Predictability, non-normality, volatility and stability in emerging Asian markets

In this research we investigated monthly excess returns in six emerging Asian stock markets i.e India, Indonesia, Malaysia, Pakistan, Philippines, and Thailand over the relevant risk-free rates for possible existence of predictable components in these countries\u27 stock markets over the relevant risk free rates. We modeled excess returns in these markets using non Gaussian state space or unobserved component models that encompass non normality to account for fat tails and conditional heteroskedasticity to account for time varying volatility that may be present in the excess return series. Our results show that statistically significant persistent predictable components exist in India, Malaysia, and Pakistan with excess returns at 5 per cent and in Thailand at 10 per cent level of significance. Likewise, our results also show an evidence of statistically significant non-normality and time varying volatility in excess return series for all the countries studied except India where market appears to have normal behaviour. Moreover, leverage effect being insignificant in all the stock markets, stability does not appear to be an issue except for Pakistan where stock price volatility does not appear to show a regular trend pattern over time. The efficiently estimated excess returns ranges between 0.6 per cent per month for Malaysia to 1.6 per cent per month for Indonesia