« Inverse Floaters, FINRA Fine, Leave HSBC with a Sinking Feeling | Main | FINRA Rule Proposal to Allow Investors to Choose Arbitration Panel with No Industry Participants »

“Quote Stuffing” is a Recipe for Regulation — FINRA Levies Seven Figure Fine Against HFT Player Trillium

By Edward Pekarek, Esq. and Brody Tice

High Frequency Trading, commonly known as “HFT,” is a controversial method of computer-driven trading that can be traced back to the 1980s. Counted among HFT ancestry is “program trading,” widely attributed as one of the causes of the near 25% decline in broad equity indices[1] that occurred on October 19, 1987, a crash known on Wall Street as “Black Monday.”[2] If HFT was born on Black Monday, it is almost certainly the child of a “SOES bandit.”

During the ‘87 crash, brokers ignored ringing telephones and their frantic customers who were calling to place sell orders.[3] NASDAQ responded to this panic-fueled nonfeasance by developing the Small Order Entry System (SOES), a trading platform designed specifically for priority processing of retail customer orders. Unscrupulous traders known as “SOES bandits” would slip large trading volume under the radar by dividing it into smaller pieces that appeared to be retail trades, gaining an ill-gotten speed advantage over other traders’ orders.

Many suspect HFT as the cause of the May 6, 2010 “flash crash,” a stunning hour during which the Dow Jones Industrial Average (DJIA) shed nearly 1,000 points, the largest such intraday point decline ever. While some regulators have scoffed at the notion that HFT algorithms (“algos”) could have driven the DJIA and other indices down at such a rapid clip, questions loom about the ethical and legal implications of this technology and the tens of billions of investor dollars that have flowed out of the domestic equity markets at a steepening pace since the “flash crash.”[4]

Basic HFT Math – Speed Equals Co-Location, Location, Location

HFT has two main elements that distinguish it from many other trading methods. The first is the use of algorithm-based analysis, computer programs that can detect or “sniff,” pending buy and/or sell orders. Critics consider this order flow data “sniffing” to be no different than a modern version of the classic “front-running” schemes Jesse Livermore and other rogue traders used a century ago, where one would step in front of the order of another to profit at the latter trader’s expense.[5]

The second key element of HFT is the sheer speed with which it can detect and analyze these trends and then skim profits from the market. Pundits point to the fact that many HFT outfits have utilized a technique known as “co-location,” which involves physically situating HFT servers at U.S. equity exchanges, for a fee, in order to shrink the gulf the digital data must travel and thereby reduce the time between HFT orders and execution.[6] This unprecedented speed facilitates the frequency – HFT trading programs buy and/or sell not only with lightning speed, but often do so tens of thousands of times during a single trading session. In fact, HFT firms represent roughly 2% of the approximately 20,000 FINRA firms operating today, yet they account for a stunning 73% of all estimated equity trading volume in U.S. markets, although more conservative estimates maintain that only half of all market volume is now the result of HFT.[7]

These elements enable HFT algos to know and act ahead of the competition, hardly a new trend on Wall Street. But pursuing these objectives has always meant navigating a fine line between appropriate and prohibited conduct. HFT blurs the line further, as “black box” algo trading software fires off orders at rates that make retail trading orders seem hopelessly outdated. Revenge of the nerds indeed!

U.S. v. Aleynikov – A US Attorney Notes HFT Might “Manipulate Markets in Unfair Ways”

A former Goldman Sachs HFT programmer, Sergie Aleynikov, was arrested on July 3, 2009 and accused of theft of trade secrets for allegedly downloading portions of an HFT system in development at Goldman to a German website. During Mr. Aleynikov’s July 4, 2009 presentment before U.S. Magistrate Judge Kevin Nathaniel Fox,** Assistant U.S. Attorney Joseph Facciponti argued that bail should be denied and supported his argument by maintaining that, “the bank [Goldman] has raised the possibility that there is a danger that somebody who knew how to use this program could use it to manipulate markets in unfair ways.” It is unclear if there are fair (or legal) ways for one to “manipulate markets” and if one who unlawfully possessed the Goldman trading software could manipulate markets, it seems eminently logical to conclude that Goldman might also have a similar capacity.

The speed with which co-located HFT can trade creates several questionable advantages. In the simplest terms, it allows trades to be executed faster than any other method.[8] The speed of HFT systems also highlights the arbitrage trading tactics employed. HFT profits from even the smallest price inefficiencies – fluctuations between various exchanges and the like that might seem trivial at first glance, at least until one tabulates them.[9] Rather than “swinging for the fences” to realize larger gains with fewer trades, HFT firms generally focus on seemingly obscure price discrepancies and seek to generate a series of “bunts” that generate small profits and high volume, tens of thousands of bunts, day after day after day. Despite the relative size of each individual HFT trade, the ability of a co-located HFT system to spot, analyze and act upon real-time order flow data makes it a potential “man behind the curtain,”[10] perhaps one to whom we should pay no mind and simply enjoy the liquidity HFT provides. However, some pundits suspect HFT has contributed to the recent low volume upward grind for the broader domestic indices since March 2009 in the face of tepid economic news and data.

