The authors revisit why a factor with over $392 billion in low volatility investment solutions remains mostly absent from standard asset pricing models. Their answer is that traditional tests assume symmetric long and short legs and frictionless markets, while real portfolios face asymmetry, costs, and shorting constraints.
Factoring in the Low-Volatility Factor
Pim van Vliet, Amar Soebhag, Guido Baltussen
Research
57 Pages
Key Takeaways
Frictionless Tests Miss: Adding low volatility to traditional long short models raised maximum Sharpe ratios by only 0.57%, explaining its omission from many asset pricing models.
Long Leg Matters: Separating factor legs improved model performance by 11.9%, with the hedged low volatility leg receiving an average 26.2% portfolio weight.
Costs Change Results: After transaction and shorting costs, low volatility still improved factor models by 13%, while PEAD fell from 3.45% to negative 0.87% annually.