Factor investing: A Bayesian hierarchical approach

This paper investigates the asset allocation problem when returns are predictable. We introduce a market-timing Bayesian hierarchical (BH) approach that adopts heterogeneous time-varying coefficients driven by lagged fundamental characteristics. Our estimates conditional expected and residual covariance matrix jointly enables evaluating estimation risk in portfolio analysis. The prior allows modeling of different assets separately while sharing information across assets. demonstrate performance U.S. equity market, our BH outperforms most alternative methods terms point prediction interval coverage. In addition, efficient achieves monthly 0.92% significant Jensen’s alpha 0.32% sector investment over past twenty years. detect technology, energy, manufacturing critical sectors decade, size, investment, short-term reversal factors heavily weighted portfolio. Furthermore, stochastic discount factor constructed can explain many anomalies.