are forecast to witness tremendous growth in the years ahead. The alarming growth of BRIC countries has attracted investors in search of suitable environments for portfolio diversification to consider BRIC countries as potential destinations for diversifying their portfolios. This paper presents a proposal to study the link between business cycles and momentum trading in the BRIC stock markets. The paper aims at understanding how business cycle risk affects momentum profits in BRIC countries. The study also seeks to provide an understanding of how momentum profits are affected by firm specific characteristics such as firm size and book-to-market ratios in BRIC countries. 2. Objectives of the study The objective of the study is to determine the impact of business cycle risk on momentum profits and thus momentum trading in BRIC countries. Research Questions The study aims at answering the following questions: Are there momentum profits in the stock markets of BRIC countries If so, what is the impact of Business Cycle risk on these profits What are the regulatory implications of momentum profits in BRIC countries Significance of the Study The study is significant to market regulators in that it will enable them design regulatory requirements aimed at reducing inefficiencies in BRIC stock markets thereby increasing their ability to attract capital. The study will also help foreign investors to gain more confidence in BRIC countries. Finally, the study will serve as a reference point for future researchers interested in conducting research on momentum profits. 5. Literature Review. A momentum trading strategy is a trading strategy that is designed based on past performance. The trading strategy is based on the assumption that â€Å"history will repeat itself†. A momentum trading strategy is therefore a strategy, which assumes that the return performance will persist in the medium term (Signos and Chelley, 1994). Momentum profits were first observed by Jegadeesh and Titman (1993). Accordingly, the study observed that stocks that performed well in a previous period also performed well in the current period, while those that performed poorly in the previous period also performed poorly in the current period. This means that a trading strategy that went long on previous winners while shorting previous losers would result in positive abnormal returns. In particular Jagadeesh and Titman (1993) observed the realisation of positive abnormal returns of 1 percent with the momentum strategy. In addition, a number of other studies have observed significant positive abnormal returns with the momentum trading strategy (e.g., Moskowitz and Grinblatt, 1999; Jegadeesh and Titman, 2001; Liu et al. (1999), Hong and Tonks, 2003; Gregory et al., 2001; Griffin et al., 2003; Gregory et al. 2001; Rouwenhorst 1998). The implication of the existence of such a Band Wagon (money making strategy) is that markets were not efficient. According to the weak- and semi-strong form efficient market hypotheses, all information available to the general public is already reflected in stock prices. This means that investors cannot realise superior risk adjusted returns by adopting a particular trading strategy such as the one proposed by momentum trading (Ross et al., 1999; Bodie et al., 2007). Attempts to attribute this finding to inefficient markets have been opposed by Fama and French (1993, 1995, 1996) who argued that observing momentum profits cannot be attributed to inefficient capital markets. Rather the single factor capital asset pricing model (CAPM) has been criticised for not being able to properly explain the variability of the cross-section of stock returns. This model suggests that stock market returns depend on a single factor (i.e., the return on the market portfolio). However, Fama and French (1993, 1995, 1996) contest this view and argue instead that stock returns could be explained by additional factors such as the book-to-market ratio and firm size. A three factor model is therefore proposed which takes into account the impact of size and book-to-market ratio and is found to perform better than the single factor CAPM (Fama and French, 1993, 1995, 1996). In addition, the three factor model was extended to a four-factor model to include a momentum factor w hich measures the difference between the return on portfolios of stocks that performed well in the previous period and the return on portfolios of stocks that performed poorly. Including a momentum factor in the three-factor model thus making it a four-factor model enabled the model to be able to explain the momentum profits observed in Jagadeesh and Titman (1993) and the other studies identified in the Literature. In summary, Fama and French argue that anomalies such as those observed in momentum trading cannot be attributed to inefficiencies in capital markets. Rather they should be attributed to inadequacies in the models that are used in explaining the cross-section of stock returns. Other explanations have been offered for the observation of momentum profits. According to behavioural finance theorist, momentum profits are a result of slow movement of information. Behavioural finance theorists are against market efficiency theorists who argue that information is rapidly reflected in stock prices. Among behavioural theorists, Hong and Stein (1999) argue that momentum profits can be attributed to slow diffusion of information across interested investors. This means that some investors receive information about stock prices earlier than others and as such appropriate action faster than others. By so doing, investors who have quick access to information are capable of making superior abnormal returns while those who do not have quick access to information tend not to make superior risk-adjusted returns by using such information as a basis of trading. Barberis et al. (1998) argues that momentum profits can be attributed to overreaction or underreaction of stock prices to news. The explanation from behavioural theorists conflict with those of Fama and French because behavioural theorists also suggest that there is nothing like an efficient market. Given the conflict between behavioural theorists and proponents of market efficiency, alternative explanations have been provided by recent studies. These studies argue that momentum profits are influenced by business cycle variables (e.g., Antoniou et al., 2007; Liew and Vassalou, 1999). Contrary to this view Griffen et al. (2002) in a study examining the link between business cycle variables and momentum profits across many countries argue that momentum profits are not a function of business cycle variables. While many studies have investigated the relationship between business cycle variables, most of these studies focus on developed markets with very little attention paid to emerging markets such as those of BRIC countries. Given the increasing role that BRIC countries play in the global economy, it is important to understand whether there are momentum profits in these countries as well as the role that business cycle risk has on momentum profits. This study is therefore a positive step toward contributing to the literature on momentum profits and business cycle risk by extending previous studies to stock markets in BRIC countries. 