The Role of the Duty of Best Execution in Off-Exchange Market Making: Persistent Issues and Their Emerging Iterations
The Duty of Best Execution, Off-Exchange Market Makers, and Their Foreign Counterparts
A recent enforcement action brought by the Financial Industry Regulatory Authority, Inc. (“FINRA”) against Virtu Americas LLC for its failure to match at midpoint certain orders routed by other broker-dealers once again has articulated the legal principle that the duty of best execution extends to off-exchange market makers.[1] The relevance of this principle, which illustrates a potential reach of the best execution standard to multiple parties in the execution process starting with customer-facing brokers, cuts through a variety of different issues in the market structure ecosystem.[2] In that respect, off-exchange market makers operating in the U.S. securities markets, commonly known as “wholesalers” or “internalizers,” are hardly unique. Their widely recognized counterparts, stand-alone liquidity providers known as “systematic internalisers” (“SIs”) under the European regulatory regime centered around MiFID II, are required, “while complying with Article 27 of Directive 2014/65//EU [which sets the best execution standard to govern the relationship between ‘investment firms’ more generally and their clients], [to] execute the orders they receive from their clients in relation to the shares, depositary receipts, ETFs [exchange-traded funds], certificates and other similar financial instruments for which they are systematic internalisers at the quoted prices at the time of reception of the order.”[3]
While both groups of off-exchange market makers are subject to the best execution standard, SIs also need to “make public their quotes on a regular and continuous basis during normal trading hours . . . in a manner which is easily accessible to other market participants on a reasonable commercial basis.”[4] Furthermore, these quotes are subject to a certain minimum quote size and need to “reflect the prevailing market conditions.”[5] One of the dimensions of the term “prevailing market conditions” is that such quotes are required to be “close in price, at the time of publication, to quotes of equivalent sizes for the same financial instrument on the most relevant market in terms of liquidity,”[6] which serves an additional mechanism of securing a proper reference price. However, SIs are also “allowed to decide, on the basis of their commercial policy and in an objective non-discriminatory way, the clients to whom they give access to their quotes [and] may refuse to enter into or discontinue business relationships with clients on the basis of commercial considerations.”[7] These quoting obligations should be contrasted to their complete lack in the U.S. securities markets, where off-exchange market makers have adopted the business model of “dark,” i.e., undisplayed, liquidity, which is typically tied to the National Best Bid and Offer (“NBBO”) with potential price improvement. On the other hand, this business model is not entirely ignored under the European regulatory regime. In addition to a broad set of exemptions available to SIs for certain large orders,[8] “in justified cases, [SIs] may execute those orders at a better price provided that the price falls within a public range close to market conditions.”[9] As another illustration, an SI may, for certain orders, “execute that part of the order which exceeds its quotation size, provided that it is executed at the quoted price.”[10] Interestingly, one empirical study analyzed SIs operating on the French equity market and concluded that “a large portion of SI transactions are not subject to pre-trade transparency requirements [based on various exemptions] [and] that transactions subject to pre-trade transparency requirements represent only 22% of the amounts traded by SIs during the continuous trading phase.”[11] The distinction between “lit” and “dark” liquidity also implicates different approaches to complying with the duty of best execution, and this factor is intricately connected to risk management.[12]
In light of numerous price-related issues, which are often directly connected to the best execution standard, the applicable tick size regime as a mandatory pricing grid that governs quoting, price improvement, and execution has quite a bit of significance. One solution under MiFID II has employed the following approach to incorporate the relevant granular / dynamic tick price regime, which was still constrained by the exemption for certain large orders: “[T]he prices published by a systematic internaliser in respect of shares and depositary receipts shall be deemed to reflect prevailing market conditions only where those prices . . . respect minimum price increments corresponding to the tick sizes specified in Article 2 of Commission Delegated Regulation (EU) No 2017/588.”[13] A subsequent regulatory change was more radical, as it went beyond quoting requirements and covered other transactions. It extended the tick size regime to SIs’ “quotes, price improvements on those quotes and execution prices,” with the reservation that “[a]pplication of tick sizes shall not prevent systematic internalisers matching orders large in scale at mid‐point within the current bid and offer prices,” and clarified that this regime applies to SIs “when dealing in all sizes.”[14] On the other end of the spectrum, off-exchange market makers in the U.S. securities markets are able to offer essentially unconstrained subpenny price improvement while providing undisplayed liquidity, as Regulation NMS specifically sanctions this practice, as opposed to the ban on displaying subpenny quotes or accepting subpenny orders.[15]