“Quote Stuffing” is a Turkey of a Trading Strategy

Given the swath of influence HFT cuts in market volume, serious concerns exist regarding the potential to manipulate financial markets using questionable techniques, such as “quote stuffing.” For example, an HFT firm might place a rapid series of trade orders, which by their very existence influences equity prices in the direction of those orders, meaning a rapid series of buy orders in out of the money Apple call options would likely drive up the price of AAPL, at least temporarily.[11] As the market for AAPL reacts to this apparent “opportunity,” an unscrupulous HFT firm might then cancel the illegitimate order(s) while simultaneously buying (or selling) to take advantage of the price increase (or decrease) created by the disingenuous orders it placed previously.

The Financial Industry Regulatory Authority (“FINRA”) recently announced its first high-profile HFT enforcement action, although the focus was on system misuses, not HFT in and of itself. On September 13, 2010, Trillium Brokerage Services, LLC accepted FINRA’s findings of market manipulation in a Letter of Acceptance, Waiver and Consent (AWC), which was based, in part, on “quote stuffing” allegations.

FINRA maintained that Trillium conducted this illicit activity on some 46,000 instances between November 2006 and January 2007, with an estimated gain of $575,000 over the three month period.[12] Fines levied against Trillium totaled over $1 million, with individual sanctions ranging from industry suspensions of six months to two years for Trillium personnel who, according to the AWC, “engaged in a repeated pattern of layering conduct to take advantages of trading, including algorithmic trading by other firms…”[13]

FINRA’s Trillium AWC describes market manipulation beyond that of the long suspected “front-running” scheme of some HFT players. Concerns remain that in the absence of human reason on a trading floor, algo logic may have exacerbated the recent “flash crash.” While various regulatory investigations into the May 6, 2010 one day market break continue, some speculate that HFT fingerprints are all over the mini crash although others suspect the U.S. markets may have been reacting to a terrorist cyber attack that caused the rapid decline.[14] Those who cite HFT as the culprit believe that numerous bearish algo models all converged, leading to the precipitous drop, or at least so goes the theory. While this quick and unprecedented plunge may have terrified traders, it presumably made perfect sense to the programs.

Hello, HAL. Do You Read Me, HAL?

HFT issues create legitimate questions for hand-wringing market watchers and policy makers alike, and perhaps an even greater concern exists regarding the permanent impact HFT may have on the domestic capital markets as billions worth of investor funds have methodically flowed away from the domestic equities market at a record pace since the “flash crash,” with dollar volume down roughly 31% in the last two months as compared to the same months in 2009. As HFT trading comprises a steadily increasing percentage of the total trading dollar volume, it is reasonable to ask what these speed-of-light transactions really mean.

The HAL 9000 computer of Stanley Kubrick’s Space Odyssey observed, “I am putting myself to the fullest possible use, which is all I think that any conscious entity can ever hope to do.”[15] The increasingly prominent role of HFT in the domestic capital markets raises significant policy questions about the proliferation of technology in trading. What role, if any, do time-honored points of fundamental equity analysis such as profit and loss, balance sheet quality and price-to-earnings ratio play in the strategies of an HFT trader? Do stocks, regardless of the underlying issuers’ financial health and/or long-term investment potential, now simply amount to vehicles for the short-term, high-volume games played by HFT? And if so, what does this say about the underlying principles of the U.S. capital markets with a Baby Boomer cohort ready to cash out what is left of their 401K accounts? What might this all bode for future generations?

HAL 9000 also said, “I know you [ ] were planning to disconnect me, and I’m afraid that’s something I cannot allow to happen.”[16]

1 http://upload.wikimedia.org/wikipedia/commons/thumb/a/af/Black_Monday_Dow_Jones.svg/600px-Black_Monday_Dow_Jones.svg.png

2 The “program trading” of “Black Monday” is believed by many to be among the origins of what later became known as HFT. In response to many retail investors being abandoned when market makers stopped accepting calls on October 19, 1987, NASDAQ improved its existing trading execution systems to provide retail investors a preference in the order flow queue. The NASDAQ reform required market makers to instantly enter trades of less than 1,000 shares into what became known as the Small Order Execution System (SOES). Because SOES was given preference, savvy traders also known as “SOES bandits,” spotted an opportunity to trade in smaller volumes with greater frequency, a development which many believe led to modern HFT systems. See Scott Patterson, How the Crash of '87 Gave Birth To High-Frequency Trading, CNBC.COM, Sept. 13, 2010, available at http://www.cnbc.com/id/39038892 (last visited Sept. 21, 2010).

