Informational Advantages of Euronext’s Trading System: Colocation and Private Feeds

In October 2020 the Dutch financial daily newspaper Het Financieele Dagblad published an article (‘Euronext provides high frequency traders with an invisible advantage’) in which it revealed that Euronext Amsterdam grants certain traders informational advantages over other traders. According to it, the trading system deployed by Euronext’s exchanges in Amsterdam, Brussels, Dublin, Lisbon and Paris, called Optiq, provides traders involved in a transaction with early data information concerning that transaction (eg trade confirmation) prior to the traders who are not involved in that transaction. The parties profiting from the informational advantage of just a few microseconds are traders, such as high-frequency traders, who are cognizant of the discrepancy and have the financial and technological capacities to exploit it to their advantage. Although Euronext stated that it would adjust its trading system shortly after the existence of this information asymmetry had reached the public domain, Euronext’s CEO did not appear to acknowledge the urgency to do so. Moreover, he stated that Euronext complies with all relevant European laws and regulations (‘Euronext under fire after problems with trading system’). We think that this statement is at least debatable.

As a starting point, consider that, in an era where the speed of trading approaches the speed of light and the typical trader found in the trading arena uses automated trading algorithms constantly rapid-firing orders, even speed advantages of just a few microseconds are crucial. Traders who receive information earlier than other traders have the opportunity to make more real-time estimates of supply and demand (including block orders), and future price trends. They can make a profit or prevent a loss based on the early information (eg by sending new orders, or by modifying or cancelling current orders), before the other traders can respond to the information. Although the informational advantage in case comprises only a few microseconds, it should be noted that the traders receiving the early information send, modify and cancel millions of orders a day, so that the informational advantage potentially adds to substantial profits. The fact that traders involved in a transaction for a split second possess more information than traders who are not involved therein can make a world of difference in the trading pits of the 21st century, especially considering that not all traders have been made aware of this reality by the trading venue involved.

Through a private feed, a colocated trader receives information (such as post-trade information related to transactions) earlier than traders who do not make use of colocation services. It could be argued that it makes sense that traders who make use of, and pay for, colocation services receive information earlier than traders who do not make use of those services. Offering colocation services and a private feed is even allowed by European laws and regulations (see, implicitly, Art. 48(8) MiFID II and Art. 6(1) and 13(1) MiFIR). The recent controversy surrounding Euronext’s trading system, however, does not in fact concern the informational deficit of non-colocated traders compared to colocated traders, but rather the informational disparity amongst colocated traders themselves. This is because colocated traders involved in a transaction receive information about that transaction via their private feeds earlier than other (colocated) traders who are not involved therein.

Trading venues like Euronext have to be transparent when providing colocation services (see Art. 48(8) MiFID II). The fact that Euronext has acknowledged that it is familiar with the difference in latency between colocated traders but did not make this difference public before the revelation in Dutch newspapers could therefore be remarkable in itself. Perhaps even more remarkable, however, is the unequal treatment of users that have subscribed to the same colocation services. European regulations oblige trading venues to provide users that have subscribed to the same colocation services with access to their network under the same conditions, including as regards, amongst other things, access to data and market connectivity (see Art. 1(2) Delegated Regulation 2017/573). Furthermore, trading venues have to take all reasonable steps to monitor all connections and latency measurements to ensure the non-discriminatory treatment of all users of colocation services that have the same type of latency access (see Art. 1(3) Delegated Regulation 2017/573). Moreover, trading venues are obliged to publish any differences in latency types on their website (see, in particular, Art. 2(a) and 2(d) Delegated Regulation 2017/573). Admittedly, shortly after the publication in Het Financieele Dagblad, Euronext published a statement in which it acknowledged the discrepancy. However, the document is only available in Dutch, which makes it doubtful whether Euronext is in fact fully transparent about its colocation services.

A different but interesting question is whether a trader trading while in possession of early information about a barely realized transaction acquired through a private feed violates the ban on insider dealing (see Art. 14(a) in conjunction with Arts. 7(1)(a) and 8 MAR). In practical terms, only the two traders involved in the transaction receive the information concerning that transaction before other traders receive it. Hence, they, albeit momentarily, are in possession of more information than other (both colocated and non-colocated) traders. In our view, there are compelling arguments to assert that early information about a transaction, which is inaccessible to, and cannot be lawfully obtained by, the rest of the investing public, is not ‘public’, and that trading while in possession of such information harms the level playing field between traders that the European legislator endeavours to guarantee by imposing the insider dealing ban. Traders that trade while being in possession of such information can derive an unfair advantage from that information vis-à-vis the rest of the investing public who are unaware of that information. Therefore, it is plausible that traders who are in possession of such early information, before it is received by other colocated traders, and who trade while in possession of that information, violate the ban on insider dealing. More generally, based on the same line of reasoning, it could even be argued that all early information acquired through colocation services constitutes inside information, and that by trading on the basis of this information the ban on insider dealing is being systematically violated. Although the European rules allow for colocation, and making use of colocation is even part of the definition of high frequency trading in Art. 4(40)(a) MiFID II, it is questionable whether all forms of informational and speed advantages derived from using colocation services and private feeds are actually permitted (compare Myklebust 2020, p. 70, contending that a definition does not determine what constitutes legal practice). Indeed, the European rules do not exempt traders using early information acquired through a private feed from the ban on insider dealing. Nevertheless, the possibility that the practice of using such information in fact violates the insider dealing ban has not been assessed by the European legislator either in the reform process leading to the MiFID II/MiFIR regime or during the reform process that resulted in the enactment of the Market Abuse Regulation. Yet, granting colocated traders informational advantages inevitably contributes to the emergence of a tiered system on the securities markets, and greatly impacts the public’s perception of the integrity of those markets.

The revelation of the informational advantage is unfortunately reminiscent of the US practice where high-frequency traders have obtained substantial advantages over the average trader from traditional exchanges and alternative trading systems. In the US, high-frequency trading lobbyists have not only lobbied for the development of latency advantages, but also for the development of special order types especially tailored to the needs of high-frequency traders, which were often unavailable to other traders and even kept secret. Similar controversial tales currently remain absent in the EU. Let us hope that European trading platforms do not have other forms of information asymmetry that have yet to be uncovered. Especially in an era where controversies related to high-frequency trading are a matter of public debate, the disclosure of adequate information related to the actual workings of trading platforms is essential.

Although Euronext has declared that it would adjust its trading system (‘Euronext adjusts the trading system to prevent insider dealing’), in its statement following the publication in Het Financieele Dagblad it also stated that the perfect synchronisation of information dissemination between traders involved in a transaction and traders not involved therein is, ‘by definition’, impossible, and that a potential information asymmetry is present on ‘all trading platforms’. Frankly, similar issues involving trading platforms other than Euronext Amsterdam have not been raised yet. Nonetheless, in a recent consultation paper the European Securities and Markets Authority (ESMA) requested market participants to express their views on the information asymmetry between traders involved in a transaction as compared to traders who are not involved therein. As private feeds have been a phenomenon on the EU securities markets for over a decade, this consultation seems long overdue.

Mathijs Giltjes is a PhD candidate at Erasmus Graduate School of Law and fellow of the International Center for Financial law & Governance (ICFG).

Arnoud Pijls is Assistant Professor of Company Law & Capital Markets Law at Erasmus School of Law and fellow of the International Center for Financial law & Governance (ICFG).