Interpretation lagged probit model


New Member
Hi there,

First of all I''m not sure if this should be posted here or under the statistical software topics but anyway... I have two time-series, one of them is binary and the other is not. Now I'm estimating a probit model with i lags varying from zero to 24. That is, I shorten my explanatory time series at the end by i observations and I shorten the binary series at the beginning by i observations. Now when I choose my preferred lag and make a plot of the binary series on the y-axis together with the explanatory variable multiplied by its beta plus the coefficient, I observe a spike preceding each occurrence of Y (Y=1). Is it a correct interpretation to say that a spike indicates that Y will occur in i periods? This is important because I want to use this model out of sample.

Kind regards,