The May 6 market disruption shows ETFs need a more robust market-making mechanism.
Assigning blame for May 6's "Flash Crash" will take considerable resources and many months, and the conclusion is unlikely to single out a primary culprit. A probable conclusion is that the momentary lapse in the market's collective reason is a symptom of a systematic gap to accommodate the increasingly complex nature of securities markets.
Many predicted years ago that relying almost exclusively on computers, rather than humans in a trading pit, would lead to eventual market meltdowns. The human element of sound reasoning and seat-of-the-pants judgment would be necessary to facilitate a functioning market in periods of great duress. During these critical moments, it appears the critics were right. Simply transitioning from a computer-based system to the "slowdown" mode, in which humans are more involved, resulted in a momentary but monumental collapse on May 6.
But before we assess the shortcomings of the current system, let's focus on the few good points.
First, while the market crash seemingly lasted hours for those of us who were following the event moment by moment, a great majority of securities were trading at reasonable prices within 20 minutes of the onset. There was an eventual return to orderly conduct, which means those of us who did not trade during the period were not immediately impacted. Furthermore, two major exchanges acknowledged that several trades were executed so far beyond the realm of reason that they would take the unprecedented step of reversing some trades.
I applaud the exchanges for doing something they had never openly considered on such a large scale before, and I appreciate the timeframe under which they acted. It is often said that, when time is of the essence, a violently executed plan is better than a well-laid plan conducted at a later date.
To be sure, time was of the essence, because a security purchased one hour ago in a portfolio could easily be sold right now. What if I had been fortunate enough to purchase an ETF that had fallen 99% during the crisis (I was not that lucky, and I still have not found an individual investor who managed to get an order executed in the 15 minutes following 2:42 p.m. EST on May 6), and decided to sell my fund later that day? If my purchase decision had been nullified, I would now be sitting on a short position on that ETF, and to no fault of my own. Any action taken by the bourses needed to be fast, and for that, they deserve credit. But that is where the kudos end.
When Something's Not Right, It's Wrong
The conundrum of the trade nullification order is that there are two right answers. On one hand, there is a solid argument for canceling no trades whatsoever. A trade conducted on an exchange is a binding contract between two parties at an agreed-upon price--no matter what that price ends up being. If you entered a market order, as opposed to a limit order, to sell a security during the turmoil, you made a decision to part with your security at whatever price the most willing buyer would pay. Granted, precedent had been set that a willing buyer almost always appears near the last quoted price, but there was never a guarantee. Bid and ask prices have long been the price discovery mechanism for securities traded throughout the day on an exchange, and the vast majority of the time, they are effective.
The other right answer is to determine which securities exchanged hands at corrupted prices and reverse all of those trades. If the system broke down for certain securities, all the trades should be nullified regardless of price.