Long-Term Contrarian Profits In The Middle East Market Indices
AL albayt University
July 30, 2015
Research Journal of Finance and Accounting, Vol. 6, No. 16, 2015
This paper examines whether there is an existence of a long-term contrarian profits at the Middle East (ME) market indices. This paper shows strong evidence for the long-term contrarian strategy in the Middle East indices. The result of this study demonstrates that the long-term contrarian profits for the Middle East markets can’t be explained by two-factor model. In spite of whether winners are smaller or larger than losers, there are long-term abnormal profits. Finally, the findings in this paper suggest that the long-term contrarian profits may be stronger and more enveloping than is usually understood.
Long-Term Contrarian Profits In The Middle East Market Indices – Introduction
In their international indices study of long-term contrarian strategy, Malin and Bornholt (2013) comprehensively examine the performances of the long-term abnormal returns to contrarian investment strategies. They reveal empirical evidence on the long-term contrarian profits in the international market indices. Their study shows that the developed markets results provide statistically significant long-term contrarian profits for all holding months, while the developing markets results produce economically important.
By employing the Malin and Bornholt (2013) methodology, this study aims to investigate whether there is a long-term contrarian strategy applied to the Middle East market indices. Within the context, the role of two-factor model will be investigated.
The remainder of the paper is arranged as follows. Next section reviews the previous results in the literature while Section 3 presents the sources of the data and discusses the empirical methodology. Section 4 provides the results for both raw and risk-adjusted returns to the zero investment contrarian strategy, the twofactor model is also applied as part to the analysis of long-run profitability. Finally, Section 5 concludes the paper.
Following DeBondt and Thaler (1985) who documented the long-term contrarian effect, using the same data, DeBondt and Thaler (1987) re-examined stocks that earned extreme long-term gains or losses. They constructed portfolios of the 50 most extreme losers and 50 most extreme winners. Their finding confirmed the evidence documented by DeBondt and Thaler (1985) that the overreaction hypothesis plays an important role in long-term return reversals after controlling for both risk and size. Thus, the differences in risk and firm size cannot explain the winner-loser effect. The results also showed that the portfolios of losers outperformed the portfolios of winners by about 30% over the following five years. DeBondt and Thaler (1987) examined the seasonal pattern of return as well showing a January effect which was related to the reversal effect. In addition, there was a negative relationship between the excess returns for the winners and the excess returns prior to December, result which was related to the capital gains tax “lock-in” impact.
At country index level, Richards (1997) investigated the comparable winners-losers reversal in 16 national markets for the period 1970-1995. Two primary methodologies have been used to assess the risk of a contrarian strategy. The first is to measure the covariance of risk and return exposures of the winners and losers portfolio. The second is to measure whether loser’s portfolios tend to underperform the winner’s portfolios either in recession periods or during large declines in the world index. The analysis confirms the finding of DeBondt and Thaler (1985; 1987) that reversals are strongly significant around the 3-year horizon. Consequently, abnormal returns have averaged at least 6% annually during the period from 1970-1995. However, the important result is that the reversals do not reflect risk differentials. In other words, the difference between test-period returns of prior winners and losers are statistically insignificant either in their performance in adverse states of the world or in terms of their standard deviations. In addition, the results show that smaller markets are more affected by reversals than are larger markets.
Exploring the source of contrarian profits, Barberis, Shleifer and Vishny (1998) provide an economic model of investor sentiment simulated by psychological evidence of how investors form beliefs and expectation of future earnings. This model is consistent with Griffin and Tversky’s (1992) idea that concentrates on making forecasts. This model produces an underreaction and overreaction to a wide range of parameter values. They point out that the three-factor model can explain the overreaction, but not underreaction evidence. In addition, Barberis, et al. (1998) presented evidence that the investors can achieve abnormal returns during the underreaction and overreaction periods without bearing extra risk, and this is considered a major challenge to the efficient market theory. Barberis, et al. (1998) confirmed the finding of DeBondt and Thaler (1985) that long-term reversals can be attributed to traders that finally do overreact. Barberis, et al. (1998) and other studies such as DeLong, Shleifer, Summers and Waldmann (1993) and Daniel, Hirshleifer and Subrahmanyam (1998) assume that this long-term return reversal can be explained by long-term correction after investor overreaction. This model relates to both the important behavioral heuristics, known as representativeness bias and conservatism bias.
Representative bias means that investors become too pessimistic (optimistic) about companies with a series of bad (good) news. Concentrating on industry news rather than firm-specific news leads investors to extrapolate performance too far for the industry as a whole producing long-run reversals in industry returns thus supporting Tversky and Kahneman’s (1974) findings. The conservatism bias is also present when investors are more conservative in renewing their previous ideas when new industry information arrives, thus, generating under-reaction in industry prices to public information. This confirms the phenomenon documented by Edwards (1968) that conservatism supports the underreaction evidence.
See full PDF below.