Alt. title: Barclay’s Algo + Waddell & Reed Futures Sale = Flash Crash?

The 104 page report by the staffs of the U.S. Commodity Futures Trading Commission and the U.S. Securities and Exchange Commission CFTC & SEC was released this week (PDF here).

I am still digesting the entire writing, but a few things leaped out. The report states that the flash crash was set off:

“At 2:32 p.m., against this backdrop of unusually high volatility and thinning liquidity, a large fundamental trader (a mutual fund complex) initiated a sell program to sell a total of 75,000 E-Mini contracts (valued at approximately $4.1 billion) as a hedge to an existing equity position.”

The seller was Waddell & Reed, using a standard Barclays algo.

That may have been the spark that lit the fumes, but it does not address the structural flaws in the market which is the ongoing gas leak. Nor does it give us much confidence that it is unlikely to occur again soon.

Consider this WSJ description of how a SKYynet feedback loop developed and caused the crash:

“As the Waddell trade hit the futures markets . . . likely buyers included high-frequency trading firms. A key feature of high-frequency trading firms is that they quickly exit trades and, by 2:41, they were also aggressively selling the E-mini contracts they had bought from Waddell, which was still trying to sell the remainder of its contracts . . .

“HFTs began to quickly buy and then resell contracts to each other—generating a ‘hot-potato’ volume effect as the same positions were passed rapidly back and forth,” the report says. At one point, HFTs traded more than 27,000 contracts in just 14 seconds—a huge amount.The Waddell algorithm responded to the high volume by picking up the pace of its selling, even though stocks were spiraling lower. This feedback loop of selling by Waddell, high-frequency traders and others helped drive the E-mini price down 3% in just four minutes.”

What I read into this is a system devoid of human judgment or rules as the underlying structural factor.

There is no adult supervision, only bots and silicon. Instead of putting SKYnet in charge of national defense, we have put it in charge of our markets and economy. The end result — minus the spectacular special effects — seems to have been the same.

Over my career, I have criticized specialists for raping and pillaging various orders at will. But a system without humans charged with maintaining orderly markets, with software bots swapping shares with other silicon-based life forms, is what has replaced that. Perhaps the cost of orderly markets was the specialist’s license to steal.

Meanwhile, SKYnet sits . . . and waits.


May 6 2010 Trading

click for larger chart

chart courtesy of

Sources and additional reading:
How a Trading Algorithm Went Awry
October 2, 2010

Lone $4.1 Billion Sale Led to ‘Flash Crash’ in May
NYT, October 1, 2010

Waddell Shows Long-Term Fund Can Unhinge Wall Street
Christopher Condon
Bloomberg, October 1, 2010

Futures Sale Spurred May 6 Panic as Traders Lost Faith in Data
Nina Mehta and Whitney Kisling
Bloomberg, October 2, 2010

Category: Regulation, Trading

Please use the comments to demonstrate your own ignorance, unfamiliarity with empirical data and lack of respect for scientific knowledge. Be sure to create straw men and argue against things I have neither said nor implied. If you could repeat previously discredited memes or steer the conversation into irrelevant, off topic discussions, it would be appreciated. Lastly, kindly forgo all civility in your discourse . . . you are, after all, anonymous.

21 Responses to “SEC/CFTC Report: SKYnet Caused Flash Crash”

  1. Cyberdyne Systems says:

    In the Terminator series, Skynet is the main antagonist — an artificially intelligent system which became self-aware and revolted against its creators. Skynet is rarely seen onscreen and its actions are often performed via other robots and computer systems; usually a Terminator.

  2. cpd says:

    You should read the CME statement yesterday. It details what happened and makes the SEC report look like a load of crap. Also, as Denninger pointed out yesterday, 71,500 ES contracts traded right at the market close the other day and nothing happened. In other words, the number of contracts Waddell traded was not really that significant. The big question is – how much lower can the credibility of the SEC go? They are beyond useless.

  3. Thanks CPD for the heads up — I dug up the CME statement — its here:

  4. dead hobo says:

    Good story. The SEC report looked like crap, but a lot was too technical for me to understand completely. I suspect it will become a new public embarrassment for the SEC after all the analysis is posted.

    Check your email.

  5. cognos says:

    Looks like a great oppty to make money.

    Can we please do this again? I would like to buy more AAPL at 200 and make 20% in a matter of hours.

    Its wierd that Jeremy Grantham was the only guy in early 2009 pointing out that companies, cashflows, and even his services… were “on-sale” at about 50% off the previous year’s price. Course, in classic GMO fashion he turned back to bearish around 900 (and l-t perf kinda sucks as it does for all perma-bears). But still — high volatility, HFT, prices of $0.01 for ACN or 50% off for PG — doesnt anyone else see this as an “easy” game?

    Buy… the dips.

