International Investors' Exposure To Risk In Emerging Markets

We examine the empirical differences in emerging market betas (β's) taken across four major currencies (U.S. dollar, British sterling, Japanese yen, and German mark) where the β's are either mean‐variance or mean‐lower partial moment β's. The mean variance β's are found to be statistically similar to lower partial moment β's in most cases, which suggests they are robust to nonnormality in the data. The difference between the two β's has become less significant in recent years as emerging markets have become more stable. Furthermore, evidence is presented that β's obtained from both risk measures and calculated from returns denominated in different currencies have the same ordinal association. This shows the primacy of local risk over foreign exchange risk. We conclude that international investors can continue to use the mean‐variance β in assessing risk in emerging markets, although investors should not give it a conventional equilibrium interpretation.