Predictors of Stock Returns in Emerging Equity Markets

We build a prediction model for a set of Emerging Markets and the US market and we evaluate its ability to forecast equity risk premia. Our models include traditional global and local variables as the dividend yield or a credit spread and we include some variables that has been rarely been used as predictor of emerging equity returns: the implied volatility of the US market. We study the period 1998-2006 for di erent forecasting horizons and consider di erent sub-samples. We nd some cointegration relation, using EG-test, that help us build our forecasting models. Combining this way the predicting variables we can explain more than 30% of the variation of the US market risk premia at a quarterly basis. And for some Emerging markets R2 is close to 23% for the whole period of study. Other preliminary results include the tests of stationarity for all the variables and we nd evidence of cointegration among some of the variables.