Disentangling Limits to Arbitrage: Empirical Evidence on Volatility-Based Limits and Cost-Based Limits

Abstract

This study investigates how limits to arbitrage contribute to persistent mispricing by decomposing them into cost-based and volatility-based components. Cost-based limits arise from borrowing costs, lending-market scarcity, and lending fragility, while volatility-based limits capture risk-related frictions such as position size constraints and margin-call risk. Using U.S. equity data from May 2002 to December 2016, this study integrates double-sorting, Fama–MacBeth cross-sectional regressions, and the Hou and Loh (2016) decomposition framework to quantify the independent and relative effects of these limits. The results show that volatility-based limits exert a significant additional effect in preventing mispricing correction on top of cost-based limits. Decomposition analysis reveals that the relative contribution ratio of volatility-based to cost based components is about 3:7, indicating that mispricing persistence is not fully driven by cost-based limits alone. Robustness tests using value-weighted portfolios, portfolio-level decomposition, alternative factor models, and interpolation methods confirm the consistency of these findings. Overall, the evidence demonstrates that volatility-based limits represent a distinct and economically meaningful dimension of limits to arbitrage. Recognizing both cost-based and volatility-based limits provides a more comprehensive understanding of why mispricing persists in financial markets. Show more