  6. KidDynamite says:

    The CME statement is funny – they are basically trying to say “hey man – don’t try to pin this on OUR market” – and I’m not saying it’s their fault at all, but the same thing happened right after the Flash Crash where NYSE and NASDAQ were throwing barbs back and forth, with NYSE saying “our LRPs did their job” and NASDAQ saying “NYSE’s LRPs were instituted, which screwed everything up” (LRP = Liquidity Replenishment Point, aka, a temporary pause in trading)

    There’s something that’s essential to notice, though – and that’s that guys like Zero Hedge and Denninger, who have repeatedly pointed out that VOLUME and LIQUIDITY are two different things seem to have forgotten their own lessons. 75k E-minis isn’t a lot of volume – but it requires a lot of liquidity. The SEC concluded something similar in their report when they talked about large volume printing, yet shares essentially being traded back and forth amongst electronic market makers.

    All I’m saying is that the “75k is not a lot of contracts argument” is nonsense -a $4B sell order into a market that was already teetering on edge after a straight line 20% 12 week rally, provided tinder for a perfect storm.

    Combine this with the quote that Barry pulled above:

    ““HFTs began to quickly buy and then resell contracts to each other—generating a ‘hot-potato’ volume effect as the same positions were passed rapidly back and forth,” the report says. At one point, HFTs traded more than 27,000 contracts in just 14 seconds—a huge amount.”

    and you can see how the sell algo, trying to be 9% of the volume, and needing real liquidity on the other side of its trade quickly spiraled downward… Guys stepped in to buy the E-minis, and sell the underlying cash basket, which is how it spread to the underlying stocks… and so it goes.

  7. zdog says:

    How long do HFT algos typically hold a position?

    Couldn’t an expanded capital gains structure, say up to 99% if a position (bet) is held under 1s, help attenuate the amplitude of the swings?

  8. [...] SEC/CFTC Report: SKYnet Caused Flash Crash [...]

  9. DeDude says:

    The problem with a bot trader is that it is like a super bureaucrat. It has a set of rules that it follows regardless of the outcome. Even when the outcome is clearly destructive to its overall mission it continues to follow the “rules and regulations” dictated by the programmer. The programmer faces the same challenges that lawmakers and elected officials face – it is very difficult to predict all contingencies, abnormalities and “black Swans” that can turn a normally sensible action into a counterproductive action. Lawmakers sometimes try to prevent destructive bureaucracy by making “rubber rules” that allows human judgment in the enforcement of a rule (which may work or backfire depending on the common sense of the bureaucrat). Unfortunately we are not at the point where you can give human judgment to bots (current surrogate “judgment”, is just another fancy program).

    The solution is still to tax or ban bot activities that only serve to enrich the rich at the expense of regular people. Some may say that HFT is fine because it is just one speculator ripping off another speculator, but that is simply not true. A lot of regular people who had sensible strategies that include stops and cost averaging ended up losing money in the flash crash. Unless someone can point out that HFT creates something good for society as a whole (in excess of the harm it clearly creates) it should be banned or taxed out of existence. I am still waiting for someone to explain to me why we should not ban sales of stocks in any other way than as physical paper (you must poses to buy or promise a sale) and ban sales of any share more than once a day. I know speculators would lose a lot in that scenario but what else but gains would there be for society as a whole?

  10. Tarkus says:

    There have been mini version of flash crashes in individual stock (PGN, AAPL, LQD). Does the SEC blame W&R for those too?

    Or is it so obvious that there is a structural problem that it escapes the SEC’s notice – like a Bernie Madoff.

    Maybe this is the SEC’s market-structure version of Bernie.

  11. KidDynamite says:

    Dedude: “The solution is still to tax or ban bot activities that only serve to enrich the rich at the expense of regular people.” Should we ban all traders who are only out for their own profit? Should we only allow people to take part in the stock market if they pass a test showing that they know a certain amount about a company – that they are a real true knowledgeable investor, and not a trader?

    “A lot of regular people who had sensible strategies that include stops and cost averaging ended up losing money in the flash crash.” – we definitively need to increase investor intelligence, and make sure that “regular people” know what MARKET ORDER means. I personally think we should ban market orders to prevent stupid people from their own lack of sophistication, but…

  12. soloduff says:

    BR says, “What I read into this is a system devoid of human judgment or rules as the underlying structural factor.” Inaccurate! In fact the system is an institutionalization (“rules”) of the “human judgment” (preconceptions) of the biggest players in the game of financial capitalism (“the underlying structural factor”).

    This is but a grand version of the broader cultural degeneration into thinking and communicating in stereotypes–with an a perceived premium on getting one’s stereotype into play in a quick semiconductor instant. The absurdity, of course, is the illusion of control sowed by the smartest guys in the asylum.

    Are we having fun yet?

  13. mikaeel says:

    The whole thing is a crock of dodo. They got an old saying, “If you can’t dazzle them with science, baffle them with a government report.

