How artificial intelligence and robotrading pose a growing threat to the global marketplace.
By Scott Thill, AlterNet
Posted on May 27, 2010
"We have found no evidence that these events were triggered by 'fat finger' errors, computer hacking, or terrorist activity, although we cannot completely rule out these possibilities," a recent Securities Exchange Commission (SEC) report on the so-called May 6 "Flash Crash" that wiped out a cool trillion in a mere half-hour weakly admitted. "Much work is needed to determine all of the causes of the market disruption."
That's another way of saying that it remains only the market makers that caused the largest single-day point decline in Dow Jones history who actually know where the bodies are buried. The rest of us, including the SEC, have a Sisyphean task of sifting through mountains of dense data. But regardless of who ends up on the end of possible criminal proceedings, the SEC is sure that the whole clusterstock was seriously exacerbated by the robotraders executing light-speed electronic transactions via supercomputers, while exposing our hyperreal economy as an Internetworked casino. If anything, the Flash Crash proved that market makers like Goldman Sachs and plenty more playing both sides of securities could be capable -- with the high-priced help of math and computer science Ph.Ds crafting up proprietary, recursive algorithms -- of wiping out any corporation's stock, perhaps any nation's economy, in a comparative instant with just the press of a button.
"It was actually amazing watching it all happen," Gina Sanchez, senior analyst for Roubini Global Economics, told AlterNet by phone. "We went from risk-aversion to risk-seeking in the matter of an hour. But it doesn't bother me so much that the algorithms went after the bids. They were doing what they're supposed to do, which is seek out arbitrage opportunities. What concerned me was how the bids got out there in the first place."
That is the primary concern of the SEC as well. But it's going to have a hell of a hard time figuring out the human brains behind the inhuman bids that remotely reduced the price of some once-reliable stocks to mere pennies. Thanks to the very technological innovations that has transformed last century's stock market into an inscrutable hyperreality programmed and deprogrammed daily by rapacious banksters, Wall Street's corruption cops are drowning in deeper paperwork, virtual and otherwise, than ever before.
"Although developments in the markets and in technology may help speed access to market data, they also greatly complicate our efforts to analyze the complex web of trading arrangements and market dynamics that have developed since 1987," SEC chairman Mary Schapiro confessed in a May 20 Senate hearing on the Flash Crash. "For example, the key day in the 1987 Market Break Study involved a trading session processing a little over 600 million shares in NYSE stocks. On May 6, the markets processed 10.3 billion shares in NYSE stocks alone."
What Schapiro and the SEC really need to crunch all that digital data are some genius math and science nerds, but they've all been conscripted by the market makers to game the global economy. It pays stunningly well, even when it epically fails. After engineering a global crisis that has so far swelled America's national debt to $13 trillion, the too-big-to-fail banksters are now bigger, stronger and armed with bonuses for their predatory efforts. In fact, that success has emboldened them to continue their intrepid devaluations. Since the Flash Crash, the overall market has experienced severe devaluations marking off technical corrections that place it out of the reach of economic recovery. Even in depressed economies, market makers make out just fine.
"We don't really have a market as such anymore," Marshall Auerback, senior fellow at the Roosevelt Institute, explained to AlterNet. "It's a high-tech casino. And that is actually an insult to casinos, which are probably better regulated and which observe rules that make the house pay when it loses. Unlike our Frankenstein investment banks, which get government handouts."
For those interested in the discrete details of the Flash Crash, the SEC's report (PDF) is a labyrinthine walk down a dark alley populated by drowsy terminology, graphs, deconstructions and worse. With unemployment still strong, industries still cratering and lending still conventionally disappeared, working people shouldn't have to give their valuable time to it, which is, of course, how the banksters win in the end. Especially since they pay more than the SEC ever will to sift through its shifty legerdemain. But the short version is simply a high-tech, high-speed variation on the type of confidence games that have been around forever. The only difference now is that they can be gamed in eyeblinks by machines whose capacity for pattern recognition and algorithmic trading is faster than ours. The shorter version is that they have no mercy, and there's nothing us puny humans can do about it.
