This paper discusses our aims in organizing this special issue. We believe that a synthesis of the field is needed to resolve the many conflicts between theory and empirical evidence and between advocates of the various forecasting methods. One step toward synthesis is to agree on the major problems in the field, the goal of this issue. Another step is to establish standards for both practice and research and some proposals are made to this end. We also give some personal opinions on the most important research opportunities in the field.
- time series