Thinking About Composite Forecasting

THinkinG abouT CoMpoSiTe ForeCaSTinG by larry lapide, ph.d. (This is an ongoing column in The Journal, which is intended to give a brief view on a potential topic of interest to practitioners of business forecasting. Suggestions on topics that you would like to see covered should be sent via email to llapide@ mit.edu). I I went to a conference about 30 years ago and attended a session by exponential smoothing guru Robert G. Brown. During the question-and-answer period of his talk, someone asked him what he thought about the composite forecasting topic that was being discussed in another room at the conference. I forget what he actually said, but effectively he was not keen on the idea. Perhaps colored by his view, I had never been a big fan of composite forecasting. I thought that it was not worth the extra effort in business forecasting where we have to generate myriad forecasts. Who has the time to play around with figuring out the best blending of two or more forecasts into a single forecast? It just seemed easier to pick the best one from a bunch of forecasts and go with that. I believe that this is consistent with the thinking of most business forecasters who have to deal ...