Chequamegon National Forest Research Articles

Statistical Appraisal of Timber with an Application to the Chequamegon National Forest

Abstract A statistical method of appraising timber is presented. It consists of predicting the high bid of a particular timber offering under competitive conditions and adjusting this value to reflect uncertainty and the goals of the selling agency. Using data from the Chequamegon National Forest in northern Wisconsin, it was found that a simple linear model using 14 variables explained 93% of the variance in high bid for competitive sales from 1976 to 1980. This model predicted well post-sample high bids for 1981 and 1982. Based on this model, three possible definitions of appraised value were investigated: (1) predicted high bid, (2) predicted high bid minus one standard error, and (3) predicted high bid minus two standard errors. The consequences of each definition on timber sales, had it been applied in 1981 and 1982, were examined. Definitions (1) and (2) would have increased receipts by 28 and 5%, while decreasing the volume sold by only 5 and 3.6 %, respectively. Definition (3) would have led to the sale of approximately the same volume, but a decrease in receipts of 26%. North. J. Appl. For. 1:72-76, Dec. 1984.