0
$\begingroup$

I have a model for predicting stock returns that classifies stocks as overbought or oversold, kind of like an RSI. It follows an OUP and I am curious about my $\mu$, $\sigma$, and $\kappa$ parameters, so I want to do a regression on all my stocks over my entire time horizon. I could average my model over all my stocks and then follow the paper below, but if I have panel data, why not use it?

https://www.statisticshowto.com/wp-content/uploads/2016/01/Calibrating-the-Ornstein.pdf

I'm using linearmodels in python to conduct my panel data regression. It's a mixed model, with both fixed effects and random effects. My question is: how should I calculate the standard deviation of my errors to get my $\sigma$? I was thinking I use this formula, but I'm not sure if I'm interpreting everything correctly:

$\sigma =sd(\epsilon)\sqrt{\frac{-2ln a}{\delta(1-a^2)}} \propto \sqrt{\frac{(1-R_{within}^2)*SS_{total}}{n-2}}$

$\endgroup$

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.