I am analysing time series data right now using gretl, and want to test for a structural break, but I am not quite sure how I have to interpret the results.
Let's say I have a wheat price and flour price, and I'm doing a VECM. Without a structural break my ECM is -0.12, so it's correcting 12%. When I do the Chow test, I break the time series in half and I get the coefficients 1: -.115 and 2: -0.072.
Does that mean that in the first half the ECM is correcting with 11.5% and in the second half with 7.2%?
Are the results reliable when only one coefficient is significant?