Uncertainty Aversion and the Stochastic Forest Harvesting Problem

49 Pages Posted: 12 Aug 2024

See all articles by Haoyu Wang

Haoyu Wang

Virginia Tech

Gregory S. Amacher

Virginia Polytechnic Institute & State University - Department of Forestry

Abstract

We examine uncertainty in a dynamic forestry model where a harvester is unsure about model specification, and where there are both benefits to standing forests and to harvesting. We extend the literature to allow for unknown drift and volatility under several uncertainties, including amenity flows and harvest prices, relaxing usual assumptions of risk aversion and known distributions to cases of uncertainty aversion consistent with pure uncertainty. Our results show that model uncertainty can influence harvesting paths in new ways, showing in some cases that the forest can be depleted or, alternatively, never harvested because of cautionary behavior. We also reexamine two basic arguments in the literature, first, that amenity values always result in less harvesting and higher forest stocks over time, and second, whether higher variances (volatility) in prices lead to lower harvesting as landowners wait for possibly higher future prices. Here we find that conventional wisdom may be flawed. Our new approach is important to the fate of forests considering that climate fluctuations may be associated with collapse of tropical forests.

Keywords: Climate change, pure uncertainty, robust control, stochastic dynamic programming

undefined

Suggested Citation

Wang, Haoyu and Amacher, Gregory S., Uncertainty Aversion and the Stochastic Forest Harvesting Problem. Available at SSRN: https://ssrn.com/abstract=4917434

Haoyu Wang (Contact Author)

Virginia Tech ( email )

Blacksburg, VA
United States

Gregory S. Amacher

Virginia Polytechnic Institute & State University - Department of Forestry ( email )

307 Cheatham Hall
Blacksburg, VA 24061
United States

0 References

    0 Citations

      Do you have a job opening that you would like to promote on SSRN?

      Paper statistics

      Downloads
      12
      Abstract Views
      75
      PlumX Metrics