I have estimated a regression and found a breakpoint using chow:
Then I created a dummy variable which is 0 before the break and 1 from the break until the end. This is the result:
I was assuming this p-value for the dummy variable means the break is not relevant. Is this correct?
I then interacted the variable with the other independent variables:
One of the interactions is very significant.
I don't know what to conclude from it. Is this break relevant and I should segment my regression or not?
If it makes any difference: These are financial returns of a company as dependent upon a stock index returns and usd exchange rates, all as log.