A mean-variance approach to short-term timber selling and forest taxation under multiple sources of uncertainty

The effects of forest taxation on timber supply under the double uncertainty of real interest rate and future timber price are analyzed in a mean-variance framework. If future timber price and interest rate are independent random variables, there exists an asymmetry between lenders and borrowers, as there is in the single interest rate uncertainty case. Borrowers cut more and lenders cut less as a response to double uncertainty relative to single price uncertainty. If timber price and interest rate are correlated, the sign of the covariance is crucial for the interpretation of the optimality conditions and for the comparative static analysis results. If covariance is negative, borrowers still tend to increase their cutting, but the behaviour of lenders depends on the relative magnitudes of risks associated with interest rate and timber price, which is indicated by variances and covariance. Finnish data are used to evaluate the size of various risks and the sign of covariance. The evidence suggests that a statistically insignificant negative correlation exists and that the covariance is zero.