The Straw That Broke the Market's Back

If you follow the markets you may have noticed the release late last week of the SEC's investigation of May 6th's "flash crash". That was the day the S&P 500 and Dow Jones Industrial Index dropped something like 700+ points in  mere minutes. I remember that quite vividly since I happened to call up a market screen in the midst of that event. I quite naturally assumed that there was a data feed error or some other misquote, but the problem persisted and was visible on other quoting services, and began to appear in live market commentary.

As the report notes (here's a short, readable summary) there was a lot of downward pressure that day for various legitimate reasons. However, this sort of drop should not be possible, or at least very improbable, due to the many checks and balances built into the market and quoting systems. Clearly this was insufficient. The transaction trail led back to the trigger, which was a large E-mini futures trade by a respectable firm that runs mutual funds as a large part of its business. The futures contract was reportedly a hedge against its longer-term investments.

E-Mini S&P, often abbreviated to "E-mini" and designated by the commodity ticker symbol ES, is a stock market index futures contract traded on the Chicago Mercantile Exchange's Globex electronic trading platform. The notional value of one contract is US$50 times the value of the S&P 500 stock index.

It was introduced by the CME on September 9th, 1997, after the value of the existing S&P contract (then valued at $500 times the index, or over $500,000 at the time) became too large for many small traders. The E-Mini quickly became the most popular equity index futures contract in the world. The original ("big") S&P contract was subsequently split 2:1, bringing it to $250 times the index. Hedge funds often prefer trading the E-Mini over the big S&P since the latter still uses the open outcry pit trading method, with its inherent delays, versus the all-electronic Globex system. The current average daily implied volume for the E-mini is over $140 billion, far exceeding the combined traded dollar volume of the underlying 500 stocks.

Although I can't pretend to know or understand the intricacies of the markets events which led to this, I am dissatisfied with this summary. Let me explain my basic reasoning. Think of a sports match such as a hockey game between the Leafs and the Sens. It's a hard fought 58 minutes where the sides take turns scoring, and there are controversial calls and a lot of penalty minutes on both sides. There are heroes and there are goats, which the crowd cheers and boos. Then two minutes before the buzzer, with the Sens having one-man advantage, a quick passing play through the Leafs' defense leads to a Sens' goal and they go on to win 5 to 4 two minutes later.

The question is: who caused the Leafs to lose or, conversely, the Sens to win? Commentators and fans scramble to lay the blame and praise (depending on whose side they're on). Was it the Leaf player who grabbed a jersey in the dying minutes and got a penalty? Perhaps it was the goalie or a defender that made the wrong move? It gets far worse when you broaden your scope further. After all, there were 59:30 minutes of action, both good and bad plays, goals and penalties, before the seminal event. Doesn't any of that count? And what about the final two minutes of play: shouldn't the Sens' time killing tactics get praise or the Leafs' failure to get a shot on goal be panned?

Yet we still focus on the goal scorer and the goalie who failed to stop the shot, which is perhaps just one of a hundred key points during the game. I feel that the SEC did much the same thing by focussing attention on that one trade in their report. As Bloomberg reports:

Jeffrey Hopson, an analyst at Stifel Nicolaus & Co. in St. Louis, said the report is unlikely to damage the company in the long term.

“It’s not ideal PR, but the activity didn’t represent anything unusual,” Hopson said in a telephone interview.

Futures options are typically asset-backed derivatives. Unlike a bet at the track that does not affect the performance of the racehorses, an option does impact the market. Someone, somewhere has to cover the asset that the option is on. If it's on the S&P 500, the asset is likely an index fund which contains a large inventory of real shares in the company that are in the S&P 500. A large futures trade will move the market.

“One key lesson is that under stressed market conditions, the automated execution of a large sell order can trigger extreme price movements, especially if the automated execution algorithm does not take prices into account,” according to the report from the U.S. Securities and Exchange Commission and Commodity Futures Trading Commission.

There are two points here. First, when you trade either an illiquid security or the trade size is a significant portion of the average daily trading volume of that security you are foolish to not use a limit order. This goes for individual investors buying or selling shares of a small technology company, but all large financial firms trading billions of dollars of shares of the world's largest companies. What the purported initiator of that E-mini trade did was in this regard foolish but not improper.

The second and what I believe is the more important point, there were lots of sellers waiting in the wings. That these sellers were programs lurking just below the market quotes does not much matter. The thing is that when the market is going down, conservative investors and also active traders sells when the market goes down, just as they are wont to buy when the market is going up. This is the opposite of many small and inexperienced investors that choose to buy when the market is down and buy when it is up. It is of course more complicated than this -- actual trading strategies are infinitely more nuanced and focussed, even for small investors -- but there is some excellent rationale behind this trading style. It's about momentum: as with a ball that starts rolling downhill, if the market sentiment is negative, it will pick up speed.

That is why all those sell orders were lurking below. Programs had set thresholds on volume and velocity of selling (and the phase of the moon for all I know) that it used to determine downward momentum, and started selling. The selling drops the price, which causes more selling, and so forth on to the bottom of the crash.

The global economy and recent news events played a key role in setting the parameters behind the programs and the mindset of manual trading activity. We should not forget that, just as we should not forget the programs selling that dug the hole deeper. Focussing on any one trade, whether this one or another one, is a distraction. It may have been important, especially since it was a large market order (no price limit) on a high-leverage index, but it wasn't the only player in the game and it was not the only play that day.

I would hope that the real lesson learned is on how to improve the market's circuit breakers -- the real-time checks and balances -- or on better trading constraints before an anomalous event, whether or not they are executed by programs. We need the markets to be more robust than a house of cards.