Forecasting the Price of Lumber and Plywood: Econometric Model Versus Futures Markets

Three methods of forecasting the price of lumber and plywood were compared: 1) the FORSIM model of Data Resources Inc., 2) the futures markets, with prices of contracts for future delivery used as forecasts of cash prices, and 3) a naive model where the predicted price is equal to the last known cash price. The comparisons used data for the period 1974 to 1981. For lumber forecasts of the current quarter and one quarter ahead there was no significant difference in accuracy between the FORSIM and futures market. For two and three quarters ahead, FORSIM was better. For all horizons, FORSIM and the futures market were more accurate than the naive model. For plywood forecasts, of the current quarter and one of three quarters ahead, there was no significant difference between FORSIM and the futures market. For two quarters ahead, the futures market was better. The naive model was as accurate as FORSIM for forecasts one and two quarters ahead, and as accurate as the futures market for the current quarter. Turning points of lumber prices were predicted best by the naive model. Most of the prediction errors of FORSIM and the lumber futures market were of a random nature, suggesting that there is little room for improvement in either approach. Some inefficiency was apparent in the plywood futures market.

You must be logged in to download any documents. Please login (login accounts are free) or learn how to Become a Member