Asset Heterogeneity and Uncommon Factors
Date:
We challenge the restrictive “one-size-fits-all” assumption of standard asset pricing models by demonstrating that the stochastic discount factor is locally sparse within economic clusters but globally dense in the aggregate. Using a novel Bayesian Fama-MacBeth Clustering Model, we identify “uncommon factors” – risks that command premia only within specific asset clusters or economic regimes. For instance, while Size is the most pervasive common factor, Momentum and Value are priced primarily among low-volatility, large-cap stocks, but they command premia in distinct macroeconomic liquidity and issuance regimes. This heterogeneity generates substantial economic value: cluster-based portfolios yield high Sharpe ratios and alphas unspanned by standard factors. We rationalize these findings through a segmented markets framework, where informational frictions create local marginal investors. Our results suggest that part of the “factor zoo” reflects segmentation: factors can be priced in some segments but not others.
Paper: SSRN link