"There's a lot of specific data not being exploited yet in the financial market"

 

As the CEO of the Divina Seguros insurance company for over fifteen years, Armando Nieto's experience in financial mathematics and computing speaks for itself. He has just completed a PhD at the Universitat Oberta de Catalunya (UOC), on the Network and Information Technologies programme, with Ángel A. Juan (Universitat Politècnica de València) and Montserrat Guillén Estany (University of Barcelona) as his thesis co-supervisors. His thesis focused on exploring how artificial intelligence can be applied to financial mathematics.

 

As the CEO of the Divina Seguros insurance company for over fifteen years, Armando Nieto's experience in financial mathematics and computing speaks for itself. He has just completed a PhD at the Universitat Oberta de Catalunya (UOC), on the Network and Information Technologies programme, with Ángel A. Juan (Universitat Politècnica de València) and Montserrat Guillén Estany (University of Barcelona) as his thesis co-supervisors. His thesis focused on exploring how artificial intelligence can be applied to financial mathematics.

The thesis looks to address a highly complex issue in the financial and insurance market: how to invest while meeting two simultaneous objectives. The first is to secure our long-term financial commitments. And the second is to do this at the lowest possible cost, or with the highest returns. The problem has been covered extensively in the scientific literature, but, due to the mathematical complexity, we do not as yet have any satisfactory answers. This has in turn led to overly restrictive financial regulation.

We aimed to adopt an entirely different approach to find solutions that were acceptable in the financial market without aiming for an exact solution, which is meaningless in a market that is constantly changing. We also analysed the regulatory restrictions on financial institutions in detail, precisely because no reliable methods have been discovered so far.

The solution is based on simheuristic algorithms, which combine heuristic and metaheuristic algorithms with simulations. Heuristic algorithms are the key to solving these problems, since we have to forget about the idea of mathematically exact solutions, because they are highly complex problems. Simulation means we can improve our solution by incorporating all the knowledge we have of the market into the algorithm.

In the first part of the thesis, we look at the usual problem of insurance companies, and the results are excellent. The algorithm finds solutions that improve the cost of real cases by 10%, and it does it in a matter of seconds, compared to an actuary, who would take at least a week. In the second part of the thesis, we applied genetic algorithms to obtain the long-term investment plan subject to budget restrictions and liabilities for the entire life of the financial institution.

Applying genetic algorithms, simulation and machine learning gives us solutions that are almost exact in just a few minutes using a home computer, while traditional solutions require days of work and huge computing power.

Artificial intelligence (AI) is already at work in many fields. It is made up of technologies that have already exploded, but which have shock waves that have not yet reached one thousandth of the impact they will have in the future. In the insurance field, it provides more efficient and safer solutions for the end client. AI also covers some financial mathematics and computing, which is the subject of the thesis, and even influences decision-making.

There's a lot of specific data not being exploited yet in the financial market. AI can be a very effective tool when making decisions. Insurers receives information from many different sources, and in very varied formats (images, natural language, etc.) and AI algorithms can help us simplify the investigative work.

As we have shown in the thesis, these technologies have great potential in the financial and insurance fields. One of the major challenges is learning how to process and format data in a way that can be useful for machine learning algorithms. But applying AI in insurance is also subject to quite a powerful restriction: legislation.

The legislation is very prudent on both the insurance and the banking sides. If we open things up without any oversight and without any knowledge, this could lead to abusive market practices, and the consumer would suffer. It is important to convince the government of this. I have already let the Spanish Directorate-General for Insurance and Pension Funds and the Spanish insurers' association, UNESPA, know about the subject of my thesis. And next year I will be in Australia at the International Congress of Actuaries to present a part of the research undertaken in the thesis.

One of the most rewarding things was working with two scientists who are at the highest level internationally: Ángel A. Juan and Montse Guillén. And the other was overcoming the challenges involved in solving a problem that had been stalled – no significant progress had been made in recent years – and being able to open up new lines of research.

It was important for two reasons. The first is that I worked with Ángel A. Juan, who is very versatile; he is always moving forward. His department and his team have produced a great deal of research and results. The second is that with its structure, the UOC makes it easier for people like him to fulfil all their potential. That is not the case at other universities, which have more bureaucracy and more rigid structures.

The heuristics department at the UOC were mainly working on transport and logistics. When I told them about the subject of my thesis in the financial field, they understood its potential and its interest. The UOC gives you the freedom to see it, and provides the structure necessary to develop it.

There are some developments to explore that I will attempt to address over the coming year. All research aims to improve knowledge and practice. Disseminating the thesis contributes to this, and could encourage other professionals to continue improving the models and to apply them in the financial industry.

 

This research supports Sustainable Development Goal (SDG) 9, Industry, Innovation and Infrastructure.

 

 

UOC R&I

The UOC's research and innovation (R&I) is helping overcome pressing challenges faced by global societies in the 21st century, by studying interactions between technology and human & social sciences with a specific focus on the network society, e-learning and e-health.

The UOC's research is conducted by over 500 researchers and 51 research groups distributed between the university's seven faculties, the E-learning Research programme, and two research centres: the Internet Interdisciplinary Institute (IN3) and the eHealth Center (eHC).

The University also cultivates online learning innovations at its eLearning Innovation Center (eLinC), as well as UOC community entrepreneurship and knowledge transfer via the Hubbik platform.

The United Nations' 2030 Agenda for Sustainable Development and open knowledge serve as strategic pillars for the UOC's teaching, research and innovation. More information: research.uoc.edu