Data applications in quantitative finance

The PhD research consists of two parts, each discussing the application of one specific type of data in the field of quantitative finance. In the first part, we examine how machine learning models can be trained to compute prices of financial derivatives. Current state-of-the-art derivatives pricing models have become complex, and one often needs to rely on time-consuming numerical methods to evaluate them and compute prices. The objective of this research is to develop a pricing method that can compute derivatives prices in a fast, efficient way. This may be accomplished by generating a data set that reflects the pricing function of a complex pricing model, and train a supervised learning model on that set. The trained model would then be able of computing prices that correspond to the complex model, but can be computed in a faster way. In the second part of the PhD research, we analyze how ESG data can be included in the selection of optimal equity portfolio investments. ESG indicators evaluate companies on their environmental, social, and governance policies. We consider two specific ESG indicators; (1) the ESG rating, reflecting the overall degree of sustainability, and (2) the greenhouse gas intensity, reflecting the CO2 emissions of a company. In our research, we examine the impact of including these criteria on the riskiness and the returns of the resulting optimal portfolios.

Jaar van publicatie:2021

Toegankelijkheid:Closed