Another important consideration is that an SI is defined in MiFID II as “an investment firm which, on an organised, frequent systematic and substantial basis, deals on own account when executing client orders outside a regulated market, an MTF [multilateral trading facility] or an OTF [organised trading facility] without operating a multilateral system.”[16] When the criteria for the mandatory application of the SI status were adopted, it was also stated such an entity is not allowed to “bring together third party buying and selling interests in functionally the same way as a trading venue” and hence constitute “an internal matching system which executes client orders on a multilateral basis . . . which [would] result[] in the investment firm undertaking matched principal transactions on a regular and not occasional basis.”[17] A subsequent regulatory change specified that “[a]n investment firm shall not be considered to be dealing on own account for the purposes of Article 4(1)(20) of Directive 2014/65/EU where that investment firm participates in matching arrangements entered into with entities outside its own group with the objective or consequence of carrying out de facto riskless back-to-back transactions in a financial instrument outside a trading venue” and clarified that it covers “the internal or external matching.”[18] The European regulators provided additional guidance on these issues, which, among other things, stresses that “the trading activity of [an] SI is characterised by risk-facing transactions that impact the Profit and Loss account of the firm,” states that an SI is not prohibited “from hedging the positions arising from the execution of client orders as long as it does not lead to the SI de facto executing non risk-facing transactions and bringing together multiple third party buying and selling interests,” and advises against “systems or arrangements, be they automated or not, intended to match opposite client orders [or] aimed at increasing opportunities for client order matching.”[19] In addition to the permissibility of certain contemporaneous offsetting transactions elsewhere, it should also be recognized that firms and their affiliates could handle clients’ orders, such as nonmarketable limit orders, outside their respective SI platforms and route them to other destinations.[20] By contrast, off-exchange market makers in the U.S. securities markets are not required to trade only as principal or engage in segregation of liquidity provision: a portion of order flow routed by other broker-dealers on the basis of negotiated arrangements could be internalized / “cherry picked” or matched, with the rest rerouted elsewhere.[21]
Payment for Order Flow and Its Discontents
The controversial practice of payment for order flow (“PFOF”), which characterizes routing arrangements between customer-facing brokers and off-exchange market makers based on monetary incentives paid by the latter to the former in the U.S. securities markets, should be singled out as a key distinction between the European and U.S. regulatory regimes. While the U.S. Securities and Exchange Commission (“SEC”) has employed a disclosure-based approach to PFOF, as exemplified by its pivotal 1994 release,[22] MiFID II, in the context of its best execution standard, prohibits an “investment firm” from “receiv[ing] any remuneration, discount or non-monetary benefit for routing client orders to a particular trading venue or execution venue” relating to certain conflicts of interest and inducements.[23] This provision has been interpreted as a de facto ban on PFOF. For instance, as asserted by the U.K. regulators, “[F]irms should be considering the regulatory changes forthcoming under MiFID II, which will place further restrictions on charging PFOF. For professional client business, MiFID II further reinforces the ineligibility of these third-party payments when executing orders on behalf of clients [under the provision in question].”[24] This conclusion was accompanied by a number of criticisms of PFOF in general, as this practice was said to “create[] a conflict of interest between a firm (the broker) and its clients because the firm is incentivised to pursue payments from market makers rather than to provide best execution in the interests of their clients,” to “undermine[] the transparency and efficiency of the price formation process . . . because the prices paid by clients include hidden costs [in the form of] the higher spread that they may additionally need to pay to take account of the fees paid by the market [maker],” and to “[f]orc[e] market makers to ‘pay-to-play’ [and thus] distort competition by creating barriers to entry and expansion.”[25] However, with PFOF being just one example, the prohibition on collecting incentives from a routing destination in connection with client orders, such as those stemming from maker-taker arrangements, is not absolute under MiFID II.[26]
Turning back to the U.S. regulatory regime, PFOF in itself has not been considered a breach of the duty of best execution.[27] Moreover, several arguments have been offered in defense of this practice, such as competition with other trading venues’ incentive programs and paid-for benefits for customers.[28] Accordingly, the focus should be on execution quality provided by off-exchange market makers and the very structure of the underlying PFOF arrangements, which need to provide for meaningful monitoring of execution quality and compliance with the duty of best execution by both off-exchange market makers and customer-facing brokers. Yet another important issue embedded in such arrangements relates to a potential tradeoff and hence a proper split between the respective amounts of price improvement and PFOF.[29]
A prominent illustration of the economics of PFOF and the corresponding implications for the best execution standard is offered by the recent push for zero-commission trades by retail brokerage firms, which was spearheaded by Robinhood Financial. Not surprisingly, PFOF “has become especially vital to companies’ bottom line after commissions went to zero.”[30] While zero-commission trades are a clear example of a PFOF-funded benefit accruing directly to customers, there may be a potentially greater cost in the form of inferior execution quality. In fact, an executive of a leading retail brokerage firm gave a rather problematic description of different tiers of “best execution” offered by that firm to customers aside from any ancillary trading products and services: “If it’s IBKR Lite with zero commissions we do what the other brokers do, we send them off to a market maker just like everybody else and there’s payment for order flow that comes back and you may not get as good of an execution. If [it’s] IBKR Pro you’ll get better execution.”[31] For one thing, while the scope of the duty of best execution could be tweaked contractually, any distancing from the depth of this standard for specific transactions by a customer-facing broker, either with or without disclosure, would raise a number of legal and regulatory issues, such as liability for damages to customers or compliance with the applicable FINRA rules and especially Rule 5310.[32] Moreover, this description coming from a prominent market player casts doubts on execution quality offered by some off-exchange market makers and hence their own compliance with the duty of best execution as a potentially widespread and less transparent problem.
