Riding the Low-Beta Wave : What drives the performance of Betting Against Beta?

Abstract

This thesis seeks to explain the driving factors behind the Betting Against Beta (Frazzini and Pedersen, 2014) portfolio. We start by replicating the BAB factor, and then construct different portfolios in order to examine the factor's robustness, to which degree returns are driven by placements in extreme-beta stocks, and whether the factor is at all driven by industries. We find that using a different method of beta estimation dampens the portfolio's returns somewhat, but still generates positive and significant alphas. Equalweighting and value-weighting the portfolio also leads to weaker returns, with the latter performing exceptionally poorly when looking at risk-adjusted returns, implying that returns are driven by heavy weightings in stocks of smaller market capitalisation. We also find that a portfolio that buys and sells the outermost beta-sorted deciles performs better than buying all stocks. This leads us to construct a version of the BAB portfolio that buys the outermost beta deciles and excludes micro-cap stocks, which again generates a higher alpha than the all-stocks portfolio, but produces lower risk-adjusted returns. Finally, by constructing three different kinds of industry portfolios, we corroborate findings by Asness et al. (2014) that BAB is not driven by industries.