December 06, 2024
11:00
LIDAM D.251
Daniele Massacci (King's College London)
Invited by Nathan Lassance
will give a presentation on :
State-dependent comovement between factor models
Abstract :
We study regime-specific comovement between two large panels of variables, each exhibiting an approximate factor structure. Within each panel, we identify threshold-type regimes through shifts in the factor loadings. For the resulting regimes, we measure the between-panel comovements generated by the common components. Our measures can then be interpreted as model-implied covariance and correlation under the assumption of an underlying factor structure in each panel. We propose estimators for our measures, derive their asymptotic properties, and develop inferential procedures to test for changes in comovement between different regimes. We then study comovement across financial asset classes and obtain three results for U.S. portfolio stock returns in downside relatively to upside periods: (i) their comovement with U.S. Treasuries becomes stronger and more negative; (ii) they exhibit significant heterogeneous changes in comovement with equity index call and put options; (iii) their comovement with U.S. corporate bond portfolios becomes positive and significant. Risk management implications are discussed.