IEX’s D-Limit Order Type
Another case study of the interaction between the duty of best execution and off-exchange market making is the SEC’s recent approval of the D-Limit order type proposed by Investors Exchange LLC (“IEX”).[33] This order type was designed to protect liquidity providers’ displayed orders from latency arbitrage through repricing to a less aggressive price for a time period of up to two milliseconds based on IEX’s proprietary “crumbling quote indicator” (“CQI”), which had previously been applied only to certain nondisplayed orders. As an illustration, during September 2019, “the CQI was on for 1.64 seconds per symbol per day on average, which is 0.007% of the time during regular market hours [and] [o]n a volume-weighted basis . . . the CQI was on for 5.9 seconds per day per symbol, or 0.025% of the time during regular market hours.”[34] In the context of the same time period, IEX also observed that, “[w]ithout an order type that leverages the protective features of the CQI, 24% of displayed volume on IEX is executed when the CQI is on, compared to only 3% of nondisplayed volume,” and that 33.7% of marketable orders arrived when the CQI was on.[35]
The process of approval and public comment for the D-Limit order type has been contentious, culminating in a pending lawsuit by Citadel Securities LLC, a leading off-exchange market maker and one of the main critics of this order type, to set aside the SEC’s decision.[36] While Citadel Securities’ arguments specifically included concerns about the best execution standard in the context of routing retail orders to IEX,[37] the business practices of this off-exchange market maker itself — and, by implication, industry practices — have been scrutinized. Based on Citadel Securities’ statement that, “[w]hen routing retail orders to external venues for execution, consistent with standard market practice, we typically enter into back-to-back transactions (one on the external venue and one with the retail broker-dealer),”[38] IEX noted that it was unclear “whether the prices [Citadel Securities] receives are in all or most cases passed directly back to the ultimate retail customer.”[39] Moreover, IEX pointed out that Citadel Securities’ execution of odd-lot and certain large orders, which sometimes took as long as 50 milliseconds on average, was substantially longer than IEX’s automatic 350-microsecond delay and questioned that “accounting for that delay in planning the routing of orders would be improper given the much larger round-trip latency over which Citadel exercises discretion.”[40] In response, Citadel Securities continued to criticize the proposed order type combined with IEX’s delay mechanism. This firm asserted that “‘[a]ccounting’ for the IEX speedbump means routing to IEX first and intentionally delaying routing to other exchanges when accessing displayed liquidity,” questioned “whether intentionally delaying the routing of marketable orders is consistent with the ‘fully and promptly’ best execution standard [under FINRA Rule 5310],” and maintained that, “[j]ust as 350 microseconds is long enough for IEX to reprice D-Limit quotes before they can be accessed, it is also more than long enough for liquidity providers on other exchanges to reprice displayed quotes.”[41]
In its approval order, the SEC even referenced a controversial comment letter by an anonymous commentator and a self-described “industry expert” making the claims that (i) “retail orders are likely sent to Citadel at random times (initiated by actual retail investors living in different parts of the world), but Citadel likely chooses to route these orders to IEX during a CQI” and (ii) “[d]uring that time [corresponding to an average speed of execution for a retail order], Citadel has a free option to hold the order and decide what to do with it [in order] to make as much money for Citadel as possible.”[42] The first claim was made in the context of a prior comment letter by Citadel Securities, which was interpreted by the anonymous commentator as stating that “a majority of the order flow sent [by that firm] to IEX represents retail investors and that 50% of their liquidity taking orders on IEX occur when the [CQI] is ‘on’” and hence described as “a completely shocking admission by Citadel in their treatment of retail orders [given that such an arrival rate] in that narrow of a time window is mathematically impossible.”[43] In reality, Citadel Securities only stated in the comment letter in question that it is “one of the top 3 takers of liquidity on IEX when the CQI signal is ON [and that] over 50% of [its] trading activity on IEX is on behalf of retail investors,” without specifying any split between the CQI- and non-CQI-related volumes.[44] In a later submission, this firm stated that its “own analysis of the retail orders routed to IEX by Citadel Securities that removed displayed liquidity during the month of May 2020 [showed that] the CQI was ON for approximately 15% of these retail orders” and that, consistent with its treatment and characteristics of large retail orders, “of the retail orders . . . being executed on IEX when the CQI was ON, 98.3% of those orders (accounting for 99.9% of the total volume) were routed to multiple exchanges.”[45] Accordingly, instead of treating the relevant arrival rate as an unambiguous indication of abuse of retail orders, it should be considered that the CQI itself may be triggered by large retail orders under certain scenarios, as a reverse causal link. On the other hand, the second claim made by the anonymous commentator appears to be in connection with “[t]he average speed of execution for a retail order, [which] according to popular retail broker websites, is between 50 and 400 milliseconds,”[46] and hence not directly connected to the CQI and the relevant time period of up to two milliseconds. In addition, any processing time attributed to customer-facing brokers rather than off-exchange market makers themselves needs to be taken into account. At the same time, the allegation of “sitting on an order” was neither discussed in detail nor backed up by data.
