Martial Finance: The Case of High Frequency Trading
Is high finance becoming a game of battle-bots? We are all familiar with financial markets’ intense reliance on technology. But in the case of high-frequency trading, technology appears to be driving, rather than serving, trading strategies. The resulting uncertainty and instability has disturbed many leading thinkers and policymakers. Ordinary regulation does not appear capable of deterring (or even detecting) dangerous activity here. Therefore, after briefly describing some recent developments, I want to discuss how the recent fusion of law enforcement and national security forces in the anti-terror context might need to spread to the financial realm.
In the article “When The Speed Of Light Is Too Slow: Trading at the Edge,” Thomas McCabe discusses the kinds of price differences HFT’ers spot and profit from:
Computers were originally introduced in trading because they are faster than us in responding to market signals. A human trader might buy up a million shares of Microsoft for $20 a share, and sell them the next day for $21, making a million dollars in profit. However, if the price of a stock is $15.67 in New York and $15.68 in London one moment, but jumps to $15.70 and then $15.69 a tenth of a second later, no human could react quickly enough to buy the stock in New York and sell it in London before the prices reversed. To solve this problem, traders over the last few years have been building automated high-frequency trading (HFT) systems that compete by making thousands of trades a minute to maximize profit.
Lately, as McCabe notes, the limiting factor in fast trading is not computing power, but communication power. Thus firms are paying to construct ultra-fast cables (not for use by the public) between financial centers. For example, a “Chicago-New York cable will shave about 3 milliseconds off . . . communication time.”
Recently, modelers have devised more extreme solutions to the time delay problem. McCabe describes the proposals of Wisner-Gross & Freer to locate computers at “optimal locations from which to coordinate the statistical arbitrage.” An “optimal scheme” would “push trading firms to build new computers [at] the exact, optimal points in between markets”—even if that happened to be in the middle of an ocean.
Many commentators have blamed automated trading for the flash crash of May (and the several “mini-flash crashes” that have followed this year). These are valuable inquiries, and I hope SEC investigations in the area result in monitoring of both automated and high-frequency trading (HFT).
But before the agency gets too deep into the technical details of the trades, perhaps we should think a bit about the ultimate social purpose of HFT. If it’s good for traders to reduce inter-regional differences in prices in an hour, is it better to do so in 30 minutes? One minute? One millisecond? Are there diminishing returns for the strategy? And might the effort to encourage rapid arbitrage ultimately encourage the very distortions it was meant to reduce (just as tranched CDOs, advertised as a new risk management technology, ended up exacerbating and spreading the very risks they were supposed to ameliorate and contain)?
In a recent article, Joseph Fuller discusses how rapid, automated trading can cause self-reinforcing volatility:
[In 1998, LTCM’s flight-to-liquidity led to a] chain reaction [that] doomed the company itself and shook the capital markets. When credit markets began to seize up in mid-2008 and the securities markets went into free fall, the models tried to figure out a suitable response. They had been programmed to avoid volatility by moving out of securities and into cash. Of course, when many models trading hundreds of millions of shares all tried to liquidate investments and move into cash, they only increased the stock declines, leading to further volatility and thus to more selling.
He says that “computer models have three inherent problems:”
The first problem is that those who created the models don’t understand the markets. Modelers are experts in math, computer science, or physics. They are not generally experts in stocks, bonds, markets, or psychology. Modelers like to think of markets as efficient abstractions, but these abstractions can never fully account for the messy and irrational actions that humans take for emotional reasons. . . .
The second problem is that managers don’t understand the modelers. Most of the current generation of senior executives on Wall Street lack the technical background to understand the models (or the algorithms that underlie them) that power their own firms’ trading strategies. . . .
The third problem is that the models don’t “understand” each other. Each model executes its own strategy based on its calculus for maximizing value in a given market. But individual models are not able to take into account the role other models play in driving the markets. . . .
HFT systems may destabilize economies, but HFT firms profit more from volatility than stability. Imagine the arbitrage opportunities that arise if the price of a stock falls by 50% and then suddenly rebounds. Fuller’s third point, that “models don’t ‘understand’ each other,” creates incentives for writing programs that can anticipate what average programs expect other programs to do. At this point, the finance system approaches an asymptote of complete self-referentiality, unmoored from any foundation of estimates of the long-term value of enterprises by the millisecond-by-millisecond demands of a turbocharged trading environment. We need a fundamental rethinking of the purpose of speed in financial transactions. But until that happens, we need to defend ourselves from the types of sudden changes in valuation that can massively disrupt our daily lives.
