Last month the Financial Conduct Authority (“FCA”) released a report (“the Report”) detailing the findings of its thematic review into the use of UK equity market dark pools (“the Review”).1 To date, the use of dark pools, and the related rise in high-frequency trading, has intrigued the mainstream press and attracted the attention of Regulators globally, particularly the US Securities and Exchange Commission (“SEC”). Many of the issues and conduct identified in recent SEC actions are addressed by the FCA in the Report. Firms, both users and operators of dark pools, may be advised to revisit their procedures and processes in light of the concerns raised and guidance given by the FCA. The Report may also be of interest to any individual Senior Management Function holder who has overall responsibility for their firm’s algorithmic trading or systems, as well as any person who will be certified as holding the algorithmic trading significant harm function.2 This article highlights some key regulatory risks, particularly those faced by broker-dealer users of dark pools who are subject to best-execution obligations.
Scope of the Review
The FCA’s Review was broad and did not involve a detailed examination of trading activity. Instead the findings are based on information and discussions gathered from a sample of both users and dark pool operators. The Report identifies areas of concern and suggests systems and procedures which would protect against related risks. For the purposes of its Review, the FCA defined a dark pool as “a trading venue with no pre-trade transparency, in that all orders are hidden as to price and volume and are anonymous”.
Drivers of dark pool use
Before summarising the instructive aspects of the Report, it is interesting to rehearse the benefits and reasons cited by firms for using dark pools. A key driver for their use is a desire to protect against order information causing a fluctuation in price, before that order is filled. Access to dark liquidity is a solution: the order information is not publicly known. Other attractions to the use of dark pools were found to be:
- the opportunity to access additional sources of liquidity;
- the increased potential for bid-offer spread capture3; and
- the reduced risk of information leakage about an order or a trading strategy (i.e. sensitive knowledge about a party’s intentions to buy or sell a share).
The Report found that the lack of price transparency was not generally a concern for users because the prices in dark pools are derived from other transparent markets. The reference price is typically the current mid-price of the primary market’s best bid-offer prices. Accordingly, it is uncommon for a price to be outside the best bid-offer range and in any event pool operators were considered by users to be generally responsive to trade queries should any outlying trade occur. However, users’ comfort in relation to the lack of price transparency hinged on the proportion of overall trading that dark pool activity represented, which is currently quite low. If it were to increase, and thereby represent a significant proportion of the equity traded, the price would be determined by a comparatively small number of trades in an illiquid market. Whilst the Report comments that the issue “needs to be monitored”, it does not express how the FCA would react, nor how it would expect operators and/ or users to react, if indeed the proportion of trading on dark pools reached a critical mass. It should be noted however that, under the provisions of MiFID II, the volume of dark trading is set to be capped as a proportion of the overall equity trading, thereby addressing this concern.
Due diligence on pools
The Review found that users did not adequately either perform due diligence in selecting a dark pool, or consider the technical or strategic rationale. The Report states that any decision to execute orders on a given pool should be based on “thorough due diligence and a clear understanding of the attributes of each pool including how these features address trading needs and ultimately benefit customers”. As well as emphasising the need to formulate a ‘clear rationale’, the Report impresses upon users the need to keep that rationale under regular review. Adherence to these principles is considered essential in meeting the best-execution obligations, under which firms must take all reasonable steps to gain the best possible result for their clients based on the execution factors.4
Monitoring transactions and data exchange
The Report also comments on the importance of monitoring completed dark pool transactions in order to assess whether they have met best-execution standards. The FCA’s Review revealed that some firms do not compare trades executed in dark pools with market data feeds and thereby fail to identify outliers. Some firms did conduct a comparison, but not promptly. Other firms were equipped with inadequately robust monitoring systems. The Report suggests that users may not be able to rely on operators to bring outliers to their attention. Users should therefore probably seek clarification from operators as to what level of internal monitoring has been implemented. Ultimately a user should assure itself that its governance and oversight processes enable adequate analysis of its performance in dark pools.
One key aspect of achieving that objective is having accurate data against which one can compare the public market information. Being provided with a highly precise and accurate record of the timing of any transaction is particularly important. The Report noted that some operators “remained out of step by not providing data reports at the millisecond clock-speed level (at least) and this made analysis including market abuse oversight difficult”. Other limitations in information exchange were also identified as compromising a firm’s ability to identify market abuse. For example, the Review found that most operators did not have the ability to analyse and report on individual client transaction-level trading activity and that the users could wait up to a week to receive reports from operators. The Report stresses that monitoring and prompt corrective action is crucial to the prevention of market abuse and to minimising the risk of clients receiving poor outcomes.
The Report expresses some concerns surrounding order-routing. First, in circumstances where an operator routes a trade to an alternative venue, for example under a reciprocal agreement, there is a risk that a user’s agreed preferences will not be preserved. The Report suggests that, even where preferences are communicated, there is no way of knowing whether the client’s order has been filled by a restricted counterparty. Users should therefore make themselves aware of their chosen operator’s policy on order-routing, and how its systems are adapted to mitigate this risk.
Second, a conflict of interest may arise where a firm, which operates a dark pool, is effecting trades on a client’s behalf. Where that trade is routed—whether to the operator’s own dark pool or an alternative source of liquidity—will depend in part on the client’s own execution strategy. However, the Review found that bank operators consistently routed orders to their own pool first. This preference was typically justified on grounds that it optimised execution with lower cost and mitigated the risks of information leakage. The Report recognises that this practice is acceptable, provided an operator meets its best-execution obligations and does not discriminate as to which venue is used. Moreover, the Report reminds operators of the need to “support and evidence” the relevant factors by “actively monitoring trade activity.”
Alternative sources of liquidity have been increasingly seen as desirable, despite the transparency and regulatory concerns that have thus far emerged. There seems little evidence that this will change. Their expansion in use was bound to attract attention—the FCA’s Review is itself evidence of this. As with any thematic review, the FCA will expect that users and operators of dark pools take note of the concerns and related recommendations made in the Report. Although the recommendations and guidance do not have the status of ‘rules’, they allow firms to appreciate those areas in which the FCA considers the rules could be compromised. This is particularly true in relation to best-execution obligations. As the Report notes, the provisions under MiFID II will have a significant impact on wholesale equity markets and will cap the use of dark liquidity. However, until the post-Brexit fog starts to disperse, the extent of and direction of any change in the UK will be unclear.
2 From 7 September 2016 algorithmic trading will be classed as an FCA-specified significant-harm function under the rules governing Senior Management Arrangements, Systems and Controls (SYSC) (see SYSC 52.49). Accordingly, a firm will obliged to certify that any person performing that function is fit and proper to do so.
3 The bid-offer spread capture is the amount saved by buying or selling within the quoted spread. For example, an execution at mid-point would represent 50% spread capture for both the buyer and the seller.
4 See the Conduct of Business Rules (COBS) rule 11.2. The execution factors are: price, cost, speed, likelihood of execution and settlement, size, nature and any other relevant consideration.