Aside from any speculative inferences about specific market players or opinions on the desirability of the D-Limit order type itself, this debate has implications for the duty of best execution in the context of off-exchange market making. Given the described practice of back-to-back transactions with routing broker-dealers and external trading venues, as opposed to pure agency-based routing, it would be contrary to the best execution standard to deprive customers of “better” prices and reasonably available price improvement opportunities in these near-instantaneous transaction pairs, which are undoubtedly driven by a number of considerations, such as fee-rebate structures on various trading venues. In any instance, rerouting of customer orders should avoid opportunistic behavior that jeopardizes execution quality, for instance, by prioritizing rebate collection. Likewise, promptness is one of the dimensions of the duty of best execution,[47] and routing practices also need to take into account various market structure wrinkles, such as delay mechanisms and other novel features and functionalities.[48] Even more importantly, a self-interested “sitting on an order” to a customer’s detriment — whether based on predictive analytics or something else — would undoubtedly violate this duty.[49]
The Enduring Calculus of Best Execution
At the end of the day, the duty of best execution remains a complex multifaceted issue for off-exchange market makers as one of the market structure pillars in the U.S. securities markets. With the sheer degree of discretion exercised by these market players while handling customer orders, the multitude of moving parts includes execution quality, execution policies and procedures, routing practices, the structure of PFOF arrangements, and the nature of arbitrage strategies that involve routed order flow. At the same time, compliance with the best execution standard cannot address broader problems related to market fragmentation, including a proper balance between competition among marketplaces and competition among orders. Even more so, this standard itself is shaped by market structure features, as illustrated by different approaches to PFOF, permitted agency functions, tick size constraints, and quoting obligations employed by the European and U.S. regulatory regimes.
The duty of best execution also interacts with the value-added proposition offered by off-exchange market makers, such as the ability to leverage order flow segmentation for enhancing execution quality and order size guarantees. In turn, this proposition is fused with certain advantages enjoyed by these market players, such as the very existence of captive order flow. As observed by one commentator, “Their aggregation of order flow also means they know the direction in which the retail segment is pushing particular names. Thus they are gaining an advantage from seeing the flows coming through their pipes.”[50] Some advantages may have an even longer shelf life: “Wholesale market-makers also, over the long term, acquire voluminous amounts of market data, including information about retail order flows, which can be used in future modeling for order-handling, trading, and risk decisions.”[51]
The aggregate amount of price improvement, as well as its allocation, provided to customers by off-exchange market makers is an important piece of the best execution puzzle, which represents the opportunity cost of obtaining price improvement elsewhere. In general, the aggregate amount of price improvement appears to be rising quite significantly in 2020, although accompanied by considerable fluctuations in price improvement per share, and this trend reflects greater order flows to several off-exchange market makers.[52] As another informative statistic, the results of one study from the first six months of 2020 suggest that the aggregate amount of price improvement by a group of off-exchange market makers is 3.5 times larger than the aggregate amount of PFOF for the sample in question.[53] Perhaps this study signals that the economic pie is being sliced increasingly more in favor of customers relative to customer-facing brokers. At the same time, metrics related to price improvement could be gamed or distorted, and the benchmark price selection itself, including the NBBO, could invite opportunistic behavior in light of some degree of flexibility in terms of timing, information sources, reference transactions, or other factors.[54] For instance, one analysis has demonstrated how reported price improvement in the case of odd-lots could in fact disguise price disimprovement.[55] Moreover, in a key enforcement action, the SEC penalized an off-exchange market maker for self-interested arbitrage between the official and private data feeds in the process of handling retail order flow, pointing out that orders “often received price improvement, but this amount often was not sufficient to equal the price difference that had triggered the [underlying] strategy” thus disadvantaging “a substantial number of smaller orders.”[56] Some comparative evidence on price improvement from a more granular / dynamic pricing grid with no PFOF is offered by the recent empirical study of SIs operating on the French equity market, which concluded that, given the looming broad extension of the tick size regime under MiFID II to SIs, nearly 40% of trading volume in the sample, representing the bulk of price improvement, would not have complied with this regime.[57] Overall, just like in the case of several other dimensions of best execution, addressing price improvement requires objective and data-intensive compliance, reporting, and external monitoring.
[1] Virtu Ams. LLC, Letter of Acceptance, Waiver and Consent №2016049752801 (Fin. Indus. Reg. Auth., Inc. July 21, 2020), https://www.finra.org/sites/default/files/fda_documents/2016049752801%20Virtu%20Americas%20LLC%20%28f%20k%20a%20KCG%20Americas%20LLC%29%20CRD%20149823%20AWC%20jlg.pdf [https://perma.cc/5CZY-C3ER].
[2] For a discussion of such issues by the author, see Stanislav Dolgopolov, Wholesaling Best Execution: How Entangled Are Off-Exchange Market Makers?, 11 Va. L. & Bus. Rev. 149 (2016), https://ssrn.com/abstract=2744904 [hereinafter Dolgopolov, Wholesaling Best Execution]; Stanislav Dolgopolov, The Market Data Infrastructure, the Duty of Best Execution, and Off-Exchange Market Makers: Connecting Regulatory Reforms and Enforcement Implications, Medium (Aug. 10, 2020), https://medium.com/@s_v_dolgopolov/the-market-data-infrastructure-the-duty-of-best-execution-and-off-exchange-market-makers-fd485b5d5f45 [https://perma.cc/6GW7-QBGN].