    Twenty years from now we’ll find out this whole thing was planned on Jekyll Island about 97 years ago.

  14. louis says:

    Live video feed from Waddell and Reed’s Kansas City Office

  15. mbelardes says:

    So … if the conclusion is “this is likely to happen again” does that mean I should set a series of open buys 30% lower for stocks like PG?

    Structural flaw? Yes. Should we fix it? Yes. Will it be fixed? No.

    Sounds like some sort of black swan buying opportunity.

  16. Except they are likely to be cancelled

  17. KidDynamite says:

    and therein lies the main problem. Canceled trades. Canceling them each time extends and prolongs the moral hazard. If we wanted to stop flash-crashes, and we don’t want to stop canceling trades (because Retail is the one getting hosed) then banning market orders would go the furthest the fastest. Every order must have a limit price attached to it – then there’s no “oh, I didn’t know i might get filled at a lower price” nonsense… There’s still “fat finger” potential, but the market order errors would be eliminated.

  18. ACS says:

    According to the report, less than half of the 75,000 contracts was sold on the way down. That really isn’t a lot of volume in the S&P futures. Clearly the W&R selling was merely the light to a market already doused with gasoline.

  19. Tarkus says:

    Maybe if the market does another flash-crash like May 6th we can name the phenomenon after Mary.

    Like, if a lot of stub quotes get hit at the same time, and the exchanges cancel them at “whatever-percent-we-feel-like-that-generates-most-commissions”, the people screwed over can start saying

    “Hey!! I was Mary’ed!!”

  20. Man Vs. Machine: Commentary—How the SEC Helps Underwrite High Frequency Trading
    Friday, 1 Oct 2010 | 10:44 AM ET Text Size

    The SEC’s focus on high frequency traders should be instead on the proliferation of high frequency products. The traders are the good guys in the Man versus Machine debate. My worries are about the 30-some sponsors of Exchange Traded Funds (ETFs) who are said to be creating all sorts of new ready-to-exploit pricing inefficiencies in small capitalization stocks, sectors, industry ETFs, ETF futures, ETF options, and ETF fund-of-funds.

    While the SEC openly worries about the speed of trading, staff there facilitates the rapid birth of high frequency products in increasingly exotic ETF packages. I am convinced that the May flash crash of stock values owes the day to ETFs and the high frequency traders who enable the product to exist.

    The wisdom of the New York Stock Exchange, which has long used a time out to clear order imbalances, and has been widely emulated in the German stock market, broke down the physical hedging needed to take the other side of the cascade of institutional ETF trades.

    Ironically, the regulator’s initial fix was to ignore ETF order imbalances. That’s where supply and demand equilibrium should be managed—they are after all derivative securities.

    Has anyone stopped to consider what happens when the fourth largest physical holder of gold bullion needs to begin destroying ETF creation units? Will the buyers of physical gold be there to unwind these ETFs as fast as they have been created over the past three or four years of unprecedented pessimism toward paper currencies?

    We can’t seem to list more U.S. companies, so instead we create enormously complex packages that wrap the same finite universe of securities and call it innovation. It reminds me of the whole financial engineering mess in mortgages.

    The opportunities provided hedge funds to exploit fleeting arbitrage opportunities between securities held in these packages generates knock-on effects that have been little discussed. Richard Bookstaber in his seminal 2007 book A Demon of Our Own Design talks about tightly-coupled systems which, if left unattended, will regularly produce events like the flash crash (where 65 percent of cancelled orders were for ETFs). The front flap of his far-sighted book describes the increasing complexity of our financial markets “that is ever edging toward disaster.”

    The speed with which we trade securities has little to do with these effects. The introduction of an amazing array of derivative securities against a static number of underlying physical equity securities suggests that a stock’s relative value in a package may be more important than its intrinsic value.

    The hedge funds and high frequency trading community worry only about temporary price discrepancies of a stock, or option, or future held in a variety of synthetic packages and overseen by two separate regulators.

    I’m told that many hedge firms have established sophisticated techniques that set up trades with little market risk, thus allowing them to stomp on the gas or the brakes of small company stocks held in a glistening array of ETF packages—all the way from broad indexes to lithium or palladium packages.

    The July 2010 report by Empirical Research Partners illustrates that the cap-weighted correlation among large cap stocks has been 60 percent for a good part of this year—a level exceeded only twice in 84 1years (1929 and 2009). The data illustrates that correlations have been rising most of the decade in both the broad market and within sectors like financials, capital goods and transportation.

    When financial assets move in highly correlated ways that should be the worry for regulators. This is a sign that markets may be misallocating capital and failing to discriminate among investments by not properly disciplining risk and rewarding success. . .


    Harold Bradley is the Chief Investment Officer, The Kauffman Foundation

  21. bjorn says:

    Has anyone else had this experience?? See this by Richard Shaw: