Blackwell Reference Online

Extract

This chapter provides an overview of techniques for evaluation of forecasts when the primary purpose of forecasting is to make better decisions. The traditional literature on forecast evaluation focuses on purely statistical measures of forecast accuracy without making specific reference to the way the forecasts are used. See, for example, Box and Jenkins (1970) , Granger and Newbold (1977 , 1986 ), and Clements and Hendry (1998 , 1999 ). The approach reviewed here assumes the forecast user's decision environment is known, and addresses the issue of forecast evaluation as an integral part of the decision-making process. Both approaches are clearly useful, as they address the same problem in different contexts. The statistical approach focuses on general measures of forecast accuracy intended to be relevant in a variety of circumstances, while the decision-based approach provides techniques with which to evaluate the economic value of forecasts to a particular decision-maker or a group of decision-makers. Theil (1960 , sections 8.4 and 8.5) provides an early discussion of the usefulness of the decision approach to forecast evaluation. Other important early contributions include Nelson (1972) and White (1966) , who are amongst the first to consider the problem of forecast evaluation from the perspective of decision theory and dynamic programming. Nelson, in particular, articulates ... log in or subscribe to read full text