I'm comparing the volatility of capital flows using a panel data regression. I would like to examine changes in volatility over different periods (e.g. before financial crisis, financial crisis, unconventional monetary policy and tentative recovery periods) using dummies. However, this periods are of unequal length, would this do anything to invalidate this comparison?
Any ideas would be much appreciated !