Panic sell in the stock market: Concerns over Greek debt or “the machines just took over”?

May 7, 2010

An interesting commentary appeared on BBC news about yesterday’s plunge in
US stock markets due to Greece’s continuing debt crisis:

“Computer trading is thought to have cranked up the losses, as
programmes designed to sell stocks at a specified level came into
action when the market started falling. ‘I think the machines just
took over,’ said Charlie Smith, chief investment officer at Fort Pitt
Capital Group. ‘There’s not a lot of human interaction. We’ve known
that automated trading can run away from you, and I think that’s what
we saw happen today.’”

Here the trader differentiates between two kinds of “panic” process
that both appear to the observers of the market as falling stock
prices: selling spells generated by machine interaction versus human
interaction. He assures that this time the plunge happened because the
machines were trading. This is a different kind of panic than what we
conventionally think of, one that is based on expectations about
European government debt, which escalates as traders are watching each
other’s moves, or more precisely, “the market’s” movement. Which kind
of panic prevails seems to be specific to the trading system of each
type of market. Another trader reassures us that today’s dive was “an
equity market structure issue, there’s no major problem going on.”

It is interesting that the traders almost dismiss the plunge as a periodic
and temporary side-effect, automated trading gone wild. Real problems
seem to emerge only when humans are involved. But if machine sociality
can crash a market and have ripple effects to other markets, then
perhaps the agency of trading software should be recognized.

12 Responses to “Panic sell in the stock market: Concerns over Greek debt or “the machines just took over”?”

  1. danielbeunza Says:

    Zsuzsi — very interesting, and I would go beyond. Your point, if I understand correctly, is that the divide between price movements caused by the computer and caused by human investors is ultimately an artificial one. Beautiful actor-network argument. In this case, however, it seems that the problem was due to a trading glitch, compounded by legitimate machine algorithms. So the NASDAQ is canceling trades that are 60 percent above or below the closing price.

    One immediate question concerns the politics of this “undoing” of the “not real” trades done by the computers/glitches. Why 60 percent? What if your operation was also due to a glitch, but happened at 59 percent?

    Another point concerns the competition between exchanges. It is notable that this happened at the NYSE (which has a people-based trading floor) as well as the NASDAQ, which does not. What does this event mean about the expansion of electronic trading introduced by recent regulatory reform?

    • marclenglet Says:

      Daniel – a short comment on the point you make on the 59/60% limit, which I think is a decisive one.

      In fact, this kind of issues, where the market materiality is being qualified as real/genuine or not, is usually met by both investment services providers and regulators when they, for example, decide the level above which they face a market manipulation. Designing tools in order to track such manipualtions is all about the fine-tuning of indicators and the relating levels above or under which a practice will (or not) be considered an abusive one. Boredering/junction effects play a huge role here

      Beyond these discussions on the levels indicating a manipulation – and the regulatory texts usually do not want to provide precise figures – remains the “reality principle”: fishing too many operations would make it impossible for controllers to process let’s say 300 trades a day, going into the details of what happened, why, whit whom etc.). And on the other side, intermediaries and markets have to demonstrate their commitment to fight market abuse.

      So this simple question just underlines how difficult it can be to describe markets when their own “reality” is at stake.


  2. The investment is certainly subject to risks, whether their investment strategy includes investment funds and brokerage stocks of NSE or other investment options. According to market volatility and the decisions of your purchase, you will reap the profits or losses. All investors are part of the profit and loss of game.

  3. zsuzsannavargha Says:

    Daniel thanks for bringing up the issue of revising the trades. The regulatory dilemmas for determining manipulation really seem to grapple with the same question, what is the reality of the market dynamic and what is just a non-market technical error.

    I wonder if there is something specific about electronic markets in how the “non-market” technical problems are treated: you can switch on the time machine. You can go back to that selected 15 minutes and pick out a pre-defined set of deals and erase them from the flow of events.

    And maybe I’m wrong but in the Dow case the puzzlement around the 60% suggests that first a general judgment was made about the movement–whether it is the normal working of the market or not–and then the threshold of 60% was somehow set. So not like in the decision on manipulation, where if I understand correctly, the indicators are worked out in advance, precisely to stand for a definition of the situation. Perhaps it’s because the exchanges have not come up with rules or indicators on machine (mis)behavior?

    This is the CNBC video report http://www.cnbc.com/id/15840232?video=1487347348&play=1. It really brings the moment close as the Dow flips downward at rocket speed. At first the commentators were not thinking it can be a glitch. They saw fear and as a consequence, evaporating buyers.


