Study: Market 'Collusion' in ETF Trades on Euro Exchange

Market makers trading ETFs on the Euronext Amsterdam are breaking the rules by colluding together to protect themselves from competition, according to new academic research conducted by the University of Oxford.

Massive growth in the adoption of ETFs globally has provided lucrative new profit streams for market makers who play an essential role in the efficient working of modern stock markets as middle-men that always stand ready to buy securities from sellers and to sell securities to buyers.

Strict conduct rules apply to all market makers to ensure that they treat customers fairly and to prevent them from acting in concert together to take advantage of investors.

However, the Oxford study, Anonymity, Signaling, and Collusion in Limit Order Books, found these rules are being flouted by ETF market makers on the Euronext Amsterdam exchange, who are sending signals about their order books to each other when the identity of the holders of these positions are supposed to be anonymous.

“The activity of ETF market makers on Euronext Amsterdam is consistent with collusion,” said Alvaro Cartea, Professor of Mathematical Finance and Director of the Oxford-Man Institute of Quantitative Finance.

ETF market makers tend to submit buy and sell orders that are very large in comparison to orders that originate from retail investors on Euronext Amsterdam. This size difference allows market makers to infer whether an order came from a rival market maker or another type of investor. Market makers then almost entirely avoid competing for large positions which they have inferred are held by other market makers.

Market Makers Involved in ETF 'Sniping' 

But the much smaller order instructions that originate from retail investors are much more likely to be taken out within milliseconds or “sniped” by a market maker aiming to win the business by offering a better buying or selling price than a rival trader.

“The market makers hardly touch each other’s ETF positions. But they nearly always go for the ETF flows from retail investors. A spread-improving limit order sent by a retail investor is over 2,000 times more likely to be sniped than a spread-improving limit order sent by another market maker,” said Cartea.

Market makers account for 82.2% of ETF orders, predominantly those from retail traders. Less than 5% of the snipes by market makers were in reaction to a limit order that improved the buying and selling price spread for an ETF made by another market maker. Retail investors do not have the sophisticated trading infrastructure needed to snipe ETF orders from other traders.

Retail investors are a much more important source of order flow for ETFs than for stocks on Euronext Amsterdam. Around 70% of transactions involving ETFs included a retail investor as either a buyer or seller or both. This dropped to just 10% for transactions involving shares.

The findings were based on a study of the trading activity of the 367 ETFs listed on Euronext Amsterdam in January 2022. During that month, there were 425 million order entries in ETFs, resulting in 390,628 ETF transactions which were together worth around €3.2 billion ($3.3 billion).

Cartea said his team looked at other possible explanations for market makers' trading activity but “none could fully explain” the behavior observed in the Euronext Amsterdam ETF market.

A spokesperson for Euronext Amsterdam said that it did not comment on the outcomes of research.

The Association of Proprietary Traders, a trade body representing market makers, said that the research "fundamentally misunderstands the legitimate activities of market makers and fails to provide factual evidence of collusion, relying instead on hypothetical scenarios."

"It is normal for market maker liquidity to be generally larger in size and symmetrical on both sides of the book and as a consequence readily identifiable. The reason market makers typically trade more frequently with retail orders is because retail orders are more likely to prioritise immediate execution. Market makers, by contrast, independently reach their own opinions on the theoretical price of an ETF by considering complex data inputs, making them more price-sensitive. The result is that a market maker is much more likely to trade with a retail order than with another market maker," said Matthijs Pars, director-general of the Association of Proprietary Traders.

One of the co-authors of the paper, Rob Graumans, works as a data scientist for the Autoriteit Financiele Markten (AFM), the Amsterdam-based regulator for financial markets in the Netherlands, which is aware of the findings.

An AFM spokesperson said the regulator was “supportive of academic research as a way to increase our collective understanding of the financial markets.”

This article was written by Chris Flood and originally published on our sister site, etfstream.com.