Trickery in [High Frequency] Trading

Coming up today on the Manipulation Monitor’s compendium of curious competition cases is an overview of claims of chicanery in the high frequency trading markets. Specifically, this post will provide an overview of five litigations, listed below, all of which concern the operation of stock exchanges and the Barclay’s Liquidity Cross (or “LX”) dark pool, and which have been MDL’d under:

  • In Re: Barclays Liquidity Cross and High Frequency Trading Litigation, 1:14-md-02589-JMF (SDNY).
  • City of Providence, Rhode Island v. Bats Global Markets, Inc. et al., 1:14-cv-02811-JMF 2014 (S.D.N.Y.)
  • American European Insurance Company v. Bats Global Markets, Inc. et al., 1:14-cv-03133-JMF 2014 (S.D.N.Y.)
  • Harel Insurance Co, Ltd. v. Bats Global Markets, Inc. et al., 1:14-cv-03608-JMF 2014 (S.D.N.Y.)
  • Flynn et al v. Bank of America Corporation et al., 1:14-cv-04321-JMF 2014 (S.D.N.Y.)
  • Great Pacific Securities v. Barclays PLC et al., 1:15-cv-00168-JMF 2015 (C.D. Cal.)

Procedural Posture

The four cases initially brought in the Southern District of New York had previously been consolidated, for pre-trial purposes, as City of Providence v. BATS Global Markets, Inc., No. 14-cv-2811. In these four cases, various investors (collectively, the “SDNY Plaintiffs”) brought claims under the Securities Exchange Act of 1934, 15 U.S.C. § 78a et seq., against seven stock exchanges—BATS Global Markets, Inc., Chicago Stock Exchange, Inc., Direct Edge ECN, LLC, the NASDAQ Stock Market LLC, NASDAQ OMX BX, Inc., New York Stock Exchange, LLC, and NYSE Arca, Inc. (collectively, “the Exchanges”)—as well Barclays PLC and Barclays Capital Inc. (together, “Barclays”). The fifth action, Great Pacific Securities v. Barclays PLC et al., 1:15-cv-00168-JMF, was initially filed in the US District Court for the Central District of California, but consolidated by order of the Judicial Panel on Multidistrict Litigation in December of 2014. In this fifth action, Plaintiff Great Pacific Securities (“Great Pacific”) sued Barclays alleging violations of California state law.

On January 23, 2015, Barclays’ and the Exchange defendants each filed motions to dismiss the second consolidated amended complaint (“SCAC”). Barclays further filed an addition motion to dismiss the claims brought by Great Pacific. All three motions were argued on June 18, 2015.

On August 26, 2015, Judge Jesse M. Furman granted all three motions to dismiss, though granting leave for Great Pacific to amend its complaint. With respect to the claims against the Exchanges, the court found the behavior complained of fell “within the scope of the quasi-governmental powers delegated to the Exchanges” by the SEC, and that the Exchanges were thus immune from suit. The court further found that, even if the Exchanges were not immune, the SDNY Plaintiff’s manipulative-scheme claims failed to meet the heightened pleading standards required for fraud under FRCP Rule 9(b), and, moreover, that no private right of action exists under the relevant statute, Section 6(b) of the Exchange Act. The claims against Barclays by both the SDNY Plaintiffs and Great Pacific, while arising under different statutory schemes—SDNY Plaintiff’s claims under federal securities law, and Grand Pacific’s under California state law—both were dismissed for failure to identify any manipulative acts on which Plaintiffs reasonably relied. The court was further of the opinion that the SDNY Plaintiff’s allegations against both defendants amounted to, at most, the contention that Barclays and the Exchanges aided and abetted the HFT firms by creating conditions through which those firms could influence the price at which securities were traded—but neither Section 10(b) nor Rule 10b-5 create liability for aiding and abetting the violation of another party, thus further justifying dismissal.

The SDNY Plaintiffs appealed this order. In a decision issued on December 19, 2017, the Second Circuit concluded first, that the defendant Exchanges were not entitled to absolute immunity with respect to the securities fraud claims; second, that Plaintiffs had sufficiently alleged that the exchanges were engaged in manipulative conduct; and, third, that Plaintiffs had alleged participation by the Exchanges in a fraudulent scheme, rather than merely aiding and abetting. Accordingly, the decision of the district court in favor of the defendants was vacated, and the case remanded for proceedings consistent with the appellate decision.

The Barclays defendants were not parties to the appeal, and the SDNY Plaintiffs did not appeal the lower court’s decision with respect to their claims under Section 6(b) of the Exchange Act, dismissed on the grounds that the section provides for no private cause of action. A court conferenced followed the appellate decision, and, as a result, member cases Flynn v. Bank of Am. Corp., 14-CV-4321, and Great Pac. Sec. v. Barclays PLC, 15-CV-168 were closed, while Plaintiff Foresta AP-fonden was terminated from both the MDL’d case and member case 14-cv-2811. For those keeping track, this leaves just three of the five actions still in play, each of which allege violations of Section 10(b) of the Exchange Act, and Rule 10b-5, against the Exchange defendants.

