The Flash Crash: A Cautionary Tale

So the finger that caused the “flash crash” May 6 turns out to have been merely heavy instead of fat.

An order to sell $4.1 billion of stock index futures as quickly as possible, and never mind the price, spooked an already-jittery market into a spectacular selling frenzy that day according to a long-waited forensic report by the Securities and Exchange Commission that was released last week.

The “hot-potato” selling among high-frequency trading programs that ensued cost the US stock market nearly $1 trillion in seven minutes, a 9.16 percent fall in the Dow Jones Industrial Average, before the broad market recovered to close down about 3 percent on the day.

The Midwest mutual fund that issued the order was quickly identified by several newspapers as Waddell and Reed Financial Inc., of Overland Park, Kansas.  The order was deliberate, not the result of a “fat finger,” or mistake, that might have added an extra zero or two.

Back in May, the firm had stated that it had no intention to “disrupt the market.” But the consensus among traders seemed to be that to enter so urgent an order to sell, irrespective of price, was, as one trader put it, “borderline irresponsible.”

Many traders complained that the report didn’t go far enough in assigning blame or proposing remedies or even disclosing various key details surrounding the lightning trade.  The SEC promised to say more in due course.

For a titillating story while you wait, consider “Signals from a Very Brief But  Emblematic Catastrophe on Wall Street,” a new essay by Paul David.

David, as many EP readers know, is an emeritus Stanford University economic historian of prodigious learning (with faculty appointments in Paris and the Netherlands. as well), best known for a couple of famous papers. In Clio and the Economics of QWERTY, in 1985, he introduced the concept of “path dependence” to explain, along with much else, how the typewriter keyboard remained the same after more than a century of engineering improvements (though people are still arguing about the details).

In 1990, in the The Dynamo and the Computer: An Historical Perspective on the Modern Productivity Paradox, he argued that familiarity with the commercial adoption of electricity between 1880 and 1930 (lighting first, power only with a substantial lag) suggested that still more productive applications of computers would be found in the future. (The commercialization of the Internet and the invention of the World Wide Web lay just head.)

In other words, Paul David has a very good track record when it comes to computers. And he certainly knows how to tell a story.  “Signals” reads like a cross between an Edgar Allen Poe short story and a Michael Crichton thriller.

What caused the alarming fall in prices on the US stock market? A growing mismatch between “hyper-connected systems” among individuals, not just market participants, and the administrative rules and cultural norms that govern their conduct, is what worries David. This is the signal he believes the flash crash sent, and to this end he reconstructs the event.

True, David has no history of writing on finance. He acknowledges the assistance of a trio of guides, professors Richard Sylla and Charles Calomiris, and an extraordinary student advisee, Elizabeth Stone.  He does a credible job of describing “liquidity refreshment points” and “stub quotes” (though he errs when he predicts that the SEC investigation will prove to be unsatisfying because “no actor will be found to have been culpable”) David is, after all, one of the world’s preeminent historians of technology; his narrative reads like literature:  in fact, it is literature.|

Still, if it’s the crash itself that is of interest, you’ll do better reading the SEC report.  It is the larger dimension of the signal that interests David. He invokes the work of French mathematician René Thom, inventor in the 1970s of a formal mathematics of sudden shifts, of qualitative breaks or discontinuities, that he called catastrophes.  “To appreciate that quality it is helpful to start from the mathematical rather than the ordinary language meanings conveyed the term,” David writes.

A commonplace physical illustration of a “catastrophic event” – in this formal sense of the term — may be experienced by letting your finger trace the surface of a draped fabric until it reaches a point where the surface (the “manifold” as mathematicians would speak of the shawl or cloak’s three-dimensional surface) has folded under itself; there gravity will cause your finger’s point of contact to drop precipitously from the surface along which it was traveling smoothly – to land upon the lower level of the drapery beyond the fold. That little passage is the “catastrophe.” In the present context, what is especially relevant about this conceptualization of the “event” experienced by your finger is its generic nature: catastrophes thus conceived are not phenomena belonging to a category delimited by some size dimension of the system in which they occur, or according to the severity of their sequelae; nor are they to be uniquely associated with processes that that operate only in one or another range of temporal velocities (whether slow, or fast). Instead, the catastrophes to which this essay’s title refers are fractal, possessing the property of self-similarity.

Similar to what? He compares the conditions that led to the May 6 price break to the phenomenon known as “flaming.” One party breaks off a previously civil exchange with hostile abusive comment.  Instead of turning it aside, others sometimes reciprocate.  Anonymity is the key part of the process; isolation seems to remove inhibitions that would brake the process if the exchanges took place face to face.  Silence ensues, or, worse yet, digital versions of what long ago were called “slam books” – compendia of the faults of others, sufficient to fragment the conversation once and for all. (This is the problem with Glen Beck and much of the rest of Fox News.)

Humans evolved to recognize from infancy the effect of their actions on others – facial expressions, body language, tone of voice, he notes.  On the Internet, however, no one knows when they may come across as a flame-throwing tank. It is why humor is so risky on the Web.

David takes his parable in a different direction. There is a parallel, he says, between what happens when human beings are deprived of feedback from their own communications what happens in the communication of diseases. Pathogens transmitted person-to-person – tuberculosis, say — have evolved in the direction of diminished virulence, as opposed to those borne by insects that have had no previous contact with a human host. “Communities of dependence” between carrier and host tend to diminish the bad results. In each case, feedback – what traders call liquidity – is your friend.

“Signals” is not likely to appear in an academic journal, or even the New York Review of Books or the London Review of Books or the Times Literary Supplement. (The essay first appeared in the European online journal Real World Economics.) The SEC may have the details, at least some of them. Yet this wise warning deserves a wider audience. Paul David knows how to tell a story, and, as noted, he possesses a proven intuition. The fragmented world of financial markets may be the lesser danger. Intransigence, the breakdown of political back-and-forth, probably is the greater threat.  Civic liquidity must be replenished, too.

One response to “The Flash Crash: A Cautionary Tale”

  1. I can’t help but note the parallels between the unintended consequences (one hopes) of faulty signaling in this instance and those in the recent episode at Rutgers. Had the W&R order hit an open outcry pit instead of pools of dark liquidity, the signals of its aberrance, both audible and visual, would have been instantly read, and the ensuing feedback loop would have self-corrected the misdirection in the market. Had the protagonist at Rutgers proposed the folly of exposing his roommate to a roomful of dorm mates, a single look of distaste would have overcome any sophomoric chittering in favor of the invasive act and one less tragedy would have occurred.

Leave a Reply

Your email address will not be published. Required fields are marked *