“So if I know about a stock’s activity one day before it’s insider trading, but if I know about a stock’s activity one second before it’s high frequency trading?”
-Samantha Bee, The Daily Show
Ms. Bee doesn’t have it quite right, but mainstream comedy shows rarely feature such topics. The term high frequency trading (HFT) encompasses various trading methods that employ sophisticated algorithms and powerful computer hardware designed for speedy trading. Wall Street banks and investment institutions spend billions on HFT. HFT algorithms can read market data and implement market-wide strategies in milliseconds. As the name suggests, high frequency traders aren’t buy-and-hold investors. HFT profits are aggregates of small profit margins on enormous trading volume. In fact, HFT is thought to be responsible for huge increases in trading volume in recent years. Estimates vary, but algorithmic HFT accounts for as much as 70% of daily equities trading volume in all markets. The enormous profit potential of HFT has fueled competition for talent and a technology arms race. Many institutions pay exchanges monthly rental fees to “co-locate” their hardware on the exchange premises, bypassing order routing infrastructure and improving latency to get ahead of other traders.
What strategies do high frequency traders use? The answer is largely guesswork, as proprietary trading algorithms are closely-guarded secrets. Among other practices, HFT facilitates “iceberging,” or disguising large orders by parceling them out into numerous smaller orders. But critics allege that HFT is used in connection with unfair or illegal practices like flash trading, front-running (placing bets in the market based on pending client orders) and bullying other market participants into disclosing price limits and giving up profits.
In a climate of heightened public scrutiny of Wall Street, HFT made waves in the mainstream media this summer. In late July, Senator Chuck Schumer urged the SEC to address flash trading, an instance of HFT where some exchange participants view and trade on price quotes immediately before they are publicly visible. When an exchange first receives a buy/sell order, it determines whether any exchange participant has publicly displayed interest in the desired security at the stated price. If not, Rule 602 of SEC Regulation NMS requires the exchange to include the order in public consolidated quotation data, routing the order to other exchanges. However, an exception to Rule 602 allows exchanges to first “flash” the order to paying exchange participants as little as three-hundredths of a second before routing. Those exchange participants with HFT capabilities are fully capable of acting on such a short timeframe.
Why would the SEC create an exception for flash trading in the first place? When Rule 602 was enacted in 1978, it apparently didn’t consider the possibility that technological developments would eventually enable automated traders to capitalize on the “flash,” rendering an administrative convenience exception into a substantial loophole.
Prompted by Senator Schumer, the SEC issued a proposed rule in September to ban flash trading in all markets, concluding that the flash trading exception is “no longer necessary or appropriate in today’s highly automated trading environment.” The Commission found that flash trading promotes a two-tiered market where material price data is available to some traders before the general trading public. Because flash trading creates disincentives for exchange participants to publicly display interest in trading (when they can instead act in secret on flash data), the proposed rule also voiced concerns about transparency, market efficiency, and the need for publicly displayed liquidity. In addition, under the Securities Exchange Act, the SEC is required to consider whether market practices damage public confidence in the fairness of securities markets. The SEC found that long-term investors might consider flash trading an unfair practice that damages the integrity of the market. To the extent that the interests of short- and long-term investors conflict, the SEC has a “clear responsibility” to protect long-term investors.
Goldman Sachs submitted comprehensive comments to the SEC concluding that flashes should be subject to the same regulations as standards quotes. In other words, co-located high frequency traders should no longer have a head start. Of course, HFT maintains a distinct edge against less technologically sophisticated competition. To allay this concern, Goldman points out that technological developments like HFT have lowered bid-ask spreads, increased market efficiency, injected considerable liquidity into securities markets, and increased accessibility to markets via automated market makers, “primarily benefitting retail investors.” In broad strokes, Goldman cautions against throwing out the HFT technology baby with the bathwater, instead proposing better regulation at the margins, where technology has outpaced SEC oversight. Flash trading lies at the margin and is per se unfair.
But is Goldman correct that the benefits of HFT outweigh the costs? The question is difficult to answer, primarily due to a lack of empirical data.
