Forecasting with Leading Economic Indicators — A Neural Network Approach

There is variety of important issues associated with the problem of business cycle forecasting, especially regarding forecast methodology and forecast evaluation. Overall, we can say that macroeconomic forecasting has a fairly poor reputation (Granger 1996). Still, even with the recognition that forecasting business cycles is a very difficult task, we find some hopeful signs for future progress. Our research on forecasting has focused on development of new approach in forecasting with classical NBER leading indicators by applying neural networks.