[3] Regulation (EU) No 600/2014 of the European Parliament and of the Council of 15 May 2014 on Markets in Financial Instruments and Amending Regulation (EU) No 648/2012, art. 15(2), 2014 O.J. (L 173) 84, 112, http://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32014R0600&from=EN [https://perma.cc/D4HU-Z2TM] [hereinafter Regulation (EU) No 600/2014] (citing Directive 2014/65/EU, of the European Parliament and of the Council of 15 May 2014 on Markets in Financial Instruments and Amending Directive 2002/92/EC and Directive 2011/61/EU, art. 27, 2014 O.J. (L 173) 349, 412–13, http://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32014L0065&from=EN [https://perma.cc/JWT4-8DJD] [hereinafter Directive 2014/65/EU]).
[4] Id. art. 15(1), at 111–12.
[5] Id. art. 14(3), at 111.
[6] Commission Delegated Regulation (EU) 2019/442 of 12 December 2018 Amending and Correcting Delegated Regulation (EU) 2017/587 to Specify the Requirement for Prices to Reflect Prevailing Market Conditions and to Update and Correct Certain Provisions, art. 1(3), 2019 O.J. (L 77) 56, 57, https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32019R0442&from=EN [https://perma.cc/3REZ-X7D4] [hereinafter Commission Delegated Regulation (EU) 2019/442].
[7] Regulation (EU) No 600/2014, supra note 3, art. 17(1), at 112.
[8] Id. art. 14(2), at 111.
[9] Id. art. 15(2), at 112.
[10] Id. art. 15(4), at 112.
[11] Iris Lucas, Autorité des Marchés Financiers, Quantifying Systematic Internalisers’ Activity: Their Share in the Equity Market Structure and Role in the Price Discovery Process 3–4 (May 2020) (unpublished manuscript), https://www.amf-france.org/sites/default/files/2020-06/202005_etude_internalisateurs_integrale_va.pdf [https://perma.cc/2XQV-8L4X].
[12] For several provisions related to risk management in connection with displayed liquidity provided by SIs, see Regulation (EU) No 600/2014, supra note 3, arts. 15(1), 17(2), at 111–12; Commission Delegated Regulation (EU) 2017/567 of 18 May 2016 Supplementing Regulation (EU) No 600/2014 of the European Parliament and of the Council with Regard to Definitions, Transparency, Portfolio Compression and Supervisory Measures on Product Intervention and Positions, arts. 14–15, 2017 O.J. (L 87) 90, 99–100, https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32017R0567&from=EN [https://perma.cc/W54Q-APLC].
[13] Commission Delegated Regulation (EU) 2019/442, supra note 6, art. 1(3), at 57 (citing Commission Delegated Regulation (EU) 2017/588 of 14 July 2016 Supplementing Directive 2014/65/EU of the European Parliament and of the Council with Regard to Regulatory Technical Standards on the Tick Size Regime for Shares, Depositary Receipts and Exchange-Traded Funds, art. 2, 2017 O.J. (L 87) 411, 412–13, https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32017R0588&from=EN [https://perma.cc/SDK2-YSPL]).
[14] Regulation (EU) 2019/2033 of the European Parliament and of the Council of 27 November 2019 on the Prudential Requirements of Investment Firms and Amending Regulations (EU) No 1093/2010, (EU) No 575/2013, (EU) No 600/2014 and (EU) No 806/2014, recital 44, art. 63, 2019 O.J. (L 314) 1, 10, 57, https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32019R2033&from=EN [https://perma.cc/FBU4-R7SM].
[15] Dolgopolov, Wholesaling Best Execution, supra note 2, at 158–59.
[16] Directive 2014/65/EU, supra note 3, art. 4(1)(20), at 382.
[17] Commission Delegated Regulation (EU) 2017/565 of 25 April 2016 Supplementing Directive 2014/65/EU of the European Parliament and of the Council as Regards Organisational Requirements and Operating Conditions for Investment Firms and Defined Terms for the Purposes of That Directive, recital 19, arts. 12–17, 2017 O.J. (L 87) 1, 3–4, 24–27 https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32017R0565&from=EN [https://perma.cc/N5VK-REM6] [hereinafter Commission Delegated Regulation (EU) 2017/565].
[18] Commission Delegated Regulation (EU) 2017/2294 of 28 August 2017 Amending Delegated Regulation (EU) 2017/565 as Regards the Specification of the Definition of Systematic Internalisers for the Purposes of Directive 2014/65/EU, recital 2, art. 1, 2017 O.J. (L 329) 4, 4–5, https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32017R2294&from=EN [https://perma.cc/84ZZ-LT75.
[19] Eur. Sec. & Mkts. Auth., No. ESMA70–872942901–38, Questions and Answers: On MiFID II and MiFIR Market Structures Topics 54–57 (May 29, 2020), https://www.esma.europa.eu/sites/default/files/library/esma70-872942901-38_qas_markets_structures_issues.pdf [https://perma.cc/AM6N-MY5W] [hereinafter ESMA, Q&A on MiFID II and MiFIR Market Structures Topics].
[20] For a detailed illustration, see Virtu Fin. Ir. Ltd., No. COM-20399, Order Execution Policy for Professional Clients (Jan. 2020), https://www.virtu.com/uploads/documents/VFIL_Order_Execution_Policy_Jan_2020.pdf [https://perma.cc/2EUN-8GT8].
