Blog article
See all stories »

Momentum Ignition: Arson for Financial Markets

ESMA has defined momentum ignition as a case of prohibited market manipulation. ICE Canada considers this to be disruptive, and so does the CME Group. One UK government report stated that momentum ignition will “induce algorithms to compete with other algorithms, can push prices away from fundamental values.”*

This leads us to the question – what exactly is momentum ignition? And how do we detect that is has occurred so that compliance teams can surveill for this prohibited activity?


The market & contract

The fictitious Jelly Bean futures contract is trading without much volatility. Hedgers and speculators are participating daily. There is book depth. Automated market tools at numerous buy side and sell side firms are participating.   

The ignited market

All of a sudden there is a large spike in filled contracts. The price jumps. Algos seeing the price change suddenly change their bid/ask quotes. Other algos respond. Accumulation strategies (e.g. VWAP) increase their market involvement. Volatility and volume increases. Price continues to change. Day traders seeing the change quickly jump in to benefit from the price move. In other words, there is an intraday mini-run on jelly beans.

The market effect

Eventually, the market participation and price settles back down to the price it was before. Without the spike in volume from the trader who “ignited” the market, the price will settle to a more market driven fundamental value.

The effect on market participants

  1. The traders who jumped in at the tail end of the ignition have been significantly disadvantaged as the price quickly dropped once the original “igniter” dropped out.
  2. The accumulation strategies, often working on behalf of fund groups, have paid a disadvantaged price.
  3. Algos have been inappropriately tricked into trading, which affects the market making buy and sell side firms.

In sum, the original igniter most likely got out of their position for a profit at the expense of those who were tricked into participation.


This scenario, while fictitious, is absolutely possible in exchange traded derivatives markets. It isn’t within the exchange rules, but it is possible. A good compliance team will be watching for this form of manipulation.

How could this have happened? The trading controls at broker-dealers prevent this, don’t they? Do exchanges have technology to prevent this?

  • Yes, the BDs have pre-trade risk controls.
  • Yes, the exchanges also have pre-trade controls which help protect the markets.

Yet these controls work to check the financial ability of the trader or the account to participate in the market. They don’t provide controls against intentional market manipulation. They can’t discern the intent of the trader.


The problem in detecting momentum ignition lies in discerning between a market move created intentionally to entice and disadvantage other participants, and a real investment. Put simply, the compliance analyst needs to discern the intent of the original trader.  

Machine learning, a form of Artificial Intelligence, provides some very positive results in detecting this complex market manipulation. Here’s how:

  1. Rather than asking a series of boolean questions attempting to discern intent on a trade by trade basis, machine learning uses a series of statistical features to determine the likelihood that the activity was an improper momentum ignition.
  2. Instead of looking on a trade by trade basis, machine learning approaches the review by looking at a cluster of related trading transactions. It reviews groups, not single transactions.
  3. Machine learning uses pattern detection techniques to see how well the collected statistical features match, or don’t match identified cases of momentum ignition.

Machine learning is a very current and relevant way to overcome the systemic problem of surveilling for momentum ignition.

Thankfully, our jelly beans are safe from ignition.


*Foresight: The Future of Computer Trading in Financial Markets (2012) Final Project Report The Government Office for Science, London



Comments: (0)