Exotic options, in general, have very time-consuming valuation models. I believe in recent years there has been some research done on using supervised machine/deep learning to predict the valuation given a training set of valuation done using traditional methods and other training features like greeks, market data etc.

Can someone point me to relevant research papers in this domain? Not necessarily for exotics, but vanilla would be good too.

  • $\begingroup$ I only know of this paper, where they apply machine learning techniques to common quant problems. It should give you some insight into the applicability of machine learning on problems within quantitative finance. $\endgroup$ – Pleb Jan 12 at 18:01
  • $\begingroup$ Josef Teichmann from ETHZ has done some work in this area. $\endgroup$ – g g Jan 12 at 19:11
  • $\begingroup$ arXiv (arxiv.org) and Google Scholar are two good ways to find papers. For example, using the keywords "deep learning exotic pricing" (in the "Quantitative Finance" subject area) turns up two references on arXiv and several more on Scholar. $\endgroup$ – Sharad Jan 12 at 20:16
  • $\begingroup$ @pleb this is something different from what I’m looking for. g g: That link has papers on deep hedging and deep calibration, couldn’t find anything on deep pricing. Sharad That is an option, but it’s so easy to get lost in wealth of papers on this topic especially since it’s such a hot area, looking for people who have already done some literature review. $\endgroup$ – Dhruv Mahajan Jan 12 at 20:25
  • $\begingroup$ JP Morgan have been publishing and talking about 'Deep Hedging' for a little while - basically trying to solve hedging from a data-driven angle Vs the convention model-based angle (eg. this paper: papers.ssrn.com/sol3/papers.cfm?abstract_id=3355706) $\endgroup$ – StackG Jan 12 at 22:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.