[21] See, e.g., Equities and Options Order Handling Agreement Dated as of November 29, 2007 by and Among E*Trade Financial Corporation, E*Trade Securities LLC, and Citadel Derivatives Group LLC § 2.1(a), reproduced in E*Trade Fin. Corp., Annual Report (Form 10-K) Exh. 10.29, at 8 (Feb. 28, 2008), https://www.sec.gov/Archives/edgar/data/1015780/000119312508041906/dex1029.htm [https://perma.cc/2ANR-S9X7] (stating that the off-exchange market maker in question “shall, in its sole discretion, determine to execute such Covered Orders [contemplated by the agreement] either as principal or by acting as riskless principal or agent, including, where applicable, determining the market center for execution of such Covered Orders”).
[22] Payment for Order Flow, Exchange Act Release №34,902, 59 Fed. Reg. 55,006 (Oct. 27, 1994) (codified at 17 C.F.R. pt. 240), https://www.gpo.gov/fdsys/pkg/FR-1994-11-02/html/94-27109.htm [https://perma.cc/52UL-UUD4].
[23] Directive 2014/65/EU, supra note 3, art. 27(2), at 412.
[24] U.K. Fin. Conduct Auth., Mkt. Watch №51, Observations from Market Maker Review 8–9 (Sept. 2016), https://www.fca.org.uk/publication/newsletters/marketwatch-51.pdf [https://perma.cc/346W-TRPT].
[25] Id. at 5.
[26] See, e.g., Commission Delegated Regulation (EU) 2017/565, supra note 17, art. 66(6), at 60 (“Investment firms . . . shall inform clients about the inducements that the firm may receive from the execution venues.”); Bank of Am. Merrill Lynch, BofAML 2017 Top Five Execution Venues for Client Orders: Regulation (EU) 2017/576 (“RTS 28”), v. 1, at 5 (Apr. 2018), https://markets.ml.com/web/public/bestex/eq/BofAML%20RTS%2028%20Qualitative%20Statement%20FY2017%20(April%202018)%20v1.0.pdf [https://perma.cc/23ZY-KUAX] (“Some European Exchanges and MTFs [multilateral trading facilities] operate a transparent maker/taker fee structure, whereby posting liquidity earns a rebate. BofAML receives such rebates in accordance with the rulebook and fee structures of these venues. BofAML has no private arrangements to receive compensation from any execution venue.”).
[27] See, e.g., Payment for Order Flow, 59 Fed. Reg. at 55,011 (“The Commission does not believe that all payment for order flow arrangements are against the customer’s best interest and must be banned per se as compromising a broker’s duty to seek best execution of the customer’s order.”).
[28] For a discussion of some of these arguments, see Dolgopolov, Wholesaling Best Execution, supra note 2, at 162–64.
[29] For additional sources addressing this tradeoff, see id. at 156 n.20, 160.
[30] Kate Rooney & Maggie Fitzgerald, Here’s How Robinhood Is Raking in Record Cash on Customer Trades — Despite Making It Free, CNBC (Aug. 13, 2020), https://www.cnbc.com/2020/08/13/how-robinhood-makes-money-on-customer-trades-despite-making-it-free.html.
[31] Id. (quoting Steve Sanders, Executive Vice President of Marketing and Product Development of Interactive Brokers); see also Interactive Brokers Group’s (IBKR) Q3 2019 Results — Earnings Call Transcript, Interactive Brokers Grp. 2 (Oct. 15, 2019, 4:30 PM), https://investors.interactivebrokers.com/download/ir/3Q19_IBKR_earnings_call_transcript.pdf [https://perma.cc/N2JY-V74W] (“Since the inception of our Electronic Brokerage business, we have prided ourselves in our commitment to offering the best platform for our customers. To us, this meant best execution by routing orders directly to a venue and not selling customer order flow. It meant routing our customers’ orders to whichever venue is likely to offer the best price for that specific security at that particular time. . . . More recently, we found that not everyone shared our definition of what is ‘best’. To some investors, ‘best’ meant paying 0 commissions first and foremost. If this is what some of our customers wanted, we decided to oblige.” (statement of Nancy Stuebe, Director, Investor Relations of Interactive Brokers)); Press Release, Interactive Brokers Grp., Inc., Interactive Brokers to Launch IBKR Lite (Sept. 29, 2019, 1:19 PM), https://www.businesswire.com/news/home/20190926005753/en/Interactive-Brokers-Launch-IBKR-Lite [https://perma.cc/99DK-NYC3] (“IBKR Lite was designed to meet the needs of investors who are seeking a simple, cost-free way to trade US exchange-listed stocks and ETFs and do not wish to consider our efforts to obtain greater price improvement through our IB SmartRouting(sm) system. . . . Like many other brokers, Interactive Brokers will route the orders of IBKR Lite clients to market makers in exchange for receiving payment for order flow. Clients that opt to use IBKR Pro will continue to receive the best prices our sophisticated algorithms can secure for them.”).
[32] 5310. Best Execution and Interpositioning, FINRA, https://www.finra.org/rules-guidance/rulebooks/finra-rules/5310 [https://perma.cc/V7YN-2Z3K] (the last amendment effective as of May 9, 2014).
