Podcast: Hans Buehler on the data science behind deep hedging
Top JP Morgan quant stresses importance of ‘de-trending’ training datasets used in machine learning
Top JP Morgan quant stresses importance of ‘de-trending’ training datasets used in machine learning
Machine learning systems are only as smart as the data used to train them. Train a neural network on equity market data from a bull market, and its simulation will undoubtedly display an upward drift.
“If you ask a machine to build a trading strategy for an empty portfolio . . . it will start selling puts and buying delta,” says Hans Buehler, global head of equities analytics, automation and optimisation at JP Morgan, and Risk.net’s quant of the year for 2022.
Of course, such a strategy would have produced significant losses this year, when US equity indexes suffered their first correction since the beginning of the Covid-19 pandemic. Removing the drift in financial data series is therefore a prerequisite to training trading strategies, especially hedging algorithms. In a recent paper, Buehler and his co-authors Phillip Murray, Mikko Pakkanen and Ben Wood describe a technique for doing exactly that.
In this episode of Quantcast, Buehler explains the importance of removing drift in financial data, and discusses the general features and uses of deep hedging, as well as his future research projects. He also reveals how the origins of deep hedging can be traced back to his time as a PhD student at the Technical University of Berlin.
Buehler’s research has come a long way since then. JP Morgan already uses a version of deep hedging to price a large chunk of its Eurostoxx vanilla index options book and an increasing portion of S&P 500 options trades. Deep hedging is now being tested for cliquet options and the bank plans to expand its use to single stocks by year-end. Autocallable products, for which deep hedging was originally designed, are likely to be the next product to be priced this way, Buehler says.
Towards the end of the podcast, Buehler shares his ambition to introduce modern machine learning and AI-based methods to JP Morgan’s sales team, and further strengthen the bank’s electronic trading capabilities.
To hear the full interview, listen in the player above, or download. Future podcasts in our Quantcast series will be uploaded to Risk.net. You can also visit the main page here to access all tracks, or go to the iTunes store or Google Podcasts to listen and subscribe.
Now also available on Spotify and Amazon Music.
Index
00:00 What is deep hedging and what is it used for?
06:30 The importance of removing drift from training datasets
09:03 Application of the de-trending technique
11:25 The origins of deep hedging
15:15 Future research projects
Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.
To access these options, along with all other subscription benefits, please contact info@risk.net or view our subscription options here: http://subscriptions.risk.net/subscribe
You are currently unable to print this content. Please contact info@risk.net to find out more.
You are currently unable to copy this content. Please contact info@risk.net to find out more.
Copyright Infopro Digital Limited. All rights reserved.
You may share this content using our article tools. Printing this content is for the sole use of the Authorised User (named subscriber), as outlined in our terms and conditions - https://www.infopro-insight.com/terms-conditions/insight-subscriptions/
If you would like to purchase additional rights please email info@risk.net
Copyright Infopro Digital Limited. All rights reserved.
You may share this content using our article tools. Copying this content is for the sole use of the Authorised User (named subscriber), as outlined in our terms and conditions - https://www.infopro-insight.com/terms-conditions/insight-subscriptions/
If you would like to purchase additional rights please email info@risk.net
More on Cutting Edge
Volatility shape-shifters: arbitrage-free shaping of implied volatility surfaces
Manipulating implied volatility surfaces using optimal transport theory has several applications
Georgios Skoufis on RFRs, convexity adjustments and Sabr
Bloomberg quant discusses his new approach for calculating convexity adjustments for RFR swaps
Infrequent MtM reduces neither value-at-risk nor backtesting exceptions
Frequency of repricing impacts volatility and correlation measures
Dynamic margining long/short equity trading strategies
A repo haircut model extends a previous solution for long-only strategies
SABR convexity adjustment for an arithmetic average RFR swap
A model-independent convexity adjustment for interest rate swaps is introduced
Joint S&P 500/VIX smile calibration in discrete and continuous time
An arbitrage-free model for exotic options that captures smiles and futures is presented
The carbon equivalence principle: minimising the cost to carbon net zero
A method to align incentives with sustainability in financial markets is introduced
The cost of mis-specifying price impact
Expected returns can be significantly affected by the wrong use of impact models
Most read
- As FCMs dwindle, regulators fear systemic risk
- Top 10 operational risks for 2024
- Top 10 op risks: AI fears drive cyber risk to record high