I'm trying to estimate factor loadings on portfolios over time for portfolios that are traded pretty frequently. I have a sense that several portfolios are loading on the Fama-French HML factor dynamically, and OLS estimates show that there is not a statistically significant loading on the HML factor.

I'd like to estimate the portfolios' loadings on the market and HML factors over time using the Kalman filter, but I also want to make sure I don't overfit a model and include extraneous variables. Is there a standard procedure for ensuring that I'm not adding unnecessary variables to a Kalman filter regression?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.