Monthly prices of four categories of Douglas-fir lumber (dry clears, green clears, dry commons, and green commons) were analyzed using reduced form econometric models covering the period January 1954 to June 1969. Numerous explanatory variables were considered for inclusion in the final models. The most complete final model was that for green clears which had five explanatory variables: average price of Douglas-fir logs and wage rates in wood processing each with a 3-month lag, and the previous month’s level of the price and of unfilled orders and gross stocks in the Douglas-fir region. The dry clears model was similar except that wages in wood processing were not included. The dry commons model was similar also, except for the omission of log price. The green commons model had only three explanatory variables: log price, the previous month’s price level and the level of unfilled orders in the previous month. The previous month’s price was a dominant variable in all cases. Unfilled orders in the previous month and wage rates were the only other important variables. Log price and the level of stocks appeared to have a negligible effect on lumber price. The potential of the models for forecasting purposes was examined using the test period September 1968 to June 1969. Average absolute forecast errors ranged from 2.9 percent to 5.7 percent.
You must be logged in to download any documents. Please login (login accounts are free) or learn how to Become a Member