A0303
Title: Factor inference under common components in volatility
Authors: Julia Koh - Tilburg University (Netherlands) [presenting]
Benoit Perron - University of Montreal (Canada)
Silvia Goncalves - McGill University (Canada)
Abstract: Considering the common components of volatility is crucial in finance. It is found that the standard factor literature does not account for these common components in volatility. Alternative assumptions are introduced that can accommodate the volatility dependence and develop inference methods for factor models. Three key applications are revisited: Factor estimation and distribution theory, factor-augmented regression models, and a test for the number of factors in group factor models. Under new assumptions, it is shown that the results for factor estimation and factor-augmented regression models are maintained. It is found that the variance of the test statistic for the number of factors in the group model is affected, and a new estimator for the variance is proposed.
