XClose

UCL Institute of Finance and Technology

Home
Menu

AIRiskLab

29 May 2021

Our Mission

At the Artificial Intelligence Risk Laboratory (AIRiskLab), we aim to redefine financial risk management by leveraging Artificial Intelligence (AI) algorithms and investigating their impacts on modern quantitative risk analysis. AIRiskLab will be specifically devoted to AI modelling in risk management and its long- and medium-term impacts on financial stability and sustainability, based on a broad interpretation of risk in finance.

We live in an era where scientific fields are very much overwhelmed with and reliant on the applications of AI, and finance is no exception. Although the use of AI in financial modelling has already begun and the prospects are promising, the complexity of AI algorithms means that it is not clear how and to what extent such algorithms will reshape finance as we know today. AIRiskLab will be at the forefront of this challenge and will provide a blend of cutting edge solutions and advice.


The Laboratory

The lab is a subunit of UCL - Institute of Finance and Technology (IFT). It is led by Dr Ramin Okhrati, who is a full-time faculty member of IFT, and supported by Prof Francesca Medda as IFT’s director. Throughout the development of the lab, IFT will provide support and assistance in both financial and intellectual terms. In summary, the lab will be an intellectual hub of innovative ideas and critical analysis of AI modelling in risk management.


Current Projects

Offline Reinforcement Learning for Contractual Pricing (Natwest Group)

Project Title: Offline Reinforcement Learning for Contractual Pricing
Sponsored by: NatWest Group
Lead Investigators: Raad Khraishi (PhD Candidate and Data Scientist, NatWest Group), Dr Ramin Okhrati (Supervisor, IFT), Prof Francesca Medda (Supervisor, IFT)
Starting Date:  September, 2021
Project summary: Raad is investigating offline reinforcement learning approaches to pricing contractual products. Traditional online reinforcement learning is impractical in many pricing applications given potential for mispricing as an agent learns from live interactions through trial and error. He is currently exploring offline techniques that allow an agent to learn a pricing policy using historical data with online fine-tuning.

Anomaly Detection and Clustering of Trading History (Consob)

Project Title: Anomaly Detection and Clustering of Trading History
Sponsored byCommissione Nazionale per le Società e la Borsa (Consob)
Lead Investigators: PhD Candidate (TBC), Prof Francesca Medda (Supervisor, IFT), Dr Ramin Okhrati (Supervisor, IFT)
Starting Date: (TBC)
Project summary: The analysis will permit the classification of all the investors under scrutiny – investors selected using simple and traditional detection alerts – into different clusters in order to assess – through an algorithm – the probability that the behavior under observation is/is not a case of potential misconduct (in so far as deserving more in-depth analysis) by simply confronting the trading under alert with the whole trading history of the same investor.

 

Underlying Risks and Financial Resilience: Gamestop Case (Consob)

Project Title: Underlying Risks and Financial Resilience: Gamestop Case
Sponsored by: Commissione Nazionale per le Società e la Borsa (Consob)
Lead Investigators: PhD Candidate (TBC), Prof Francesca Medda (Supervisor, IFT), Dr Ramin Okhrati (Supervisor, IFT)
Starting Date: (TBC)
Project summary: Network science provides the instruments to address the problem of amplification and propagations of small shocks that hits the market system. We will complement social network data in order to identify coordinated price manipulation schemes (pump and dump, Ponzi), as was recently done for cryptocurrencies.