Flow-Based Asset Pricing: A Factor Framework of Cross-sectional Price Impacts

Abstract

We develop a framework to study how noisy flows impact the cross section of asset prices. Instead of treating flows as additional risk factors, our framework posits that flows into canonical factor portfolios drive time-series fluctuations in these factors' risk premiums. In the cross section, each asset’s price impact depends on its (return) risk exposure to the factors. To build this framework, we introduce flows into classical portfolio tools, including Sharpe ratio, Fama-MacBeth regression, Fama-French portfolios, and Gibbons-Ross-Shanken test. We estimate the model using U.S. equity mutual fund flows data. The model-implied strategy that optimally profits from flows improves the investment performance of most existing anomaly portfolios.

Keywords: cross section, factor, flow, price impact, risk

JEL Classification: G12