Forecast Evaluation of different models

[jugheadj]

Hi all,

I've estimated various error correction models of which I've also done a 1 year ahead in-sample forecast. What I want to know is that since these ECMs are all forecasting different variables in different periods, can I still do a forecast evaluation and compare them with MAPE, MAE etc? Or is forecast evaluation just for different models trying to forecast the same thing?

How do I know if my forecasts are accurate if they are all forecasting something different? Eg, one ECM forecasting AUD/USD from 2010 to 2015 whilst another is forecasting CHF/USD from 2006 to 2010?