"Artificial intelligence is becoming so deeply integrated into our economic ecostructure that some day computers will exceed human intelligence," American inventor, futurist and hedge funder Raymond Kurzweil explained in a speech to investors and other quantitative speculators in 2006. "Machines can observe billions of market transactions to see patterns we could never see."
What those machines saw during the Flash Crash, according to the SEC, were ridiculously priced bids valuing reportedly healthy companies at next to nothing. Since then, we need, the SEC report explained, "to examine the use of 'stub quotes,' which are designed to technically meet a requirement to provide a 'two sided quote' but are at such low or high prices that they are not intended to be executed." Even so, the algorithms did their job, as Sanchez explained, and processed those quotes without what would constitute a rational human pause. That lead directly to what Schapiro called "the absurd result of valuable stocks being executed for a penny" in her Senate testimony, and further evidence that when algorithms attack, they do so according to their programming, even it doesn't make a lick of human sense.
Of course, there's the alternate theory, which is much less labyrinthine: They were programmed specifically to make no conventional economic sense, just literal cents. Exacerbated by what the SEC report called "the withdrawal of liquidity by electronic market makers," it's pretty reasonable to assume that the Flash Crash's robotraders attacked on purpose, searching and destroying their targets with callous accuracy, while the banks and traders supposedly committed to a "best execution" duty disappeared and let the machines tear up the trading day. Some lost out, as thousands of trades were canceled. Others meanwhile, scored huge.
For example, whomever programmed their computers to buy up shares of world-beating Apple Inc. at $199 locked in a price that hadn't been glimpsed by the iPhone and iPad king in months. Better yet, they scored an immediate 20 percent payout for that timely investment in a matter of mere minutes, as Apple's more conventional stock price returned to around $250 that same day. No trader with an ounce of sense would ever expect to nail Apple at $199, given its recent performance. But an enterprising investor placing a hopeful stub quote on a nonplussed computer at a laughable $199 could get lucky, provided he or she is supplied with the appropriately timed left-field Flash Crash. The computer, certainly, isn't going to stop Apple's financially insane stock fall or subsequent rise from happening, if it's expected to profit from it.
But the Flash Crash is the new normal for our economy, in which unregulated stratagems like credit default swaps (valued at least at $36 trillion and counting), naked shorts, high-frequency trading, dark pools and more deliberately inscrutable mechanisms all suck up taxpayer and investor money without ever creating anything of actual value for most of us. In fact, the mounting evidence -- despite all of the digital front-running strategies, securitized derivatives webworks or members-only traders' clubs -- is spectacularly clear: These technological and scientific stock-market innovations have chiefly resulted in sociopolitical and economic misery for the world-at-large.
"These kinds of products serve no social purpose whatsoever," Auerback told AlterNet. "And they have come to excessively dominate trading. We seem to confuse means with ends."
Sure, if one is expecting the worse of us to care about the distinction. But in our new millennium, humans and computers alike are executing gamed trades using next-generation technology and algorithms whose morality is colder than math itself. With them, there is no distinction between means and ends, nor is there supposed to be. All that matters is the payday, in the end. Everything else is collateral damage.
And we should have seen it coming, because this has often been the case with technological and sociopolitical evolution. With each new innovation, we have found even newer ways to exploit our surroundings and each other. We took weapons designed to keep us well-fed, and used them to kill each other to the point of genocide. Our excessive fossil fuel consumption -- which has increased our mobility, convenience and connectivity -- has also led us to a climate crisis that could conceivably leave most of our planet uninhabitable within a couple cosmologically short centuries. And especially lately, we have used our economic, mathematical and technological ingenuity to find ways not to make us all richer, but to wipe out trillions in wealth in the relative blink of an eye. It's no wonder our algorithms are assholes.
No comments:
Post a Comment