5. Research Methods This study will employ an econometric model to study the relationship between momentum profits and three sets of variables: (i) business cycle variables; (ii) firm specific variables (iii); and behavioural finance variables. The relationship between momentum profits and these variables can be represented using the following econometric model: (1) Where is a measure of the momentum profit of country i at in year t; is a vector of firm specific variables; is a vector of the past cumulative raw returns; and are the sensitivities of the momentum profits to changes in firm-specific variables and past cumulative returns respectively. The magnitude of the effect of these variables will be determined by testing the significance of the parameters at the 5% level of significance. In order to study business cycle variables, a model was developed by Chordia and Shivakumar (2002) and later extended by Antoniou et al. (2007). The model is an econometric model which establishes the relationship between momentum profits and business cycle variables. The model can be stated as follows: Where is the return (inclusive of dividends) of firm i in month t, BC is a vector of j (j=1-6) macroeconomic variables representing business cycle variables (DY, Rf, TERM, DEF, FX, and GDP), and is the error term of stock i in month t. DY is the dividend yield; Rf is the risk-free interest rate; DEF is the premium for default risk premium which is estimated as the difference between the yield on long-term corporate bonds and the yield on long-term government bonds; The term spread (TERM) is the difference between the yield on long-term government securities minus the yield on short-term government securities; FX is the foreign exchange rate; and GDP is the change in GDP (Antoniou et al., 2007). As earlier mentioned, stock returns depend on two factors: market factors and firm-specific factors. There is a trade-off relationship between the manner in which each group of factors affect stock returns. That is the higher the impact of firm-specific factors, the lower will be the impact of market factors and vice versa (Antoniou et al., 2007). To estimate equation (1) equations 3 has to be estimated and its parameters used as inputs to equation (2). After estimating equation (2) its parameters can then be used as inputs to equation (1). In this study, both time-series and cross-sectional regressions are used. Cross-sectional regressions are preferred over time series regressions because they help to avoid data-snooping biases which tend to occur in time-series regressions. In the time-series regressions, individual stocks are used which help to reduce the degree of loss of information that tends to occur when portfolios are used. Using first-pass time series regression, which allows the parameters to also fluctuate with firm-specific variables. The firm-specific factors include firm size and book-to-market ratio. The first-pass time-series regression can be stated as follows; is the return on firm i at time t, BC is a the vector of business cycle risk variables identified earlier, FF (Fama and French factors) are the firm-specific variables. Once equation (3) has been estimated, the parameters will be used as inputs to the second pass regression equation (4) below: Where is the output of equation (3). It is the unexplained variation from equation (3). These include the intercept coefficient and the residual term (+) of the regression equation (3); is a vector of firm characteristics, which include firm size and book-to-market ratio for security i at time t. represent the three sets of past cumulative raw returns (for m=1-3) over the second through third (RET 2-3), fourth through sixth (RET 4-6) and seventh through twelfth (RET 7-12) months prior to the current month t. (Antoniou et al. 2007). 6. Data Stock price data for stocks in the BRIC countries will be retrieved from the Thomson Financial Datastream Database. Data on dividend yields will also be retrieved from this database. The database also reports data on exchange rates. GDP, interest rate and exchange rate data will be retrieved from the IMF International Financial Statistics (IFS) database. Stock price data will be used to calculate the monthly return for each stock over the 60 monthly holding periods from January 2007 to December 2011. The returns will be used as inputs to the first-pass regression. References Griffin, John M., Martin, J. Spencer and Ji, Susan, â€Å"Momentum Investing and Business Cycle Risk: Evidence from Pole to Pole† (March 18, 2002). AFA 2003 Washington, DC Meetings; EFA 2002 Berlin Meetings Presented Paper. Available at SSRN: http://ssrn.com/abstract=291225 or DOI: 10.2139/ssrn.291225 Antoniou A., Lam H. Y.T., Paudyal K. (2007). Profitability of momentum strategies in international markets: The role of business cycle variables and behavioural biases. Journal of Banking & Finance volume 31, issue 3, pp. 955-972. Liew, Jimmy K.yung Soo and Vassalou, Maria, (1999). â€Å"Can Book-to-Market, Size, and Momentum Be Risk Factors That Predict Economic Growth?† Available at SSRN: http://ssrn.com/abstract=159293 or DOI: 10.2139/ssrn.159293 Rouwenhorst, K.G. (1998). International momentum strategies, Journal of Finance 53, pp. 267–284. Wu, X. (2002). A conditional multifactor analysis of return momentum, Journal of Banking and Finance 26 (2002), pp. 1675–1696 Jegadeesh N., Titman S. (1998). Returns to buying winners and selling losers: Implications for market efficiency, Journal of Finance 48, pp. 65–91. Barberis N., Shleifer A., Vishny R.W.(1998). A model of investor sentiment, Journal of Financial Economics 49, pp. 307–343. Fama E.F., French K.R. (1996). Multifactor explanations of asset pricing anomalies, Journal of Finance 51 (1996), pp. 55–84. Hong H., Stein J.C. (1999). A Unified Theory of Undereaction, Momentum Trading, and Overreaction in Asset Markets. Journal of Finance. Vol. 6, pp 2143-2184 Chelley-Steeley, Patricia and Siganos, Antonios, (2004). â€Å"Momentum Profits in Alternative Stock Market Structures†. Available at SSRN: http://ssrn.com/abstract=624583