SEMINAR:The QLBS Model within the presence of feedback loops through...

Guest: Umur Özsoy

Title: The QLBS Model within the presence of feedback loops through the impacts of a large trader

Time: March 20, 2024, 13:40 - 14:30 

Location: FENS G032

Abstract: The seminar will start with the concept and implications of the existence of a large trader. To grasp the regulatory side of the concept of a large trader, we will go back a few decades in history to tell what has changed in the front office vs back office modeling perspectives.  Afterwards, we will introduce our approach to a large trader as an agent making sequential decisions in a reinforcement learning setting. We will explore our reformulated version of the QLBS model, one of the early examples of machine learning approach in option pricing,  and show that we obtain an optimal hedging strategy  for the large trader so that there will be much lower transactions costs, and possible less regulatory issues in leaving an impact in the course of the market. We will conclude by discussing the possible contributions of reinforcement learning could offer in option pricing under market impacts as well. 

Bio: Dr. Umur Özsoy graduated with a bachelor in industrial engineering from TOBB University of Economics and Technology. To receive more focused training and further education with his ongoing interest in matters of quantitative finance and trading, he enrolled at the Institute of Applied Mathematics at Middle East Technical University. There he completed his master and doctorate degrees in Financial Mathematics. Dr. Özsoy enjoys research topics such as computational finance, machine learning in finance and algorithmic trading. Currently, he works as a researcher at Turkish Management Sciences Institute (TÜSSİDE) of TÜBİTAK.