[33] Order Approving a Proposed Rule Change by Investors Exchange LLC to Add a New Discretionary Limit Order Type Called D-Limit, Exchange Act Release №89,686, 85 Fed. Reg. 54,438 (Aug. 26, 2020), https://www.govinfo.gov/content/pkg/FR-2020-09-01/pdf/2020-19204.pdf [https://perma.cc/MH7S-8H4C] [hereinafter SEC’s D-Limit Approval Order].
[34] Id. at 54,440 & n.35.
[35] Notice of Filing of a Proposed Rule Change by Investors Exchange LLC to Add a New Discretionary Limit Order Type, Exchange Act Release №87,814, 84 Fed. Reg. 71,997, 71,999, 72,001 (Dec. 20, 2019), https://www.govinfo.gov/content/pkg/FR-2019-12-30/pdf/2019-28024.pdf [https://perma.cc/M3VJ-PRBL].
[36] Petition for Review, Citadel Sec. LLC v. SEC, №20–1424 (D.C. Cir. Oct. 16, 2020).
[37] Stephen John Berger, Managing Dir. and Global Head of Gov’t & Regulatory Policy, Citadel Sec. LLC, Comment Letter to the SEC on IEX’s Proposed D-Limit Order Type 4–5 (July 2, 2020), https://www.sec.gov/comments/sr-iex-2019-15/sriex201915-7378994-218874.pdf [https://perma.cc/XZF4-SQP7] [hereinafter Citadel Securities’ Second Comment Letter to the SEC on IEX’s Proposed D-Limit Order Type].
[38] Id. at 2.
[39] John Ramsay, Chief Mkt. Policy Officer, Inv’r Exch. LLC, Comment Letter to the SEC on IEX’s Proposed D-Limit Order Type 11 n.46 (Aug. 3, 2020), https://www.sec.gov/comments/sr-iex-2019-15/sriex201915-7534417-222147.pdf [https://perma.cc/K23F-EGBG] [hereinafter IEX’s Third Comment Letter to the SEC on IEX’s Proposed D-Limit Order Type].
[40] Id. at 12 (citing Citadel Sec. LLC, Q1–2019 FIF Supplemental Retail Execution Quality Statistics (n.d.), https://s3.amazonaws.com/citadel-wordpress-prd102/wp-content/uploads/sites/2/2016/09/09175131/FIF-Rule-605-606-WG-CitadelSecurities_Retail-Execution-Quality-Stats_Q1_2019.pdf [https://perma.cc/JNS4-NHQ3]).
[41] Stephen John Berger, Managing Dir. and Global Head of Gov’t & Regulatory Policy, Citadel Sec. LLC, Comment Letter to the SEC on IEX’s Proposed D-Limit Order Type 4–5 (Aug. 14, 2020), https://www.sec.gov/comments/sr-iex-2019-15/sriex201915-7653454-222402.pdf [https://perma.cc/CQ82-CTH5] [hereinafter Citadel Securities’ Third Comment Letter to the SEC on IEX’s Proposed D-Limit Order Type].
[42] SEC’s D-Limit Approval Order, supra note 33, at 54,440 n.38 (quoting Anonymous Indus. Expert, Comment Letter to the SEC on IEX’s Proposed D-Limit Order Type 2 (n.d.), https://www.sec.gov/comments/sr-iex-2019-15/sriex201915-7182916-216802.pdf [https://perma.cc/BR8Z-JAML] [hereinafter Anonymous Comment Letter to the SEC on IEX’s Proposed D-Limit Order Type]).
[43] Anonymous Comment Letter to the SEC on IEX’s Proposed D-Limit Order Type, supra note 42, at 2. The anonymous commentator also argued that Citadel Securities itself is likely to engage in CQI-like predictive analytics for deciding whether to internalize or reroute order flow accordingly and that such rerouting may seek to collect rebates on other exchanges and ultimately disadvantage certain retail orders. Id.
[44] Stephen John Berger, Managing Dir. and Global Head of Gov’t & Regulatory Policy, Citadel Sec. LLC, Comment Letter to the SEC on IEX’s Proposed D-Limit Order Type 2 (Apr. 23, 2020), https://www.sec.gov/comments/sr-iex-2019-15/sriex201915-7110555-215933.pdf [https://perma.cc/9NHE-AH9L].
[45] Citadel Securities’ Second Comment Letter to the SEC on IEX’s Proposed D-Limit Order Type, supra note 37, at 3–4. In response, IEX criticized the relevance of the data sample in question and stressed “the infinitesimally small chance that an order sent by a retail investor to her broker, which routes it to a wholesale broker, which then processes the order and elects to send it to one or more exchanges, would happen to arrive on IEX during a 2-millisecond window of time when the CQI is on.” IEX’s Third Comment Letter to the SEC on IEX’s Proposed D-Limit Order Type, supra note 39, at 9, 11. Citadel Securities retorted by emphasizing the scenario “when the execution of the retail investor order is actually what causes the CQI to turn ON in the first place.” Citadel Securities’ Third Comment Letter to the SEC on IEX’s Proposed D-Limit Order Type, supra note 41, at 3.
[46] Anonymous Comment Letter to the SEC on IEX’s Proposed D-Limit Order Type, supra note 42, at 2.
