Another Look at Stock Return Comovement: Some New Evidence and Test

Abstract

The study of the comovement between asset returns reflects an ongoing effort by economists to understand investment risk in financial markets. Building on previous findings, in the current thesis I provide some new evidence on this topic with a focus on large-cap stocks and highlight an innovative way to evaluate the statistical significance of comovement asymmetry. In the first part of the thesis, I revisit the question of how large-cap stock return comovement varies with volatility and market returns. I propose the use of an eigenvalue-based measure of comovement in a multivariate semi-Markov-switching framework. I conduct various model evaluation checks and compare the new results with that based on a benchmark. I estimate models with two to four regimes and consider the impact of sample selection and outlier reduction. Contrary to the sweeping sentiment that comovement is highest when market is down and volatile, I illustrate the significance of comovement differential across states and find in most case studies evidence that suggests otherwise. In the second part, I propose a test of asymmetric stock return comovement across states. The test can be viewed as a variation of Kendall’s τ conditional on the state and has an asymptotic χ^2-distribution. A refined version of the test is derived based on the Markov chain theory of regenerative cycles which substantially improves finite sample size and power. I show that the test has power against local alternatives, which is nonetheless compromised due to a finite sample convergence bound put on the implied local alternative data generating process. I evaluate the new test against traditional correlation-based measures and demonstrate power attrition due to nuisance parameters when states are ignored. I find that asymmetric tail dependence becomes much less significant when considered state by state. A list of related tests is given as an extension at the end.