I'm trying to detect return predictability by regressing stock returns on the first couple of principal components of a set of macroeconomic variables. I'm doing this for different stock styles such as small, value and momentum stocks. Since I don't know at which lag I should insert these principal components, I'm hoping to find someone who can help me out.

enter image description here

Later on I would use these principal components in another regression that looks like this:

enter image description here

  • $\begingroup$ Maybe you should first implement an automatic selection algorithm which starts with a lot of possible lags then gradually removes them by performing new regressions. It stops as soon as all your coefficients are statistically significant. Let data talk. $\endgroup$
    – Lisa Ann
    May 1, 2019 at 16:07


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.