[47] See, e.g., Magnum Corp. v. Lehman Bros. Kuhn Loeb, Inc., 794 F.2d 198, 200 (5th Cir. 1986) (“The implicit agreement between customer and stockbroker is that the latter will use reasonable efforts to execute the order promptly at the best obtainable price.”).
[48] See also SEC’s D-Limit Approval Order, supra note 33, at 54,442 (“[W]ith or without D-Limit orders, if a broker-dealer does not seek to maximize its fill rates while minimizing information leakage by accounting for latencies (e.g., technological, geographic, or access delay) when it routes portions of large orders to multiple venues near-simultaneously, the broker-dealer runs the risk of missing out on executions at displayed prices.”).
[49] For a discussion of the hypothetical scenario of an off-exchange market maker sitting on an order, see Dolgopolov, Wholesaling Best Execution, supra note 2, at 192, 195–96.
[50] Chris Hall, The Incredible Shrinking Market, Trader Mag. (Aug. 20, 2020), https://www.tradersmagazine.com/xtra/the-incredible-shrinking-market/ [https://perma.cc/6MZ3-XKX6] (quoting Larry Tabb, Head of Market Structure Research, Bloomberg Intelligence).
[51] U.S. Sec. & Exch. Comm’n, Staff Report on Algorithmic Trading in U.S. Capital Markets 32 (Aug. 5, 2020), https://www.sec.gov/files/Algo_Trading_Report_2020.pdf [https://perma.cc/VXG4-G6P5].
[52] Paul Rowady, Wholesale Market Makers: Adding Price Improvement to the PFOF Analysis, Alphacution (Oct. 9, 2020), https://alphacution.com/wholesale-market-makers-adding-price-improvement-to-the-pfof-analysis/ (subscription required).
[53] Larry Tabb (@ltabb), Twitter (Aug. 13, 2020, 7:31 AM), https://twitter.com/ltabb/status/1293887815835684864.
[54] See also Roundtable on Market Data Products, Market Access Services, and Their Associated Fees, U.S. Sec. & Exch. Comm’n 152 (Oct. 25, 2018), https://www.sec.gov/spotlight/equity-market-structure-roundtables/roundtable-market-data-market-access-102518-transcript.pdf [https://perma.cc/P2GD-GH7H] (raising the issue of “benchmark reference price arbitrage” in connection with the concept of competing consolidators of market data and voicing concerns about conflicts of interest posed by the same entity “executing my trades and managing the price that they’re benchmarked against” in the context of off-exchange market making (remarks of Oliver Albers, Global Head of Sales and Strategic Partnerships, Nasdaq)).
[55] Tyler Gellasch, Exec. Dir., Healthy Mkts. Ass’n, Comment Letter to the SEC on the Market Data and Market Access Roundtable 2–13 (Mar. 5, 2019), https://www.sec.gov/comments/4-729/4729-5020185-182987.pdf [https://perma.cc/8KDY-YFN9].
[56] Citadel Sec. LLC, Securities Act Release №10,280, Exchange Act Release №79,790, at 8–9 (Jan. 13, 2017) (settled proceeding), https://www.sec.gov/litigation/admin/2017/33-10280.pdf [https://perma.cc/E5WE-MQ9Y].
[57] Lucas, supra note 11, at 4. In fact, a key industry group criticized the anticipated implementation of this measure on the following grounds: “The application of the tick size regime above Large-in-Scale (other than trades executed at mid-point) will in our view not contribute to the price discovery process for [Large-in-Scale] trades and may actually inhibit appropriate price formation between systematic internalisers and their clients. Furthermore, the ability to execute Large-in-Scale trades on a sub-tick basis provides meaningful price improvement for clients trading in large sizes which brings benefits to end investors.” Eur. F. of Sec. Ass’ns, EFSA — Initial Comments on MiFID 2/R Review 3 (Jan. 20, 2020), https://www.afme.eu/Portals/0/DispatchFeaturedImages/EFSA%20comments%20on%20MiFID%20Review%20dated%2020%20January%202020.pdf [https://perma.cc/KFS9-EAFA]. On the other hand, several stakeholders had previously argued that the option of de minimis price improvement by SIs in combination with the best execution standard applicable to SIs themselves or other parties conferred an unjustified advantage and could ultimately harm the entire marketplace. Stephen McGoldrick & Gargi Purandare, Deutsche Bank Glob. Equities, MiFID II: Definition of Systematic Internalisers and Industry Response to EC Consultation app. 1, at 6–7, 9 (Sept. 28, 2017), https://autobahn.db.com/microSite/docs/MiFID2_Examined_Briefing_Note_1_Definition_of_SI.pdf [https://perma.cc/2J2B-U2FU]. The European regulators also asserted that “marginal price improvements on quoted prices would challenge the efficient valuation of equity instruments without bringing any real benefits to investors [and that] to ensure that price improvements do not undermine the efficient pricing of instruments traded, price improvements on quoted prices would only be justified when they are meaningful and reflect the minimum tick size [regime].” ESMA, Q&A on MiFID II and MiFIR Market Structures Topics, supra note 19, at 58. At the end of the day, finding an optimal combination of a level playing field for competing market players, material price improvement for specific types of orders, and effective price discovery in the marketplace remains a difficult detail-driven exercise for any regulatory regime.