The Remaining Claims

Now that we’ve established the somewhat convoluted procedural posture, let’s delve into the allegations being made in the remaining actions.

Who are the “SDNY Plaintiffs,” anyway? Each of the plaintiffs in the remaining cases are institutional investors who participated in the buying and selling of shares of U.S.-based stock exchange listed stock for investment purposes, including on the Exchanges named in this action. These include City of Providence, Rhode Island; Plumbers and Pipefitters National Pension Fund; Employee’s Retirement System of the Government of the Virgin Islands; and State-Boston Retirement System. The case was brought as a class action on behalf of all public investors who purchased and/or sold shares of stock in the United States between April 18, 2009, and the present, and generally alleges a “scheme and wrongful course of business” whereby the Exchange Defendants (and Barclays, though this write up will, given the result of the appeal, focus on the Exchanges) employed various devices and manipulations to defraud investors in a manner that, by design, manipulated the U.S. securities markets and trading on those markets, diverting “billions of dollars annually” from other market participants, and generating every more in kickback payments for the Exchanges.

Some of the basics of the stock trade are important here, so let’s take a step back to look at the rise of high-frequency trading before delving into the specific behavior that Plaintiffs claim to underlie the manipulation. Until (what some may consider to be) quite recently, there was no centralized, national market for the exchange of stocks. In 1975, however, Congress amended the Exchange Act to give the SEC the authority to issue rules stitching together the various disparate exchanges together. The SEC duly promulgated a host of regulations to fulfill this unified vision, all of which were consolidated into a single rule—“Regulation NMS” (“NMS” standing for “National Market System”)—in 2005. Regulation NMS also requires the publication of market plans (“NMS Plans”) which, among other things, compel the exchange to transmit real-time information regarding transactions on that exchange to a centralized entity. This centralized entity, known as the “Processor,” consolidates the information into a single, consolidated data feed.

Immediate acceptance of the most competitive offer by a broker or exchange matching a buyer to a seller is an additional requirement of Regulation NMS, and one that is facilitated by the existence of the centralized feed. This combination of a consolidated price and “best available price,” requirement, together with the authorization, in 1998, for electronic platforms to register as exchanges, has ushered in the rise of high-frequency trading (“HFT”). HFT is a catch-all term that refers to the practice of using computer-driven algorithms to buy and sell stock positions very quickly, taking advantage of small differences in stock prices, often across different exchanges. In the last fifteen years or so, the volume of HFT orders placed has risen from around ten percent of the Exchanges’ trading volume to nearly three-quarters.

This success of the arbitrage described above relies heavily on the ability of high-frequency traders to react expeditiously to information about the stock market. Plaintiffs identified three different practices undertaken by the Exchanges that improve the speed at which high-frequency traders can execute their trades. The Exchanges were motivated to implement these different practices, so the complaint alleges, because they directly benefit from the promotion of high frequency trading practices. Namely, the Exchanges make a commission on each trade that passes through their system: more trades, more profit, and, given the proportion of daily trades conducted by the high-frequency traders, more reason to cater to that group. Plaintiffs believe that these incentives resulted in the exchanges rigging their markets in favor of the HFT firms. To quote from the SCAC (itself quoting market experts),

The primary purpose of the stock exchanges has devolved to catering to a class of highly profitable market participants called [HFTs], who are interested only in hyper-short term trading, investors be damned. The stock exchanges give these HFTs perks and advantages to help them be as profitable as possible, even if doing so adversely affects you, the investors, because HFT firms are the exchanges’ biggest customers.

To satisfy the demands of HFT firms and to attract greater order flow—and yes, more commission fees—the Exchanges (1) designed and implemented complex “order types,” or commands that traders could use to tell the Exchange how to handle their bids and offers; (2) provided HFT firms with enhanced proprietary data feeds; and (3) permitted “co-location,” or the installation of the HFT firm’s servers in close proximity to the servers used by the Exchanges.

Complex Order Types

The implementation of new order types, which the SCAC describes as “exceedingly complex,” has the basic goal of providing customers of the Exchanges different ways to interact with the market. First, Plaintiffs allege that the Exchanges’ disclosures of complex order type functionality and handling practices to the SEC and to the public at large are wholly insufficient “for even the most sophisticated investor” to understand or utilize the new order types. As a result, these complex orders types are only utilized by the HFT firms, to the detriment of Plaintiffs.

Specifically, through these complex order types, HFT firms are able to engage in what the complaint describes as “superior queue positioning.” Standard practice would typically rank bids by price and time of receipt; here, the complex order types allow the bids made by the HFT firms to be placed at the top of the “order book,” rather than positioning them appropriately in the sequence. This lets the HFT firms execute “predatory strategies,” allowing them to profit on their trades to the detriment of unknowing and unsuspecting investors, and further gives the firms an increased opportunity to collect “makers” rebates—and avoid paying the “taker” fee—from the Exchanges. As a result, investor orders that would have been eligible for the makers rebates were converted into unfavorable executions incurring taker fees. The complaint further argues that this practice essentially inserts HFT firms as intermediaries between legitimate customer-to-customer trades, and further results in discriminatory handling of investor orders during sudden price movements.