  4. […] גם דיווחים כי מדובר בתקלות מחשוב, או מעניין יותר – בתוצאה של שימוש בתוכנות מסחר אוטומטיות המוכרות מנייה משזו […]

  5. jck Says:

    “It is interesting that the traders almost dismiss the plunge as a periodic and temporary side-effect, automated trading gone wild.”
    It is a side effect of *regulations* governing automated trading. If the market is hit by a large public order like a stop or a large margin call or a “fat finger” event, it is impossible for (automated) market makers to comply with rule 610/611 of reg NMS, therefore they pull out, further market orders will be filled in a vacuum wherever there is a bid if any, just like the proverbial shoeshine boy in 1929 who bought stocks at $1 that were going for $100 2 days before.
    Complying with the rules is difficult and fails happen nearly every day and there are clear rules to handle this, for ex, if a $55.00 stock was to trade up or down more than 3% from the previous trade it is a “clearly erroneous trade” and the injured party can ask for a cancel within a short time frame.
    Thursday was a system wide event (probably margin calls, but just guessing), in which case Nasdaq can invoke rule 11890(b) that allows it to do whatever it wants without appeal, in this case cancel all trades +/- 60% away from the last reported price on the consolidated tape at or before 14:40pm. It should be obvious that this very arbitrary rule further discourage market makers to step in since they don’t know if the trades will hold.

  6. Yuval Millo Says:

    Nice post!
    Yeah, as you imply, Zsuzsi, making a distinction between human and non-human agency is pretty much futile, both from a sociological and a regulatory perspective. In fact, I would guess that the combination of human and non-human is responsible for the fact that prices were sticky downwards and didn’t bounce all the way back, which, you would expect if it were only an erroneous trade and traders would be quick to take advantage of the under-valued stocks. Maybe we see here another version, in a smaller scale, of the volatility smirk/smile effect in stock options post October 1987?

  7. zsuzsannavargha Says:

    jck thanks for clarifying the rules and regulations that were involved here. In brief, everyday rules are in place for establishing the “marketness” of individual stock movement. But the 60% was not an existing rule for handling that. First the “event” was defined as system wide extreme volatility, which then invoked the do-what-you-want regulation and the post-hoc limit…

    By the way, this sentence could have been written by you: “some experts warned that Wall Street’s effort to undo tainted trades could encourage the very behavior that led to the market plunge”.

    (http://www.washingtonpost.com/wp-dyn/content/article/2010/05/07/AR2010050705087.html?hpid%3Dmoreheadlines&sub=AR)

    But these experts turn out to explain the drop with high-speed trading (algorithmic selling was in place triggered by pre-set price limits) and not margin calls or the market maker problem. They argue the firms doing fully automated trading *badly* (i.e. mindlessly, I suppose) shouldn’t get away with it, but they do if the trades are canceled. The respondents seem to have a particular human axe to grind…

    Yuval, on regulatory perspective it will be interesting to see the results of the SEC (now CSI) “forensics” of the crash. And the stickiness issue is a good question. What would the post-1987 volatility smile say about the current mix of rules, traders and algorithms?

    P.S. thanks to Marc L. for the insightful “reality” discussion. This was left off in my previous response.

  8. Juan Pablo Says:

    Hi everyone,

    This was a great and timely post (unlike my reply!). Algorithmic and high frequency trading are fantastic topics for SSF. On this particular case, I wonder if one of the causes of the fall wasn’t the differentiated standards on circuit breaking in extreme situations. There seems to be little written on how different trading venues apply circuit breaks in the event of a large fall, and how differentials in such application can actually exacerbate price movements (the explanation could go as follows: as some participants start to leave the market, there is less liquidity and more unfulfilled orders, possibly providing algorithms with signals that lead them to a frantic race to the bottom). This case might also highlight the limited diversity of algorithms running in the market or, more tellingly, the existence of conditions under which the diversity decreases rapidly leading to extreme events.

  9. zsuzsannavargha Says:

    Juan Pablo, you’re right and that’s exactly how the exchanges are beginning to treat it, as a problem of different slowing-down rules that should be harmonized. Good point on the limited diversity, I wonder if Daniel and David Stark’s finding on reflexive modeling would be applicable here as well.

  10. nar price Says:

    machines work on patterns and algorithms they do everything in milliseconds, what happens to the small individual investors ?


  11. […] point on HFT missing from mythical accounts of Wall St, and continuing thoughts from an earlier post on the May 2010 stock market plunge: here is an interesting commentary from Ars Technica on High […]


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: