Carhart 4 factor model

[John Cron]

Hi there,

I am currently doing some analysis looking at the affect of CSR on stock returns during a crisis and non-crisis. I have created a long short portfolio which goes long on high CSR firms and is short on low CSR firms. I have used Carharts 4 factor model for both time periods and during the non-crisis period i get a negative and significant alpha but the momentum factor is not significant. Whereas in the crisis period my alpha is positive but not significant, but the momentum factor is now significant (still negative). I also ran the fama-french 3 factor model and during the crisis period i get a highly positive and significant alpha but when the carhart model is used the result is not significant. My question therefore is, is there a reason that the momentum factor during the crisis period is now significant and what could be some reasons for it impacting the alpha so significantly.

Hope that makes sense

[Joro Kolev]

This is a topic of another paper: How does the momentum factor behave in the times of crisis. (And what crisis do you mean, the Covid? Or general crises that you have defined some way?)

In any case you care just about the alpha.

At the time I use this methodology
Kolev, Gueorgui I. "Underperformance by female CEOs: A more powerful test." Economics Letters 117, no. 2 (2012): 436-440.
Kolev, Gueorgui I. Behavioural Biases and Chief Executive Officers Compensation. Universitat Pompeu Fabra, 2010.

the norm was that we try both the Fama French three factor model, and the Carhart 4 factor model.

And you just observe how the alpha changes in the two conditions.

(I think now more factors have cropped up, but the idea is that what factors we use is a specification issue, and at the end we care only about the alpha.)