A looter, yes. But the cause of the Flash Crash...Noooo
Yesterday's news was that the authorities caught a rogue trader named Navinder Sarao. Allegedly, Sarao, manipulated the futures market lower on the day of the "Flash Crash".
According to the Wall Street Journal (subscription required), Mr. Sarao used "a souped-up version of commercially available software to manipulate a stock-market index futures contract, laying the groundwork for the index's decline, in which hundreds of stocks momentarily lost nearly all their value." This has, understandably, triggered a flurry of comments. One of the best comments I read was from Bob Pisani, on CNBC, this morning. He made three critical points that deserve amplification:
- The strategy allegedly used by Mr. Sarao is manipulative, but not the cause. The sheer scope of the drop and the traded volume on May 6, 2010, however, identified him as a small contributor to the crash, not the primary cause.
- The ability to detect such behaviors clearly needs major overhaul; the fact that it took 5 years to build this case is troubling, to say the least. Many improvements to manipulation detection can be made, but a good starting point would be to expedite rulemaking to increase order routing disclosures in equities, ETFs, options and futures. For example, detecting such behavior would be much easier if our proposal for Rule 606 reform was implemented and extended to prop trading firms, as well as the futures market.
- The markets vulnerability to the Flash Crash was, in major part, due to insufficient and poorly coordinated "circuit breakers" in both equity and futures markets. This is actually good news as the equity market adoption of Limit up / Limit down (LuLd) rules provides substantial protection against a recurrence. That said, cross asset class coordination could certainly be improved including rules for both futures and options in addition to ETFs and individual equities.
The first point is that people should calm down and not blame the flash crash on this one trader. Let's say the allegations are true and the "layering" of the book, by Mr. Sarao, likely exacerbated the crash; it certainly wasn't the cause. You can think of him as a "looter" during a blackout; he contributed to the problem and deserves punishment, but Mr. Sarao could not have caused such a large move on his own. The numbers simply don't add up. As Bob Pisani pointed out, the "$170 million-$200 million in sell orders that Sarao allegedly put in seem like small potatoes" compared to the $4.1 billion futures order that the SEC blamed. In addition, I would, actually, be very surprised if there were not more "looters" that day, since many proprietary strategies trade more during periods of excess volatility. While most of these trades are not manipulative, it is important for the confidence of market participants to identify and punish those that are consciously manipulating the market.
The key to stopping such intentionally misleading trading behaviors, is to have both clear rules and proactive enforcement. Rules such as those against "spoofing" and "layering," as well as future rules against "momentum ignition" strategies, which were targeted by Chairman White as a key SEC priority, need to be clearly defined with appropriate enforcement tools. Improvement to both enforcement and rulemaking is essential because manipulative activity does, in fact, magnify volatility. That said, it is important to keep in mind that these behaviors could not cause a crash of this magnitude, without both major uncertainty and very large "triggering" trades. (I will not reprise all of the factors at play on May 6th, as they have been covered, extensively).
The second point is that enforcement actions are taking much too long. While enforcement actions are not easy, there are several rules that could help these efforts, including our proposals on rule 606. We have advocated for comprehensive routing disclosures, on a monthly basis, to help restore investor confidence, but this rule could also help flag potential manipulation. If such disclosures were required by proprietary equity traders as well as for traders in the futures markets, it could provide a database that could be used for early warnings.
The infrastructure required in order to calculate our proposed disclosures would categorize trades as non-marketable, whether they were at the NBBO, near the NBBO, or away from the NBBO, and then calculate the number of orders, share values, duration of orders, and whether or not they were executed. These calculations could be used, in turn, as the basis for statistical filters to identify potential manipulation. If, for example, a report measured routing of non-marketable, near the NBBO, unexecuted orders and it showed a large jump in those orders, relative to traded volumes, it would flag that trading strategy as suspicious for subsequent detailed analysis. This type of statistical reporting is one of the best ways to manage the enormous amount of trading data that is created every day. By focusing analysis on specific strategies during time periods identified by the report as being potentially manipulative, it would streamline surveillance processes. This would enable better internal trading reviews by the compliance departments of trading firms and the regulators could use it as a tool for targeting their examinations.
The last point, and an important one, is that systematic vulnerability in the markets must continue to be addressed. As a practical matter, we should all be supportive of the SEC rule making that created LuLd, but should call for extensions to the rule. There are improvements to be made, particularly for ETFs, which have limits that have no relation to their NAV in the rule, but generally LuLd is a good rule. Despite flaws, the existence of LuLd makes a full recurrence of May 6 far less likely for equities. That said, LuLd should be extended to other asset classes, including options and futures. In our current regulatory structure, that might be problematic since it will require much better coordination between the SEC and CFTC. The goal would be for related markets, such as equity index futures and underlying equities, to have similar rules, so trading halts could be synchronized. I believe that most participants would agree that it is better for markets to stop trading for short periods of time, rather than be allowed to move in an uncontrolled manner.
David Weisberger, Managing Director, Trading Services at Markit
Posted 22 April 2015
S&P Global provides industry-leading data, software and technology platforms and managed services to tackle some of the most difficult challenges in financial markets. We help our customers better understand complicated markets, reduce risk, operate more efficiently and comply with financial regulation.
This article was published by S&P Global Market Intelligence and not by S&P Global Ratings, which is a separately managed division of S&P Global.