A major challenge in asset pricing theory is to explain the cross-section of asset returns. To address this challenge, a large number of systematic risk or common risks have been identified and examined by the literature. This multitude of factors is often called the ‘factor zoo’.
Despite considering hundreds of systematic risk factors, factor models still have sizable pricing errors. In the final session of the Inquire Europe Autumn Seminar in Cologne on 26 September, Raman Uppal addressed these issues in his presentation based on research from his paper ‘What is Missing in Asset-Pricing Factor Models?’, in which he and his coauthors provide an avenue for resolving the factor zoo.
For those who were unable to attend his presentation, a summary is available below:
“One of the most important challenges in finance, and specifically for asset managers, is to figure out why one stock has different expected returns compared to another stock. This is important, not just for the first moment of returns, but also to adjust expected returns for risk. If we want to do asset allocation, stock selection, or risk management, we need to have a good model to understand what drives stock returns. The first model proposed to address this challenge was the CAPM; we now know that empirically, the CAPM doesn’t perform very well. Researchers have then explored other candidate models with hundreds of systematic risk factors (factor zoo). However, there is still a sizable pricing error in returns, typically called alpha, a large proportion of which cannot be explained by existing asset-pricing models. In this presentation, we ask: What is missing in asset-pricing factor models?
Existing models typically assume that only systematic risk is compensated in financial markets and that unsystematic risk is uncompensated. In our research, we also allow compensation for unsystematic risk and show that this single departure from the traditional risk-return tradeoff provides an avenue for resolving the factor zoo.
Theoretically, we demonstrate this key observation using the framework of the Stochastic Discount Factor (SDF), assuming the Arbitrage Pricing Theory (APT). Additionally, we validate our argument by demonstrating that an equilibrium model, such as Merton’s presidential address to the American Finance Association in 1987, is consistent with our insight.
Empirically, we demonstrate that the component of the admissible SDF reflecting unsystematic risk, which is a linear combination of unsystematic shocks, accounts for more than 70% of the variation in the admissible SDF. Moreover, the Sharpe ratio linked to an investment strategy exposed solely to unsystematic risk is 0.80 annually. Thus, what is missing in asset-pricing factor models is compensation for this unsystematic risk. This insight holds significant importance for empiricists aiming to resolve the challenge of pricing assets and for theorists wishing to develop micro-founded asset-pricing models”.
View the research in full via: https://www.inquire-europe.org/event/autumn-seminar-2023/