Back in 1993, New York Times editor (and economist) Joel Kurtzman warned about these problems in a prescient book entitled The Death of Money.* Kurtzman endorsed many proposals to dampen the volatility arising out of computerized trading, including a financial transactions tax. Unfortunately, none of his ideas influenced a policy space besotted with theoretical “liquidity” (however quickly it dries up in the very crises it’s supposed to help solve). We have now reached a point where “70% of trading volume on the major exchanges is conducted by high-frequency traders who hold a stock for an average of 11 seconds,” and even free-market mavens like the Economist magazine have expressed concerns about the social utility of ever-faster trading.
Both the CFTC and SEC are interested in better monitoring EFT, but may not get a chance to do so. Big names in HFT are using their financial gains to influence Washington regulators. As Bloomberg News reports,
In addition to writing proposed rules, the SEC’s enforcement division is investigating whether computer-driven traders have manipulated prices. “You have to be concerned every time there’s a lack of transparency into a market practice, particularly one like high-frequency trading that is so prevalent,” Robert Khuzami, the SEC’s enforcement chief, said in an interview.
The scrutiny has spurred the industry to seek friends in Washington. . . [The] companies . . . have more than quadrupled their political giving over the last four years. . . .Last January, Representative Cantor of Virginia, the second highest-ranking Republican in the House, sent [SEC Chair Mary] Schapiro a letter saying her agency’s ideas for regulating fragmented, electronic markets, including a proposal that would prohibit exchanges from giving high-frequency traders and other market participants a split-second peek at stock orders, “appeared ad hoc in nature.”
Hmm . . . powerful men tell a careful and deliberate financial regulator that she just doesn’t appear to know what she’s doing . . . where have we heard that before? Cantor says that the SEC should “collect all the facts and develop coherent and rational policy objectives before adopting any potentially far-reaching rulemaking proposals,” but I’d bet dollars to donuts he’ll be supporting his colleague Spencer Bachus’s apparent ambitions to defund entities created under Dodd-Frank, including the Office of Financial Research. The OFR is one of the few entities actually empowered to accomplish what Cantor is calling on Schapiro to do.
I can understand Cantor’s idees-fixes here, and certainly Dodd-Frank as a whole may have some utterly perverse unintended consequences. But have any of the deregulators stopped to think about the national security consequences of HFT-associated volatility, or disparate computing and communication power in finance generally? Instability in US financial markets might drag down the dollar, which in turn could make oil and other basic commodities extremely expensive. The Pentagon has “war-gamed” financial instability, and participants expressed concerns:
Several . . . said the event had been in the planning stages well before the stock market crash of September , but the real-world market calamity was on the minds of many in the room. “It loomed large over what everybody was doing,” said [Yale professor Paul] Bracken. “Why would the military care about global capital flows at all?” asked another person who was there. “Because as the global financial crisis plays out, there could be real world consequences, including failed states. We’ve already seen riots in the United Kingdom and the Balkans.”
Some might be uncomfortable with a blurring of the boundaries between economic regulation and national security. However, as Danielle Keats Citron and I document in our recent work Network Accountability for the Domestic Intelligence Apparatus, that distinction is already long gone in many other areas. We aren’t entirely comfortable with the blurring of law enforcement and domestic intelligence. But if the U.S. can deploy agents to monitor honey-baked ham stores, we certainly should be thinking about much larger threats to the nation’s economic infrastructure. As the financial economy continues to grow and dwarf the real economy in size, it becomes a source not merely of “systemic risk,” but of far more profound disruptions.
In a world of unmonitored, ever more rapid capital flows, concentrated wealth can purchase power in a way that fundamentally threatens the state’s monopoly on the legitmate use of force. However, the very monitorability of electronic data flows could also help smart states avoid that scenario —and the varied lesser challenges to state authority that lead, on average, to over $1 trillion of illicit financial flows each year. Britain is taking some good steps in the right direction, but far more needs to be done. For example, if there is any communications channel deserving of “warrantless wiretapping,” it is the high speed trading cables described here.
As Philip Mirowski demonstrated, military thought has thoroughly shaped modern economics. Financial wars, both real and metaphorical, have grabbed headlines recently. It is now time for the US intelligence apparatus to shift more attention to the financial flows that could cripple the nation’s ability to maintain order in times of crisis.
*Admittedly, Kurtzman appears to be reconsidering his ideas currently.