Model Combination and Stock Return Predictability

Bayesian Model Averaging (BMA) has recently been discussed in the financial literature as an effective way to account for model uncertainty. In this paper we compare BMA to a new model uncertainty framework introduced by Yang (2004), called Aggregate Forecasting Through Exponential Reweighting, which has as well a Bayesian interpretation, but enjoys several attractive features not shared by BMA. The AFTER algorithm has nice theoretical properties if the true model does not belong to the class of considered models and can easily incorporate 'stylized facts' of financial data in the weighting scheme, such as time-varying volatility and fat tails. Most importantly, the determination of model weights in AFTER is based on pseudo out-of-sample performance and not on within sample criteria as it is the case for BMA. This seems rather attractive from an investment perspective.