How To Kill Your Trees

The White Paper provides a detailed exploration of the limitations and inefficiencies of tree-based numerical methods, particularly binomial and trinomial trees, in the context of computational finance. The author, Andreas Binder, critiques these methods by highlighting their instability and lack of accuracy, especially when applied to mean-reverting models commonly used in financial instrument pricing.

The paper discusses alternative approaches, focusing on advanced numerical techniques such as finite elements and streamline diffusion, which offer greater stability and precision in solving partial differential equations. These methods are not only more robust but also have been successfully applied in industrial mathematics, particularly in processes like the simulation of blast furnaces. The author concludes by advocating for the abandonment of traditional tree methods in favor of these more sophisticated techniques, illustrating their superiority through practical examples and numerical experiments.