What are some of the issues relevant to forecasting with multiple regression models?
Just because a multiple regression model is statistically significant, does that mean it will do a good job forecasting?
Because we are now dealing with a statistical model, we have the ability to develop different types of forecasts.