At a glance: Algorithmic trading regulatory review in Europe

On 18 December, ESMA published its anticipated consultation paperopens in a new tab on the impact of requirements under MiFID II/MiFIR regarding algorithmic trading. The paper is based on a holistic approach to algorithmic trading and reviews all related MiFID II requirements together (e.g. direct electronic access or tick sizes), with the aim of improving the existing regulatory framework.

Parts of the proposal may have significant business, operational and cost impact on firms engaged in algorithmic trading. Below we set out a brief summary of the paper: 

Areas of focus

Extending scope to SIs. ESMA proposes to extend the definition of algorithmic trading to include trading in financial instruments OTC by systematic internalisers (SI) and selectively apply some of the requirements currently set out in Level 2. Key requirements at SI level for OTC algorithmic trading would include (i) governance arrangements for trading systems and trading algorithms, (ii) controlled deployment of algorithms (iii) kill functionality and other risks controls.

DEA. As per ESMA’s proposal, persons having DEA access to only deal on own account (DEA clients) would no longer need to be authorised as investment firms, except where they qualify as HFT firms. Also, to address the uncertainty that has arisen in relation to DEA sub-delegation and Tier 2 clients, ESMA proposes to amend the definition of DEA to include DEA sub-delegation. This would allow clarification around which Tier 2 clients should be considered DEA users for the purposes of MiFID II obligations relating to DEA.

HFT. Since current HFT definition is based on high message intraday rates, ESMA will be seeking input to understand whether static daily message rates remain a relevant criterion or whether other approaches should be considered. In addition to this, ESMA also intends to have more stringent rules for third-country HFT firms by introducing a requirement to be authorised as an investment firm when they qualify as an HFT firm on an EU trading venue.

The most significant scoping change would be related to the SI extension, however, the impact will depend on the size of a firm. While many large banks already applied these requirements to their OTC business, it may have greater impact on those who have previously excluded such algorithms from their governance and control arrangements. Also, it may bring in scope third-country HFT firms using DEA to access EU venues who will need to seek additional authorisation. 

Testing and testing environments. ESMA proposes to clarify the meaning of testing for “disorderly trading conditions”, which should refer to a market where the maintenance of a fair, orderly and transparent execution of trades is compromised. Also, the regulator considers that it would be useful to produce additional guidance covering the expectations for behavioural testing. Such clarifications could indicate that this type of behavioural testing should focus on the interaction between the tested algorithm and the market, and that it should in particular detect whether the tested algorithm contributes to amplification of market movements that are unrelated to real economic value.

RTS 6 self-assessment and validation. ESMA notes that currently there are very different outcomes in the level of detail of the submitted self-assessments and considers this exercise to be a proper due diligence assessment, noting that it should be more ambitious than a statement of compliance. To increase the quality of these assessments, ESMA proposes to:

• consider a specific format for self-assessment, harmonised at EU level
• ask investment firms to submit their self-assessment to their local regulators for review
• amend frequency of self-assessment to every two years (except stress testing requirements)
• require investment firms to report more information on testing environments

ESMA is also of the view that similar changes should be made to the RTS 7 self-assessments produced by trading venues.

Orders to transactions ratio (OTR) for venues. ESMA’s analysis suggests that at the moment there is a very significant divergence in terms of maximum OTR limits allowed across different trading venues. Therefore, it proposes to develop technical standards to set out the maximum OTR ratio, calibrated per asset class.

Monitoring of compliance with exchange rules. Given a number of operational incidents which took place in 2020 that led to significant halts of trading across a number of exchanges, ESMA would like to improve the process surrounding the monitoring of compliance with exchange rules. It proposes to develop a procedure to enable efficient and timely notification from trading venues to their local regulators and ESMA in case of IT incidents and systems outages.

Data for traders. Another proposal from the regulator is to require algorithmic traders to always use at least two different reference points to ensure there is always a possibility for trading activity to migrate from the main market to another trading venue in the case of an outage.

This section of the paper may potentially result in quite a significant uplift required by firms and deserves greater attention. While the industry would welcome lowering the frequency of self-assessments to 2 years, ESMA is considering a new obligation to report the self-assessment results to regulators. Also, more prescriptive and detailed testing requirements and using additional sources of data for traders would need to be looked at more closely as the draft rules are presented by ESMA.

ESMA intends to streamline the MiFID II market making regime by the following:

• Limit the application of its market making requirements to continuous order trading books  
• Broaden the obligation of having market making schemes to all instruments and types of trading systems

While the regime may be potentially extended to additional products, any changes should not pose significant operational challenges for firms. 

The regulator is also keen to explore some recent concepts such as the introduction and functioning of speedbumps (mechanisms implemented by some exchanges, consisting of a delay applied to incoming orders before they enter the matching engine for execution). Overall, the aim of speedbumps is to regulate the speed of high-frequency traders and to curb ultrafast trading strategies. While this mechanism is already approved in the US and Canada, it is more recent in Europe (e.g. the FCA granted its first approval to the LME in March 2020). ESMA will consider whether venues which introduce speedbumps, should be required to tighten their market making requirements to ensure the liquidity is sufficient.

Also, ESMA is considering the impact of public and private transaction data feeds. The current discussion revolves around the discrepancy in the timing of the two feeds, and on market participants not receiving market data at the same time. ESMA’s proposal is to assess the implications of the timing of public trade messages versus private fill confirmations and the advantages and disadvantages of one preceding the other.

For the time being it is difficult to conclude whether this area will materialise in a concrete proposal by the regulator. Given the increasing usage of speedbumps, businesses should assess commercial impacts of such proposals as they emerge to understand the implications for their execution strategies.