I spoke recently with James Angel about the high-tech nature of today’s markets, the glitches that result, and the conclusions we should (or shouldn’t) draw from same. He offered me a compare-and-contrast exercise involving three distinct market crises: one in 1987, with technology that now seems ancient, one in 2010, and the most recent on August 1, 2012. The lesson: plus ça change, plus c’est la même chose.
Angel has his finance Ph.D. from the University of California at Berkeley. He consults with broker-dealers and law firms on the structural features of the financial markets, is on the board of directors of Direct Edge, and currently teaches at Wharton School, University of Pennsylvania.
I mentioned to him that one of the issues that concern some people about the rise of the high-frequency traders is that there is a somewhat nostalgic idea about the functions the old (human) market makers used to serve, and the robot-trades are thought especially dangerous given the impression that the functions of those market makers are left unperformed.
Reconsider the Good Old Days
As the Securities and Exchange Commission and the Commodity Futures Trading Commission put it in their joint report about the May 2010 flash crash, in an environment “where high frequency and algorithmic trading predominate and where exchange competition has essentially eliminated rule-based market maker obligations, liquidity problems are an inherent difficulty that must be addressed.”
Angel said that the equity markets as of 2012 “make up a complex technical system.” There are rough edges to this system, and “we have to be aware of that.” But he proceeded to pour some cold water on the flames of nostalgia.
“Even back in the days of physical market makers, when things went bad, as for example in the crash of ’87, the market makers would head for the hills,” he said. That point was one of the themes of the Brady Commission report in 1988. That commission, a presidential task force headed as its name implies by Nicholas Brady, who would become Treasury Secretary later that year, said that while no default occurred in the ’87 crash, “the possibility that a clearing-house or a major investment banking firm might default, or that the banking system would deny required liquidity to the market participants, resulted in certain market makers curtailing their activities and increased investor uncertainty.”
What (or who) is a market maker. The dictionary definition is one who stands ready to buy or sell at publicly quoted prices, and in round lots. The head-for-the-hills notion seems to contradict the ‘stand ready’ element in that definition. But they refer to different situations.
In a market not experiencing any crisis, it will nevertheless often be the case that Joe Smith’s “buy” order and Jane Doe’s “sell” order won’t come in to any common exchange or trading platform at precisely the same time. As Angel puts it, “We can either wait until they are both available and matched, or we can pay someone for immediacy. Providing this immediacy is a valuable service, but market makers are not and have never been a price or crash insurance system.”
Turn the Computer Off or Leave it On
Let’s get back to 2010, the Flash Crash. Did this come about (or was it exacerbated) either because there are no market makers? or because they didn’t perform properly? Those two questions have one answer: no. As the SEC/CFTC report quoted above also notes, there are market makers, even if they have changed form. High frequency traders themselves “often engage in multi-market arbitrage activities that essentially result in liquidity provision to and across markets.”
Later in the same paragraph, that report observes that many of these traders “chose to withdraw on May 6 as a reaction to the level of uncertainty.” Was that wrong of them? Something they should be discouraged or prohibited from doing hereafter?
Angel contends that they got it right. “What really happened is that the market was overwhelmed by a tsunami of message traffic, feeding into well-documented technical issues. Market data got scrambled. The market makers didn’t believe what they were seeing on their computer screens, so they shut the screens down, they stopped trading.”
They can hardly be faulted for doing so. Angel contrasted this favorably with the most recent crisis we discussed, which he called the Knight-mare, the August 1, 2012 disruption in the trading of equity in 140 companies. In this case, the data available on the computer screens was again scrambled, but “the market makers didn’t turn off their computers, and continued to churn out erroneous trades. The Knight-mare resulted.”
None of this is to suggest that everything is for the best, or even that the proverbial glass is half full rather than half empty. But discussion of what is really wrong with the way markets are made today, or of what is really worrisome about HFT, needs in either case a more precise focus.
“What we need,” Angel said, “is a fail-safe or fail-graceful system, one that recognizes that glitches will happen and that can contain the damage they can do through appropriate circuit breakers and shock absorbers.”