Critics contend that HFT promotes a two-tier system – HFT-armed institutions in the first tier, retail and individual investors in the other. To some, HFT is a “sophisticated bid-rigging scheme” that amounts to a multi-billion dollar tax on unsophisticated investors, redistributing wealth to large banks with institutional savvy and the resources needed to exploit technology-capability disparities and loopholes in SEC rules. Paul Krugman came out guns blazing, arguing that HFT also fosters excessive speculation and concluding unequivocally that “what [high frequency traders] do is bad for America.” Viewing HFT through the lens of social utility, Krugman cites scholarship showing that speculation based on private information “combines private profitability with social uselessness,” wastes resources, and undermines market integrity. To the extent that HFT represents a sub-optimal level of speculation, it falls on the negative side of the ledger. The problem: we don’t have enough data on HFT to quantify the scope of these problems.
What about liquidity and pricing? There’s no denying that HFT provides liquidity, and lots of it. Liquidity – the ability to safely buy and sell an asset – is so important in financial markets that many exchanges offer rebates to liquidity providers like high frequency traders. Some markets might not even exist without HFT. Critics respond that HFT injects unhealthy, destabilizing liquidity into the financial system. HFT is also said to contribute to price discovery. One rebuttal to this claim is the argument that HFT algorithms primarily trade against each other in a game of speculation, as opposed to value investing that contributes to proper pricing of securities. Again, limited available data makes it difficult to draw solid macroeconomic conclusions on these issues.
Some have suggested that HFT poses a systemic risk to financial markets and the broader economy. An HFT algorithm with free reign could transform into a “rogue algorithm” that inflicts extreme market fluctuations. Extraordinary events could trigger a chain-reaction of massive unloading by HFT algorithms, a not-so-far-fetched scenario given one likely cause of the 1987 stock market crash. HFT may have contributed to excessive volatility during the oil spike of 2008 and the recent market-wide collapse. The presence of substantial systemic risk may delineate a line of optimal technological innovation in the financial markets. A public policy discussion of HFT should at least consider the possibility that we have crossed this line.
Proposed approaches to the HFT problem range from laissez-faire attitudes, an armistice in the technology arms-race, investing in new technology (e.g., anti-HFT algorithms), fiscal policy measures (e.g., a tax on all trading, aiming at HFT speculation), and, most helpfully, enhanced regulatory oversight. The necessary starting point for more regulation, however, is properly measuring and quantifying the effects of the HFT “problem.” This means more transparency, and the SEC’s approach, focusing on greater disclosure and reporting requirements, is right on track.
First, on October 21 the SEC voted to draft a proposed rule concerning dark pools, largely unregulated private exchanges that implicate many of the same issues as HFT. Second, the SEC is planning a reporting system for HFT firms. After the 1987 market crash, Congress enacted the Market Reform Act of 1990, amending the Securities Exchange Act. Under Section 13(h), the SEC may require persons meeting the Act’s definition of “large trader” to file with the SEC and self-identify when they trade. The prospective reporting system will gather information to enable the SEC to determine the impact of HFT on the marketplace, with special interest in the possibility of harm to long-term investors, disparities of market access and speed, and market efficiency.
At a recent Senate committee hearing on HFT, flash trading, and dark pools, Senator Edward Kaufman articulated widespread concerns about systemic risk and a two-tiered “trader’s market” where retail investors are at a substantial disadvantage. The CFTC has also expressed concerns on the impact of HFT in the energy futures market. Coordinated regulatory action – and possibly new legislation – will be necessary to the extent that loopholes, concessions and limited statutory authority restrict the scope of SEC regulation. A comprehensive effort to achieve transparency is essential to understanding the full impact of HFT technology in the financial markets.
Given investor reliance on highly beneficial liquidity, it’s unlikely that HFT is on its way out. But valid concerns about the impact of HFT have been raised, and regulators need to learn more about what they’re dealing with. In light of recent market failures and inadequate regulatory oversight of financial markets, it would be unwise to give the market free reign on HFT technology.
Trackbacks/Pingbacks