This post isn’t about e-discovery or computer forensics.  Not at all.  But as it’s about the near-fatal, self-inflicted wound Knight Capital suffered from a software snafu early on 8/1/12, it will touch on the immense power of technology

Doesn’t the whole fiasco bear an uncanny similarity to the old Matthew Broderick movie, “War Games,” where the NORAD W.O.P.R. computer thinks its being tested in a bracing game of “Global Thermonuclear War” but is actually connected to live warheads poised to annihilate hundreds of millions?  Indications are that Knight Capital’s shiny new software was running test trades in the 24 hours after its installation, but no one at Knight Capital realized that W.O.P.R. was actually executing those trades on the New York Stock Exchange!  Forty-five minutes and $440 million in losses later, the Big Board interceded, perhaps sparing us all from another financial meltdown.

Someone in IT needs to start packing up his or her Star Trek bobbleheads.

Yesterday, I took a position in the stock, moments before it was announced that a massive capital infusion from much larger investors would likely dilute my newly acquired interest by some 75%.  I blanched when I read the sweetheart terms extended to the “rescuers,” but fortunately my trade executed at the nadir of the issue’s precipitous decline and I managed to generate a 10% one-day return that (knock wood) hasn’t evaporated.  But this post isn’t about that either.  I mention it only so you can leaven anything I say here with the knowledge that I’m now a (very minor) Knight Capital shareholder.

I tend to be a black swan investor.  That is, I buy interests in otherwise solid companies when big, bad things happen, like BP,  Marsh & McLennan, Ford or Goldman Sachs.  I sometimes mistime the trough and ride the price down for a while or I come late to the party and miss some of the upside; but, so far, all my black swan investments have rebounded handily.  I was a plaintiffs’ personal injury attorney in a prior life; so, I’m well-practiced in profiting from the misfortunes of others by sharing their risk and investing in their recovery.

The scale of the Knight Capital loss is staggering; but, someone will always be in a position to precipitate unimaginable chaos.  It may be the guy who silences the alarms at the refinery or the cruise ship captain who gets too close to the reef.  The technical term for this is “shit happens.”

It brought to mind a case I settled in trial on a warm Spring day many years ago.  The facts were a Palsgraffian nightmare.   In the eighties, in the then-peaceful and prosperous city of Monterrey, Mexico, most of the water put into the distribution system never reached customers because the pipes were beset by leaks.  Imagine a big, modern city having water for only an hour or two a day!  The water and sewer department had but a single, primitive set of paper maps that poorly correlated with the locations of the pipes and mains.  When crews went out to fix leaks, they dug here and there, trying to find the pipes.  Water, it seems, not only seeks its own level, but seeks its own path, too.  

A Texas company that mapped water and sewer systems was hired to develop a geodetic database to help the city locate and repair its system.  For months, crews with special sensing equipment (and not-so-special shovels) walked the streets, dug holes and made extensive field notes.  When the work was done, the maps and the field notes were stored in a secure, windowless interior office on the fifth floor of an glitzy office building in west Houston, waiting to be digitized.

Two floors above, an oil company had an employee break room with a soda fountain–the sort where carbonated water is mixed with syrup to become Coke.

Just FYI: in Texas, all soft drinks except iced tea are called “Cokes,” so it’s not strange to hear a server ask, “What kinda Coke dy’all want,” and the sensible answer being , “I’ll have a Dr. Pepper and Bubba will have Big Red.”

Anyway, inside the soda fountain, was device called a carbonator.  It  consisted of a water tank resting on a spring assembly such that the tank would descend as it filled and trigger a level control micro switch that would, in turn, cut power to the pump and cease filling the tank.  It was all a bit Rube Goldbergian, but it worked.  That is, until one fateful Friday evening when the micro switch failed by arcing, sticking in the “on” position.  The pump pumped away and the tank finally overpressured, causing the relief valve to relieve and water to shoot out of the tank at high pressure…for hours and hours and hours.  The seventh floor flooded, then the sixth.  Finally, two floors of water crashed through the ceiling of the fifth floor.  And wouldn’t you know that, of all places, it played Niagara right into the windowless storeroom where all the original paper data was stored?

Copies?  We don’t need no stinking copies!  We got pulp!

And with that ridiculous turn of events, a company was decimated and the people of Monterrey were left more parched than previously.  Coming to a drive-in near you: The Micro Switch that Ate Monterrey!

It wasn’t very funny at the time, but it was all an indelible lesson in how great oaks of misfortune from little acorns of error grow.  As I said, stercus accidit.  Everything sounds better in Latin.

Trolling the many news reports and analyses of the Knight Capital self-immolation, I’m struck by how many see it as a further symptom of hollow markets, Wall Street corruption and the need for regulation.  While I share all the cynicism and disgust for Wall Street, the Knight Capital disaster is not a further cause or symptom of the putrid puss that passes for banking and brokerage nowadays.  The software error was a problem in a system, not a systemic problem.  It isn’t like the LIBOR scandal, where people knew it was wrong and acted out of greed or a misplaced sense that they were lying in support of a greater good.

The Knight Capital disaster was a terrible one-time mistake that might have been thwarted by better testing, better vigilance or better circuit breakers; but, maybe not.  In any event, no regulation will serve as a better deterrent than the $440 million haircut Knight visited upon itself and its hapless shareholders.

But that’s just how I see it, lacking expertise in finance, without firsthand knowledge of the Knight Capital failure and, above all, not seeking to influence your investment decisions.

Advertisements