If you've watched any of the Terminator movies, you know that computers are just waiting to attack the human race. While we embrace technology's role as an integral part of our daily lives, computers are sitting back, waiting for SkyNet to go online and usher in their reign.
Of course, I'm joking a bit. But, in the wake of the past month's flash crash and one frightening, but mostly unpublicized event since then, it's hard to ignore the possibility that computers have been trying to destroy us lately.
What is going on?
It's been suggested that last month's lightning crash in the market, which sent the stocks of several unrelated companies including Procter & Gamble (NYSE: PG ) , Boston Beer (NYSE: SAM ) , and Accenture (NYSE: ACN ) plummeting, began as a result of a "fat finger" trade, a physical error on the part of a single trader. But even after weeks of investigation, the authorities still aren't sure what caused the flash crash.
We can definitely confirm today, however, that the crash wouldn't have much of a crash if it weren't for the thousands of trades automatically triggered by computerized, high-frequency trading platforms working across the entire system. The cascading momentum of so many unmonitored pieces of semiconscious software codes kicking into effect -- not a stupid trader or a blown order -- made the crash a real crash.
SEC Chair Mary Schapiro explains: "...the interconnections among markets and among equity securities and derivatives have grown immensely more complex over the past few years. Orders in one stock directed to one market can now ricochet to other markets and trigger algorithmic executions in other stocks and derivatives in milliseconds."
In the last 20 years, the New York Stock Exchange has gone from processing a few hundred million trades per day to many, many billion, all thanks to the interconnected strength of advanced technology. The trading world has undergone extraordinary evolution. But is that a good thing?
Beware the Black Swan
Ninety-nine days out of 100, the widespread use of advanced technology isn't a problem. The pieces of code that dictate the actions of high-frequency trading platforms are configured well enough to handle normal gyrations in the market. In fact, we're unlikely to even notice their existence. But that one seriously abnormal day, well, it tends to cause big problems. Readers of Nassim Nicholas Taleb's The Black Swan will understand this idea well. Computers work effectively ... until they don't. Then things get ugly.
Today's Wall Street
Modern trading institutions leave much of a firm's overall activity to automatically calibrated algorithms like these. Consider one recent Wall Street Journal article, which discussed high-frequency traders that scrape information from the market "100-200" milliseconds faster than the competition. According to the article, high-frequency trading of this type now "account(s) for about two-thirds of U.S. stock market volume." Man alive!
Assuming these programs are properly written, investment houses can generate incredible trading profits in mere seconds. But are these programs properly written? If the flash crash were merely an anomaly, what explains the near flash event that occurred on the Osaka Stock Exchange just last week?
Land of the rising HOLY SMOKES!
On June 3, the Osaka Stock Exchange quietly announced that Deutsche Bank's (NYSE: DB ) computers had mistakenly ordered approximately 7,500 Japanese futures contracts worth a staggering $182 billion. Fortunately, only 0.3% were executed before the mistake was fortunately caught (yes, only about $550 million worth). Imagine the chaos that might have ensued, should that order have been filled in its entirety. Depending on the pricing of these orders, the market could have soared out of control or plunged into new depths -- all without any kind of legitimate justification.
How many other near-events have occurred as a result of computer error without anyone knowing about it? How many times have we gotten this close to outright disaster?
Here's the problem
Software and technology, like human beings, are fallible -- but in much, much different ways. High-tech tools can marginalize the notorious emotional errors that we bring to the table, and work at speeds that are fully incomprehensible. But, at the same time, they very clearly lack the advanced reasoning that we bring to the table (at least, for now).
Unlike a human, a computer cannot take a step back when a $180 billion futures order comes across the screen, say, "Something doesn't seem right about this," and cancel the order, unless it is explicitly instructed to do so beforehand. Pending future developments, software is also written by humans, meaning it's bound to suffer from at least some of our more basic intellectual shortcomings.
Why I'm scared
We are exposing ourselves to problems associated with human-less decisions, at rates that we may not be able to appreciate. Because humans don't have the physical capacity to monitor high-frequency trading platforms on a real-time basis, I'd argue that we are not playing a substantively active role in our own stock market -- at least, not during the hours between 9:30am and 4pm. This is a terrifying realization.
Mary Schapiro explains once again: "One of the challenges we face in recreating the events of May 6 is the reality that the technologies used for market oversight and surveillance have not kept pace with the technology and trading patterns of the rapidly evolving and expanding securities markets."
The natural result of this situation is a world far more prone to the type of Black Swan, extreme-volatility events that we saw over the last month.
The Foolish bottom line
We should probably relinquish the idea that we still have control over the same entities that hold our financial future. We can still play the game, for sure, but it does not belong to us anymore. The game belongs to the computers. Whatever mistakes are made when creating these tools will likely rear their ugly head, on a scale and at a time that we least expect. We're just passengers now, along for the ride.
More than anything, it's critical that our governing bodies understand the role that these technologies play in our larger marketplace, and determine whether role is fully appropriate given today's economic environment. Mary Schapiro, for one, appears to see the light. I'm waiting for others to be similarly awakened.