A parameter based approach to single factor stochastic process selection for real options applications

Abstract

The single factor stochastic diffusion processes most commonly used for Real Options Valuation are the Geometric Brownian Motion and Mean Reversion Models. Nonetheless, the choice of process to model asset price dynamics is still one of the main challenges for researchers and practitioners in the field. Particularly, in investment projects where there is significant managerial flexibility, the project value and the investment rule may depend in large part on the process used to model the underlying uncertainties. In this article, we develop an approach based on the parameter values of the model, which has some advantages over the methods currently used for stochastic process selection. We use the half-life and normalized variance of the time series to be modeled to determine the optimal choice and discuss related theoretical as well as practical issues concerning the application of this approach to real options valuation. Numerical examples are used to illustrate the method, and a guideline for implementation of this approach is provided.