Abstract (eng)
The main objective of this work is to formulate, implement and evaluate a pharmaceutical investment following the real options approach. After a presentation of different models the Least-Squares Monte Carlo model by Schwartz (2004) is presented and implemented. The input data for the model is based on the case of a milestone payments agreement between GlaxoSmithKline and the biotech company Apeiron. The data was collected and validated through interviews with the management of GlaxoSmithKline and Apeiron. After the implementation of the single-stage model, the results were again discussed with GlaxoSmithKline and Apeiron and the objectives for an extension of the model were set. The extended model considers the research and development phases as separate stages with different cost structures and probabilities of failure. This thesis is structured as follows: After the introduction the second chapter gives an overview of the real options theory, starting with general characteristics of financial options. The third chapter provides an overview of the structure of the pharmaceutical research and development process and discusses various methods for estimating cost and duration. The fourth chapter describes the Monte Carlo simulation in general and develops the Least Squares Monte-Carlo model for the single-stage process. In addition, explanations as to why the Least-Squares Monte Carlo approach was chosen over other approaches and why it generates better results than the standard net present value analysis are presented. Before the results of the single-stage model are presented in chapter six, the milestone payments agreement between GlaxoSmithKline and Apeiron is presented in chapter five. In chapter six the results of the single-stage model for the milestone payments agreement between GlaxoSmithKline and Apeiron are presented and discussed in terms of convergence and sensitivity analysis. Furthermore as in chapter seven, the implementation of various basis functions such as Legendre or Chebyshev polynomials is discussed. Lastly, in chapter seven the extension of the model of Schwartz (2004) in a multi-stage model is introduced and the achieved results are compared with the results obtained from the single-stage model in chapter six. The results, which are achieved by both the single-stage and the multi-stage model, display an improvement on net present value analysis if the underlying data is suitable.