The Economic Benefit of Forecasting Market Components for Mean-Variance Investors

Abstract

Existing studies focus on variables’ predictive quality with respect to the aggregated stock market, which per definition contains a minimum level of idiosyncratic risk and provides a favorable environment for such applications. Economic intuition suggests that the level of out-of-sample predictability decreases as we climb down the ladder from market aggregates to industries and ultimately single stock returns. Thereon, we ask the central question: ‘Do forecasting errors from direct predictions of market components out-way the additional errors introduced by an intermediary asset pricing model?’ This is an early stage research note and we welcome any feedback and comments.

Publication
Working Paper