Particle physics helps to visualise manipulation in JPMorgan spoofing case

Published on
February 22, 2022

Researchers from Wageningen University & Research, the European Organization for Nuclear Research (CERN), and The Commodity Risk Management Expertise Centre (CORMEC) have developed a unique visualization method for financial markets using particle physics tools. With this tool, financial institutions or regulators can analyse all activity on the trade floor and dissect market manipulation, such as spoofing, in detail.

Spoofing is a form of manipulation where a trader intentionally manipulates the market by placing orders they intend to cancel, hoping prices move to improve their market position. One of the largest banks in America, JPMorgan, paid a record-breaking settlement of $920.2 million in 2020 for spoofing markets from 2008 to 2016. The order of the regulator, the Commodity Futures Trading Commission (CFTC), outlined nine detailed examples how JPMorgan manipulated the market by this spoofing method.

The researchers used their new visualization method to analyse these spoofing examples in detail. The method uses data-analysis framework ROOT, which is originally designed for particle physics analysis by among others CERN. How does it work? The researchers visualised the limit order book (LOB) of the markets JPMorgan manipulated. The LOB shows all the prices traders are willing to buy or sell a financial product for, including the quantities they want to buy or sell.

Milli- or nanoseconds

Marjolein Verhulst, author of the article Unravelling the JPMorgan Spoofing Case Using Particle Physics Visualization Methods published in European Financial Management: ’The traditional way of looking at LOB data in close detail is to take snapshots. You can compare snapshots to taking pictures: each second, or minute, you take a picture of the market at that particular moment. However, you do not see what happens in between these pictures. For some research this is fine, but we are looking at markets with high-frequency traders who sometimes trade in milliseconds or nanoseconds. Market manipulations and high-frequency trading activities are often done at such a high pace, taking snapshots does not show what really happens.’

To illustrate, the spoof orders from the nine JPMorgan examples were visible in the LOB between 0.31 and 5.56 seconds. Marjolein: ‘With the use of particle physics, we are now able to see in detail what happens in between the snapshots. We can now visualise all market activity and have the full picture.’

Spoofing further explained

Spoofing has been illegal under the Dodd-Frank act since 2010 and is defined as “bidding or offering with the intent to cancel the bid or offer before execution”. The word “intent” is key here: it involves placing an order that is not intended to result in an actual trade. A simplified example of how spoofing works is as follows. Let’s say we have a trader who bought a corn futures contract, and they would like to sell it for a higher price using spoofing. They place a large spoof order on the buy (bid) side of the market. This is a spoof order because they do not actually want to buy more corn contracts: they want to create a false impression that lots of people want to buy (so the price might rise). To not miss out on an increasing price, other market participants react to this false information by buying which in turn increases the price. The spoofing trader then sells their corn futures contract for a higher price than before and cancels their large spoof order. Thus, the spoof order never resulted in a trade and was used to move the market. Since the high buy pressure is now gone, the price will decrease to its previous level and the trader succeeded in getting a higher price for their corn futures contract than before.

More understanding and research into spoofing and market manipulation

The researchers found several indicators from the JPMorgan case suggesting that in some circumstances, JPMorgan did not spoof to move the price, but to attract more traders to the market (attract liquidity). This has been unreported in science so far. With the new visualization method and deep dive into the JPMorgan case, the researchers hope to give more insights into spoofing and what happens to the markets when spoofing occurs. The researchers hope that this research helps all stakeholders, such as regulators, the exchange and market participants, to improve their understanding of spoofing and its effect on the markets. Moreover, they hope the visualization will function as a steppingstone for further research on market manipulation.

The researchers would like to give a special thanks to CERN, CORMEC, the Province of Limburg, the Chicago Mercantile Exchange Foundation, the Office for Futures and Options Research at the University of Illinois at Urbana-Champaign, the Dutch Authority for the Financial Markets (AFM), Euronext, the Dutch National Bank (DNB) and SURF SARA.