Binomial Models for Interest Rates

Recombining binomial tree models are very convenient for derivative pricing. But traditionally trinomial tree approximations are often used and have become a standard way of approximating continuous time interest rate models. We provide a methodology for using binomial models rather than trinomial models for such approximation and provide comparison with these trinomial approaches. We will demonstrate some advantages of these binomial models over the popular trinomial tree models. We believe our approach is easier to apply and will perform better computationally and will help to improve some of the methods used by other researchers promoting lattice methods. The probabilities produced by our methodology are never negative for any discretization step length. The binomial model constructed by our method also preserves properties of the continuous time model like mean reversion.