پاداش ها برای ریسک نامطلوب در بازارهای آسیایی

Distributional properties of emerging market returns may impact on investor ability and willingness to diversify. Investors may also place greater weighting on downside losses, compared to upside gains. Using individual equities in a range of emerging Asian markets, we investigate the potential contribution of downside risk measures to explain asset pricing in these markets. As realized returns are used as a proxy for expected returns, we separately examine conditional returns in upturn and downturn periods, in order to successfully identify risk and return relationships. Results indicate that co-skewness and downside beta are priced by investors. Further testing confirms a separate premium for each measure, confirming that they capture different aspects of downside risk. Robustness tests indicate that, when combined with other risk measures, both retain their explanatory power. Tests also indicate that co-skewness may be the more robust measure.

The purpose of this paper is to explore whether measures of downside risk contribute towards an explanation of the risk/reward relationship for individual shares in emerging Asian markets. The notion that measures such as co-skewness and downside beta should matter to investors is well established in the literature. Kraus and Litzenberger (1976) show that utility functions with non-increasing absolute risk aversion imply a preference for positive skewness. Huang and Litzenberger (1988) demonstrate that the risk premium on assets will depend on their co-skewness, with investors preferring assets with positive co-skewness. The rational Disappointment Aversion (DA) utility function attributed to Gul (1991) implies that investors display a larger aversion to losses relative to the attraction for gains. Ang et al. (2006a) demonstrate how, in a DA utility framework, cross sectional asset pricing will incorporate a premium for downside risk measures such as downside beta. Emerging markets merit separate examination, as there is evidence that asset returns exhibit very high volatility and are not normally distributed (Bekaert and Harvey, 1997). Bekaert et al. (1998) identify significant skewness and kurtosis in emerging market returns, and they observe the persistence of skewness over time. Co-skewness of returns is our primary measure of downside risk. Using individual share data, Harvey and Siddique (2000) find that co-skewness has explanatory power for share returns, after allowing for other established explanatory factors. Downside beta is a further indicator of downside risk. Ang et al. (2006a) provide a detailed empirical examination of the explanatory power of downside beta for individual shares in the US market. They show that the shares which co-vary strongly with the market during market downturns do have higher average returns. Pedersen and Hwang (2007) also demonstrate that downside beta will explain a higher proportion of individual UK share returns than will beta alone. The issue of downside risk in emerging markets has already been addressed in the literature. Studies so far have examined this issue at aggregate market level, but not at the individual firm level. Estrada (2002) uses market indices to provide evidence on the explanatory power of downside beta. Using a measure developed from the comparison of investment returns with market portfolio returns, when each is below their respective means, he reports stronger results than for beta. When compared with downside beta, Galagedera and Brooks (2007) find that co-skewness is the better explanatory variable of emerging market monthly returns. Galagedera (2009) also reports that, when compared with beta and downside beta, co-skewness is a better measure of risk. However, when assessing developed market indices, he finds that both downside risk measures perform poorly when compared to beta. Using daily data from emerging Asian markets, we present a series of empirical examinations of whether downside risk is independently priced in cross-section. We exclude Singapore and Hong Kong, as they would not normally be categorized as emerging markets. We also limit our investigation to eight markets, as remaining markets in this region are so small that they have relatively few actively traded shares. For our investigation, we compare realized returns of individual companies with individual risk measures computed in each market. As realized returns are a proxy for expected returns, we analyze returns during market downturns separately from market upturns. Our results offer a significant contribution to the unraveling and understanding of risk measures in emerging markets and the manner in which investors are rewarded for assuming those risks. In outline, we find that investors in emerging markets are clearly rewarded for exposure to both co-skewness and downside beta. Control tests confirm a separate premium for each, indicating they capture different aspects of downside risk. When combined with other risk measures, both co-skewness and downside beta retain explanatory power. There is however some evidence that co-skewness may be a more robust measure, as it tends to retain greater significance. Our paper proceeds as follows. In Section 2 we outline details of the markets included in the study, we also address the issue of thin trading and share selection. Section 3 describes the research methodology. A separate sub-section models expected relationships between exposure to risk measures and investor returns. Section 4.1 presents an assessment of relationships between individual downside risk measures and returns, Section 4.2 offers the results of control tests on the potential impact of interrelationships between risk measures, and Section 4.3 contains results of regression tests on the explanatory power of risk measures, when in combination. Section 5 concludes.

We find evidence of relationships between downside risk measures and returns on individual shares in emerging Asian markets. Our measures are negative co-skewness and downside beta. In a separate investigation of upturn and downturn markets, we find that both are priced. In upturn markets, exposure to negative co-skewness and to downside beta is rewarded with larger returns. These exposures are penalized with greater losses during downturns. If upturn and downturn data is combined, average returns obscure the pricing relationship. We do not however believe that this implies major differences between emerging and developed markets, as previous research did not include a separate analysis of upturns and downturns. Our findings are not directly comparable. Control tests confirm that each downside risk is separately priced. They do not capture similar effects in company returns. When combined with other risk measures, we find that downside beta has reduced significance. There is no evidence of a reduced significance for co-skewness. During upturn years, we estimate an average premium of 7% for negative co-skewness risk, and an average premium of 12% for downside beta risk. We estimate losses of a similar scale during downturns.