Content for the HPC community and innovation enthusiasts: tutorials, news and press releases for users, engineers and administrators
- All News
- Aerospace & Defence
- Artificial Intelligence
- Cloud Platform
- E4 Various
- European Projects
- Kubernetes Cluster
- Latest news
Deep Learning on Exotic Derivatives Portfolio
This is the second article about the theses of some undergraduates, students of the Master’s Degree Course in Quantitative Finance at University of Bologna.
These theses are all about Deep Learning applied to financial markets, showing very interesting aspects. The students used a high performance computing infrastructure provided by E4, to make their analysis.
The work of this article is named “Deep Learning on Exotic Derivatives Portfolio” , written by Behnam Lari, Marco Bianchetti and Pietro Rossi.
In the last decades, the finance industry has consistently tried to exploit the computational power of modern hardware. Besides, we have witnessed the rapid growth of tools such as artificial intelligence (AI) and machine learning (ML) from supervisory agencies (suptech) as well. These advances in computational power have unlocked the application of algorithms that were not reachable in the past. In the present study, we aim to apply machine learning in order to price, simultaneously, the multiple payoffs of a portfolio that includes eight different exotic derivatives, in particular the performance of baskets of four stocks, and the respective vega hedges, composed of 16 plain vanilla options on the corresponding individual stocks. We present a novel combination of the Quasi-Monte Carlo (QMC) method and the Deep Neural Network (DNN) framework, where a massive dataset with 226 training examples on 120 payoffs (including all possible combinations of four stocks is generated with high accuracy, equivalent to 1M Monte Carlo (MC) scenarios…
To know more, download the PDF.