Liquidity, time-varying betas and anomalies: Is the high trading activity enhancing the validity of the CAPM in the UK equity market?

Abstract

In the last decades, a large number of multifactor assetpricing models have emerged with the aim of correcting the estimated equity risk premiums for some well-documented market anomalies. In any case, recent research on asset pricing shows how the higher liquidity resulting from the globalization of financial markets has significantly reduced returns tied to many strategies based on market anomalies. In this framework, questions arise about the possible renovated validity of classic assetpricing models. On this basis, in this paper we study to what extent the capital assetpricing model (CAPM) has recovered its past explanatory power. Specifically, we propose a time-varying beta CAPM in order to control for the variable nature of beta risk to changes in the market liquidity, using the variation of the Amihud illiquidity measure to account for the degree of trading activity. We test both the time-varying and constant beta models on different sets of anomaly portfolios for the UK equity market, and we compare their performance to that of the Fama–French threeand five-factor models. Additionally, we test the constant beta model on a set of actively managed portfolios, formed according to the variation in market illiquidity in the previous year. Our results show that the pricing errors of the CAPM have significantly decreased with respect to those of previous literature. Furthermore, the time-varying beta model performs similarly to the Fama–French models in most cases. These results are consistent with increased trading activity that reduces arbitrage opportunities and, therefore, enhances market efficiency.