3 Bob Pisani, Man Vs. Machine: How Stock Trading Got So Complex, CNBC.COM, Sept. 13, 2010, available at http://www.cnbc.com/id/38978686/ (last visited Sept. 21, 2010).

4 http://www.zerohedge.com/sites/default/files/images/user5/imageroot/shirakawa/Cumulative%20Fund%20Flows.jpg

5 High Frequency Trading Trading: Harmful for Fundamental Investors, With ONE Exception..., SEEKINGALPHA.COM, Aug. 28, 2010, available at http://seekingalpha.com/instablog/195387-logical-thought/90200-high-frequency-trading-trading-harmful-for-fundamental-investors-with-one-exception (last visited Sept. 21, 2010).

6 HFT systems are typically linked to servers that are within close physical proximity of major exchanges. While this may seem excessive (because of the phenomenal speed of these systems), this is done to gain trade execution advantages that would be lost if the server was located further away. See Man Vs. Machine: Tracing Trades Through Electronic Maze, CNBC.com, Sept. 15, 2010, available at http://www.cnbc.com/id/38974297/page/2/ (last visited Sept. 20, 2010).

7 See Allen Zaydlin, HFT is Fair Game, ADVANCED TRADING, Sept. 15, 2010, available at http://www.advancedtrading.com/
(last visited Sept. 21, 2010).

8 Id.

9 At one New Jersey-based HFT outfit, a “slow” day meant the firm had only traded 21 million shares for a $7000.00 profit. See Electronic Maze, supra note 3.

10 Memorable quotes for The Wizard of Oz, INTERNET MOVIE DATABASE, available at http://www.imdb.com/title/tt0032138/quotes (last visited Sept. 21, 2010).

11 According to recent remarks by Securities and Exchange Commission Chairman Mary Schapiro, “The SEC and other regulators are looking carefully at certain practices in this area to assess whether they violate existing rules against fraudulent or other improper behavior.” Jonathan Spicer and Herbert Lash, SEC Probes “Quote Stuffing” Practices: Schapiro, ABC NEWS.COM, Sept. 7, 2010, available at http://abcnews.go.com/Business/wireStory?id=11576459 (last visited Sept. 21, 2010).

12 See Financial Industry Regulatory Authority, Letter of Acceptance, Waiver and Consent (AWC) No. 20070076782-01, available at
http://www.finra.org/web/groups/industry/@ip/@enf/@ad/documents/industry/p122044.pdf (last visited Sept. 21, 2010).

13 Brandon Rowley, Trillium Fine Nothing To Do With High Frequency Trading (HFT), ISTOCKANALYST, Sept. 14, 2010, available at http://www.istockanalyst.com/article/viewarticle/articleid/4499037 (last visited Sept. 21, 2010).

14 Some claim HFT had nothing to do with the “flash crash.” See CME Absolves High-Frequency Traders in Flash Crash, NY TIMES DEALBOOK, May 16, 2010; John W. Labuszewski and Richard Co, CME Group Report, What Happened on May 6th?, available at http://www.scribd.com/doc/31546905/CME-Group-Report-on-the-Flash-Crash (last visited Sept. 21, 2010). However, many market observers continue to posit that HFT had something, if not everything, to do with the massive one day drop. See Graham Bowley, Stock Swing Still Baffles,With an Ominous Tone, NY TIMES, Aug. 27, 2010, available at http://www.nytimes.com/2010/08/23/business/23flash.html (last visited Sept. 21, 2010); Tyler Durden (pseudonym), How HFT Quote Stuffing Caused The Market Crash Of May 6, And Threatens To Destroy The Entire Market At Any Moment, ZEROHEDGE, June 23, 2010, available at http://www.zerohedge.com/article/how-hft-quote-stuffing-caused-market-crash-may-6-and-threatens-destroy-entire-market-any-mom (last visited Sept. 21, 2010).

15 Memorable quotes for 2001: A Space Odyssey, INTERNET MOVIE DATABASE, available at http://www.imdb.com/title/tt0062622/quotes (last visited Sept. 21, 2010).

16 Id.
* Mr. Pekarek is a Visiting Professor and Supervising Attorney at Pace Law School and serves as the Assistant Clinic Director for the non-profit Pace Investor Rights Clinic of John Jay Legal Services, Inc. **He served as a law clerk for the Hon. Kevin Nathaniel Fox during the criminal proceeding discussed above.

Mr. Tice is a third-year law student at Pace Law School and a student-intern for the non-profit Pace Investor Rights Clinic.

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)


This page contains a single entry from the blog posted on September 23, 2010 2:31 PM.

The previous post in this blog was Inverse Floaters, FINRA Fine, Leave HSBC with a Sinking Feeling.

The next post in this blog is FINRA Rule Proposal to Allow Investors to Choose Arbitration Panel with No Industry Participants.

Many more can be found on the main index page or by looking through the archives.