Another IT model success: forecasting [using] exchange rates

At the beginning of the year, I used the IT model to forecast a relative fall in Canada's GDP compared to the US via the decline in the exchange rate, and that turns out to have happened. Here's the original graph:

And here's the latest data through 2016 Q3 (new and revised data are darker blue/darker yellow = brown):

Note that due to data revisions in GDP that was used in the original fit, the fit changed a tiny amount.

Update: Also, it seems, like usual, the exchange rates over-react. So this is useful direction information, but not necessarily magnitude.

Update: The title should read forecasting using exchange rates. The exchange rate data is available daily while GDP tends to come out a month after the quarter it represents ends.