科研成果详情

We consider the problem faced by mining companies to set their extraction rate over time in orderto meet their production targets, from a real option perspective.As part of their strategic planning, mining companies usually set a production target for a given,short-term to medium-term, time horizon. The extraction rate is constrained by this target, butremains flexible around a certain base rate. Because of the uncertainty in commodity prices, thisflexibility has a value. Basically, by slowing down production when prices are low and speeding upproduction when prices are high, the miner can still meet its production target exactly, while at thesame time significantly increase the value of the mine.Mathematically, this resource extraction problem can be described as a multi-regime constrainedstochastic control problem. The control is the extraction rate, to be set for every time step, andthe constraint is the total extraction volume over the whole time period. We solve this problemnumerically using an extension of the simulation-based Least-Squares Monte Carlo algorithm. Todeal with the endogenous reserve variable, we use the so-called control randomization technique.The estimated mine value increase would be nothing without the roadmap to achieve this improve-ment in practice.  Fortunately, our numerical solution provides the optimal extraction regime toapply over time, for any commodity price level and any reserve level. We summarize the informa-tion about the optimal regime switching surfaces into intuitive graphics that can assist industry fortheir sequential production decisions under uncertainty in practice.