Construct the 1-step-ahead forecast. Because you have reserved some of your observations (prediction sample), evaluate how good

Question:

Construct the 1-step-ahead forecast. Because you have reserved some of your observations (prediction sample), evaluate how good the 1-step-ahead forecasts are by implementing a test of unconditional predictability. Compare the 1-step-ahead forecasts from the VAR model with those that you would obtain from the best univariate model that you could find for each series.