Because of the complexity, limited marketing, and general lack of clarity surrounding these new order types, Plaintiffs believe that they were created by the Exchanges at the behest of their best customers, the HFT firms, through “exclusive, backroom communications.” The SCAC quotes several different traders and others market participants, including Eric Hunsader, founder of market data firm Nanex; Hunsader is quoted as having observed that “[e]xchanges are losing out to dark pools, so when HFTs ask for a new order type, they get a new order type.” The result was hundreds of new order-type options, many of which are detailed in the complaint.

It is important to note that the creation of these new order types was approved by the SEC. Accordingly, a portion of the complaint was dedicated to detailing how the Exchanges thwarted the rule-making process by failing to include important information about the functioning of the order types in their regulatory filings. The failure to provide the SEC with full information in its filings not only “deprived the SEC of information essential to performing its statutory regulatory function,” but, according to the complaint, derived the investing public of both adequate notice of the new order types, and of the opportunity to comment on the same.

Enhanced Data Feeds

As described above, all of the Exchanges are obligated to transmit real-time transaction information regarding to a centralized entity, the public securities information processor (“SIP” or “Processor”). The Exchanges are not permitted to release market data to private recipients before disseminating the data to SIP, but they do provide the opportunity to subscribe to enhanced or direct data feeds. The information in these feeds is inclusive of the data sent to SIP—or, in the case of enhanced feeds, more detailed—and is sent directly to the private subscribers at the same time as it is transmitted to the SIP. However, because the transmittal of information to SIP is consolidated, the HFT firms subscribing to the enhanced feeds receive market data more quickly than those who are dependent on SIP. Because the cost of receiving these alternative data feeds is beyond the reach of the majority of traditional buy-and-hold investors, the practice means that HFT firms are able to pay for access to public information sooner than the investing public—and thus trade on information before it is publicly disseminated. The allegations in the complaint explain that the more timely receipt, and particularly of enhanced data, allows the HFT firms to track when an investor changes price on their order or how much stock an investor is buying or selling, allowing the firms to predict short-term price movements “with near certainty.”

Regulation NMS requires that all trades be executed on the exchange offering the best price at the time of the order, and this—the national best bid and offer, or “NBBO”—is calculated based on data from SIP. While Regulation NMS did not establish a maximum or minimum speed at which the data destined for SIP must be collected and transmitted, it did require that the SIPs transmit such data so as to be received by all market participants at the same time. Provision of data to the HFT firms either more quickly, by way of direct feeds, or simply with detail more enhanced than that provided to the Processor, via enhanced feeds,” place the Exchanges outside of the ambit of their regulatory function. The high fees charged for these alternative feeds are highly profitable, and, as a result, the Exchange Defendants offering these feeds have a clear incentive to direct their resources towards development of these feeds at the expense of the SIPs and the investors who rely on them.

Co-Location

The last of the practices alleged by the Plaintiffs and still in play today concerns a much simpler method of manipulation: physical proximity. Essentially, what is complained of in this instance is the practice of the Exchanges of permitting the installation of servers belonging to the HFT firms at or extremely close to the servers used by the Exchanges themselves. As anyone with a wireless router just slightly too small for their living space has probably realized, the miracle of wireless transmission is in fact impacted by proximity. The fractions of seconds shaved off transmission time benefits the HFT firms in the same manner as the direct receipt of trade data allegedly does.

Like the direct and enhanced data fees, the cost of co-location is typically beyond that which a traditional buy-and-hold investor could afford. The complaint cites to a 2010 report in the Daily Finance which claims that the Exchanges were at that point collecting more than $1.8 billion per year from HFT firms for co-location fees alone. While both alternative feeds and co-location are in theory available to all investors, they are cost-prohibitive for all but those entities making frequent, speculative, short-term investments. And as the complaint notes—why would the HFT firms be willing to pay such high fees, if not for the “informational and technological advantage” that they receive in return?

Watch this space . . .

This summary merely skims the surface of a lengthy complaint and contested motion practice. While I have focused here on the claims that will be moving forward, I encourage anyone interested in the issues here to review the case documents. The Exchange defendants filed their renewed Motion to Dismiss on May 18, so please stay tuned for the Manipulation Monitor’s review of that brief, and Plaintiff’s opposition, in the coming weeks.

This post was written by Alexandra M.C. Douglas.

Stay informed!
Sign up for email alerts and notifications here.

2 thoughts on “Trickery in [High Frequency] Trading

  1. Pingback: Trickery in High Frequency Trading: Part 3 – Lundin PLLC

  2. Pingback: Trickery in High-Frequency Trading: Part 4 – Lundin PLLC

Comments are closed.