The Flash Crash Plus Two Months

By 2pm on May 6, 2010, it was clear that the stock market was having a bad day. The Greek Crisis was going through one of its spasms of fear and uncertainty. It was also the day of the UK general election, and by mid-Traders Crop afternoon in New York it was becoming clear that the result would be some flavor of a hung parliament. The S&P 500 was down –2.9%, a decline that only a few years ago would have been considered headline news but is now mundanely bad.

Then over the next 46 minutes, the decline accelerated to the point where it became severe, then horrendous, and finally absurd. At about 2:46 the S&P was down about –8.6%. I say “about” because at that point the basic fabric of the marketplace was breaking down and exactly what the S&P was trading at, and when, is not clear even months later.

And then, as if the market fairy who had caused this twitched her wand in the other direction, the market staged its best ten minute rally in history. By 2:55 the S&P was down only -3.9% on the day. It would close down –3.2%.

The immediate response to this was a unanimous “What? Huh? Hold on, what just happened? I just went to the men’s room and …. Jeez.”

Regulators announced investigations. Pundits theorized. Both houses of Congress held hearings. And two months later we are no closer to finding the person or persons who screwed up to cause this.

We did uncover a few problems in the way that the market functions, technical issues that seemed pretty harmless up until 2:40 on May 6.

For example, the various sub-markets that collectively make up the American stock market have a procedure under which they can declare “self-help” and break their connections with the other sub-markets. The intent is to stop a computer glitch or other malfunction from spreading from one market to another.

Of course, the events of the 2:40 – 2:50 period sure looked like an error, so the people running some of the markets declared self-help and those markets broke apart, making each one an island of less liquidity. That made things worse. Just when you would have most wanted the markets to be tightly wired together, the humans running the computers decided to disconnect them.

And there were the stub quotes. A stub quote is (and I realize this is meaningless to you under-40s) a test pattern. It is a fake pair of offers to buy and sell at ridiculous prices that a broker-dealer puts out just to make sure that the computer network that carries the quotes is working. If you go on-line before the market opens in the morning sometimes you will see a bid-ask for a stock of $0.01 and $100,000. Those are stub quotes and, taken literally, they mean that you can sell shares for a penny or buy them at a hundred grand.

As far as anybody can recall, nobody ever accidentally traded at a stub quote price prior to May 6. That day, when real quotes disappeared in the confusion, several stocks traded at stub prices, although only very briefly and only on the fully automated, that is, non-NYSE, exchanges. Accenture went for a penny a share. And Apple traded at $100,000, up 40,000% on the day. If you are feeling bad that you missed an opportunity to buy at penny, don’t be. All those trades were “busted” or cancelled afterwards.

I must say that I object to this. The exchanges cancelled all trades made between 2:40 and 3:00 at prices more than 60% from the price at 2:40. 10,000 trades for 1.4 million shares were busted. As James Stewart at the WSJ points out, the traders who sold at a penny and bought at $100K were sophisticated users of automated algorithmic trading systems who have only themselves to blame. They agreed to sell at one cent, or authorized a computer to agree for them, and a deal is a deal. Suck it up.

Moreover, breaking extreme trades sets a bad precedent that may make things worse next time.

Imagine that because of some fluke XYZ is down from $50 to $25 in the course of a few minutes. What you want to have happen is for some brave soul to stop the panic and buy all he can at $25. That takes bravery because although it is generally true that a sudden drop of 50% is unwarranted, and probably nuts, there is the chance that the sellers know something that you don’t, and you don’t have the time to research the point. You need to balance off the prospect of a quick killing if the price drop is unjustified with the possibility of getting killed if it is. Add in the scenario that if the price is later considered to be an error of some kind your profitable trade may be cancelled, and the scales are tipped in favor of doing nothing. Which thins the ranks of brave traders willing to stop a panic.

Indeed, I think a lack of traders willing to step in front of the thundering herd of Chicken Littles is partly to blame for the events of May 6 and the volatility we have been experiencing over the past two years. Before the summer of 2008, the big investment banks maintained large “prop desks” full of well bankrolled traders who made a dangerous living buying and selling at stupid prices. They would take the other side of the trade when panic or enthusiasm got the best of investors. That helped stabilize the market, blunting wild price swings.

Post-2008, there are fewer investment banks, the ones that are left have less money to work with, they are more risk averse, and, it must be said, making money this way is something of a political liability. So the market is a lot more volatile.

What nobody seems willing to say out loud about the events of May 6 is that, by and large, the market worked the way it was supposed to. The basic presumption that something so extraordinary could only happen because of some exceptional act of stupidity or malfeasance is false.

This particular sequence of events had never happened before, and will likely never happen again, but in a general way things like this have always happened in the stock market. The only thing that is different now is the speed. What took only twenty minutes on May 6 would have taken an entire day or even two in previous decades. And that’s a good thing.

categories Government, Investing | | datetime July 6, 2010 11:59 am

2,477 Comments

Other Links to this Post

  1. What Other Bloggers are Saying | Options for Rookies — April 24, 2011 @ 6:27 pm

  2. child porn — September 3, 2023 @ 7:16 am

  3. fuck google — September 4, 2023 @ 10:27 am

  4. fuck google — September 5, 2023 @ 8:24 am

  5. porn — October 8, 2023 @ 8:01 am

  6. porn — October 9, 2023 @ 7:19 am

  7. porn — October 10, 2023 @ 3:48 pm

  8. porn — October 11, 2023 @ 5:25 am

  9. porn — October 11, 2023 @ 8:31 pm

  10. child porn — October 13, 2023 @ 6:54 pm

  11. porn — October 17, 2023 @ 6:58 am

  12. child porn — October 18, 2023 @ 6:22 am

  13. child porn — October 20, 2023 @ 10:28 am

  14. child porn — October 20, 2023 @ 11:41 am

  15. child porn — October 22, 2023 @ 9:20 am

  16. porn — June 8, 2024 @ 8:29 am

  17. porn — June 29, 2024 @ 5:06 am

  18. child porn — December 25, 2024 @ 3:13 pm

RSS feed for comments on this post. TrackBack URI

Leave a comment

WordPress Themes