Algo Trading: Life in the Cross-Hairs

A recent paper by three attorneys of the multi-national law firm of Covington & Burling LLP says that algorithmic trading is now “squarely in the cross-hairs of securities and commodities regulators,” in large part due to the flash crash of May 6, 2010.

The authors/attorneys are: David L. Kornblau, Allison Lurton, and Jonathan M. Sperling (respectively, as below). Their paper, which appeared in the Securities Regulation & Law Report in February, cites a joint report on the flash crash by the Securities and Exchange Commission and the Commodity Futures Trading Commission, issued September 30, 2010, that blamed the sudden loss of 700 points on “automated execution programs and algorithmic trading strategies.”

The intensified focus has had enforcement consequences. For example, the paper notes, FINRA fined and suspended 11 of the employees of Trillium Brokerage Services LLC, and fined Trillium itself $1 million for alleged manipulation and related matters. More recently, as Kornblau et al observe, “the Department of Justice and the SEC launched an investigation into whether high-frequency traders are placing and cancelling waves of orders in an attempt to manipulate the market.”

The authors seek to place this new intensified scrutiny into the context of an old dispute over what constitutes actionable market manipulation. One theory holds that a trader’s intent in making a specific trade is the critical factor. If the trader’s “sole intent” in making even a quite ordinary buy or sell order is to move the price, then the resulting trade is market manipulation.

The D.C. Circuit upheld this theory in 2001, in Markowski v. SEC.

Looking for Something More

But the Second Circuit took a narrower view of what constitutes manipulation, in ATSI Communications v. Shaar Fund (2007). It said the actual trade must be combined with “something more”specifically, an effort to put inaccurate information into the marketplace. A classic boiler room pump-and-dump includes, as the hyphenated phrase suggests, both of these elements. The manipulative trade, the “dump” is preceded by an effort to talk up a stock, the “pump.”

Covington’s lawyers say, though, that firms that engage in algorithmic trading should not draw too much comfort from such precedents as ATSI Communications. The SEC continues to espouse the “sole intent” theory.

“Unless and until the law in this area is clarified by the Supreme Court or Congress, the SEC and FINRA are unlikely to limit their market manipulation cases to ‘pump and dump’ and other schemes involving the injection of false or misleading information into the market place.” Rather, regulators may well investigate whether the trading programs themselves manifest manipulative intent.

Signs of manipulative intent likely to matter to the SEC include: order cancellations, financial motives aside from gain or loss on the specific transactions at issue, and market domination.

The lawyers also observe that overly colorful emails or instant messages will continue to pose problems.

In the actions against Amaranth, for example, certain i-ms sent by Brian Hunter played a prominent part. Hunter had messaged a natural gas trader at another firm that a certain trade of his was a “bit of an experiment mainly,” and that he was “waiting until 2:20” to complete his selling. This suggested that he was experimenting with painting the close. As Covington’s lawyers note, this sort of thing can “cut through layers of complexity.”

Amaranth also illustrates the role that may be played by an indirect financial motive in an action for manipulation. Hunter was allegedly painting the Nymex tape in order to improve the value of his positions at the IntercontinentalExchange (ICE).

Bottom Line

What can and should algorithmic traders do to reduce the likelihood that they will become the targets of an investigation for market manipulation? As many firms have discovered over the years, the eventual outcome of the litigation is not the main event – as soon as you have become a target of an investigation, you have in an important sense lost. The direct and indirect costs will prove enormous.

Covington’s lawyers offer several suggestions. For example, firms have to have an internal look-out for the sort of suspicious trading activity that regulators would zero in on. Better you know it’s going on before they do.

As to comments like “a bit of an experiment,” the lawyers suggest a firm should review its traders’ communications on a regular basis to “help flag actual instances of problematic trading or identify individuals with poor e-communications practices.”

Be Sociable, Share!

3 Comments

  1. Peter Urbani
    March 8, 2012 at 7:13 pm

    In the good old days of open-outcry single-capacity floor-trading under which I came up through ( The ‘Big Bang’ deregulation which first allowed dual capacity happened in London in 1986 but only in 1995 in South Africa ) a dealer had to be good for a minimum of 100 shares at any price they bid for. There was also the uptick rule which required that short sales only take place after a price uptick – abolished in 2007. Whilst a return to these gentler times, before the zeitgesit of ‘rip your clients face off’ became the order of the day, may be impractical it does strike me that a single constantly recalculated on-balance-volume ruling price would simplify things greatly, remove the confusion and cost of the spread and also clearly delineate what is an agent and what is a principal trade*. Under this scenario the ruling price would always be the weighted average price of all non-limited buying and selling orders currently in the system. They would be filled automatically and commissions allocated to exchange members on a pro-rata to their number of seats/shares basis. This would have the effect of making all orders ‘real’ at the minimum order size and hopefully reduce the scope for manipulation by spurious later cancelled orders.

    *The single OBV Ruling Price (RP) would always be assumed to be the agency price and any prior buying of shares or ‘warehousing’ in the old vernacular would be considered to be a principal trade at the buyer / sellers risk and subject to wherever the Volker rules end up. participants would not however be able to hide behind the market / liquidity making excuse / mantra currently being used to obfusticate the issue.


  2. Chris Vermeulen
    August 25, 2014 at 7:16 pm

    I would love to see a follow up piece on this now that we are two years down the road. Has there been an increase in identifying and going after problematic trading?


Leave A Reply

← Video: Ken Shoji, Founder & Managing Partner, Stissing Lake Advisors Yes, the Bubble Burst: That’s What Bubbles Do →