Identifying latent factors based on high-frequency data

This paper tests whether the continuous component of an observable candidate factor is in the space spanned by the counterparts of latent common factors with high-frequency financial data. We introduce two identification strategies corresponding to two types of regressions: the regressions of intraday asset returns on the estimated factors and the candidate, and the regression of the candidate factor on the estimated ones. We construct the test statistics by adding randomness to the statistics obtained from residuals of the regressions, and demonstrate the consistency of the novel randomized tests. Simulations are conducted to evaluate the performance of the tests in finite samples. We also perform empirical applications to identify the relationships between some candidate factors and the latent ones, and further use the factors selected by the tests for portfolio allocation.(c) 2022 Elsevier B.V. All rights reserved.

Automated summary

This paper tests whether the continuous component of a candidate factor can be explained by the latent common factors in high-frequency financial data. Two identification strategies are introduced for two types of regressions: regressions of intraday asset returns on the estimated factors and the candidate, and regressions of the candidate factor on the estimated ones. The test statistics are constructed by adding randomness to the residuals of the regressions, and the consistency of the randomized tests is demonstrated. Simulations are conducted to evaluate the tests' performance in finite samples, and empirical applications are performed to identify relationships between candidate factors and latent ones.