Enhanced real options valuation with Machine learning : Applied case to energy finance

Abstract

This thesis explores real option valuation in the energy industry using deep learning methodologies. Despite the theoretical foundation of real options in financial analysis, their practical application in the volatile energy sector remains under-explored. This study bridges this gap by integrating advanced data science techniques with traditional financial models. Utilizing machine learning architectures, particularly deep learning, the study evaluates these models’ efficacy in capturing the uncertainties and dynamic investment opportunities in energy projects, comparing their performance against traditional financial approaches and integrating predictions within the Black-Scholes-Merton model. The empirical case focuses on the European energy generation industry. This research validates deep learning’s utility in enhancing cash flow prediction and optimizing investment decisions under uncertainty. The thesis contributes to finance, energy economics, and AI, providing valuable tools and techniques for industry practitioners and researchers.