Abstract

The paper presents a study on model credibility, which was completed in Japan. More than 10 economywide models were selected and GNP results compared. The economists found that the model with the smallest prediction error was a very simple model: a linear extrapolation of the trend in the previous year. In other words the hypothesis that the GNP growth rate during a year is the same as the rate during the previous year proved to be an accurate forecast. Undoubtedly not only modelers but those who are not familiar with modeling methodology would be surprised.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call