I don’t know how many of the readers here pay attention to what’s happening in Wall Street. Yesterdays trade was quite spectacular, although from many points of view it is terrifying and it is a day that will probably live in infamy. From the academic point of view I’m sure it will be studied to exhaustion.
The most interesting thing that happened was a huge spike where the stock market plunged about 8% to recover within 20 minutes about 7% upward. This spike shifted a huge amount of money. If you want to read more about it, you can go to one of the articles in the New York Times here. IN the parlance of statistics, this was a large fluctuation, beyond those that are normal. This is where you know that the Gaussian errors don’t necessarily apply and its most likely that whatever model one has is wrong.
For me I had just finished writing some reports, and was at a break point when I started reading the news before going to lunch. This was coincidentally just as the stock market had plunged some 8%. I immediately called my broker and told him to buy everything, because it had to be a mistake. When he did buy everything as I instructed him, I had saved the day and made a tidy profit. Well, the part about the broker is fiction (as well as the profit ), but I realized that the news hadn’t been so bad that one would expect such a large deviation, and if I had any money to use I would have not hesitated in buying at the moment. Indeed, all of this instability can be attributed to Greece and the sense that they should be punished for their misdeeds, as is the custom when fate is involved . After all, we are talking about a Greek tragedy. (Ok bad joke.)
The most interesting thing of all about this episode are not my personal thoughts on what the markets are worth, but rather the rapidity with which this episode took place. Someone seems to have made an error. An army of robots sells. The market goes down. People realize that this is wrong and an another army of robots buys everything. All within a few minutes. The rest of us were lucky to even see it happen.
This shows the power of automated buying on large scales. Most hedge-funds, etc have in the end similar strategies for when to hit the buy/sell buttons with regards to fluctuations in the market. This means that when an anomaly shows (the large fluctuation), the effect is going to be exaggerated by the inherent non-linearity of the response: when things start to go down, you trigger sell buttons, and as soon as you hit some ‘floor of the day’, you hit buy. This happens even while you are at lunch with your friends. It is completely automated. Because it is so fast, one can make (or lose) huge amounts of money. The winner is usually the one with the fastest button whose algorithm gets a first hint of a whiff of trouble.
To malappropriate the words of a Wall Street insider: “They did god’s work“. If they mean causing what are considered acts of god, like earthquakes and other events that generate enormous amounts of poverty and mayhem on a single stroke, they might even be right about that one. I’m sure they would like to take back those words.
At the end of the day, everyone was left scratching their head because stuff like this is not supposed to happen. After all,
markets are self correcting
so we should be buying into the idea of markets. Except that this blip shows that this is just a fiction and that current technology makes the markets very unstable for short terms. There are going to be a lot of investigations surrounding this episode, and also cries for more regulation. With the usual crying against regulation by the entrenched interests that are happy with the system as it is. I still want to know who is going to clean that mess up and how are they going to prevent stuff like this from happening in the future.
In the meantime, and for the rest of us, life just goes on. I had chicken for dinner and it seems that I am just a little poorer today than yesterday. After all, I had to replace my water heater yesterday and bills have to be paid.


By Occam’s razor, the simplest explanation is likely to be the right one.
At the moment of the “human fate or error”, it was said that using Ariel by men who shave themselves by Gilette and clean their teeth by Oral-B, assuming that their babies have Pampers older than one day, causes global warming proportional to the number of the Duracell batteries they own: the albedo from the clean laundry, coupled to the missing energy from the batteries etc. creates extra CO2 which warms the planet.
Consequently, people began to sell Procter & Gamble which produces all the brands above. At the minimum, the traders who sold P&G were sent to mental asylum, so the stocks recovered.
As you see, it was simple – but you shouldn’t make it simpler than it is, so please remember all the P&G products whose combination causes global warming.
What is more spectacular about all this is that it also illustrates the power of that one can wield by understanding a complex interacted system. The stock market is still a input output machine (whether mechanized by computers or human based). Certain deliberate inputs can lead to rapid cancellation of damping mechanisms as the inputs compound. The use of computers in trading can make predicting the effects of inputs more predictable and allow for predictable rapid changes at the hands of a some really intelligent agent (or by some random process as well). Although I think this most recent fluctuation was truly “an act of god” in that there was no deliberate human action, the danger going forward is that even if we identify the cause for this particular fluctuation, some really smart people are going to be able figure out how to do the same thing using some other strategy. Of course all this is very speculative, but not out of the art of the possible.
One more strike against Wall Street; can it retain its preeminence after first demonstrating utter incompetence in pricing mortgage securities, and now revealing that it can’t even manage day-to-day trading?
I’m guessing yes, b/c there is no serious contender in the world yet – but if they don’t get their house in order, that business will eventually go elsewhere. Nobody wants to trade with firms that operate in bad faith, nor trade on exchanges where you can be ruined by a programming error.
These kinds of things demonstrate that the financial markets are not random in that the tail fluctuations always begin to the downside.
What bothers me more than anything else, is that all those computer traders are in principle predictable. They run on algorithms that are relatively straightforward. If some informed person knows the exact algorithm and details in the 10 or so packages that are out there, and that those algorithms are responsible for a huge volume of trading, then he can use that information to game the system.
When everyone does something one way (or ten ways), it makes fluctuations predictable a priori which consequently breaks the efficient market hypothesis (which is supposed to be a great stabilizer).
So in the future it become profitable to provoke crisis’ like the one that happened, b/c you know that your computer (with a brand new algorithm designed to beat the competition for a short period) will make a killing shorting the stocks for a few minutes, then buying everything back before the competitors.
This has a tendency to worsen both the frequency and intensity of the volatility and one day one might worry that we don’t simply lose an entire index in 4 minutes before a single human can flip the on/off button.
The subject of autonomous trading models can lead one down many paths and none have clear answers. I can offer my opinions, however.
Having traded for twenty years (many of those as a market maker in currencies) covering just about every product I feel that a few years ago we moved from a place where automation was a tool which helped traders in their jobs to one in which it has destroyed their jobs. In virtually every market now trading is to a large extent dominated by the various trading model systems and liquidity bots. The days of having a feel for the market, either by learned instinct or as a function of order flow, are long over. There are actually sytems out there who’s whole existences is to simply earn liquidity credits (a kickback for being a price giver) off the bid/ask spread.
Spreads have eroded significantly as well in all markets. We trade pennies now in stocks which I’m sure you all are familar, but in things like currencies, what was once a bid/ask of 0.0005 is now 0.0001.
The result of all of this is that there are fewer interested parties at any given time. Where there used to be a dealer quoting a two way price in the hopes of making the spead and gaining insight from the trade/order flow, now there is none (well, far fewer). Where once customers and other traders would book their buy and sell orders or stops ahead of time, now there is far less depth away from the current market. Put it all together and it is fairly easy to see how events like last week happen.
Look at Accenture. It went from $40 to $0.01. How does this happen? Because there are no buyers and no market makers. In stocks at least, market makers used to be responsible for providing an orderly market and to be buyer (or seller) of last resort. Most have stepped away from that job. Where were the buy orders? There were a few, but not of remotely sufficient depth to absorb the automated stupidity.
While the rebound was just as fast (gee, where did the sellers go?), I think this serves as an excellent example of why naked short selling must be curbed. The markets are structurally no longer equipped to absorb the kind of selling which these systems can generate. Perhaps if the overall risk were symmetric I would not feel the same way, but it is not. As a companies stock price is pummeled, rumor of its demise begins and lenders get nervous and ultimately all avenues of liquidity for daily operations dry up resulting in chapter 11.
I’d close by saying what you describe at the end of your post is the equivalent of the way trading used to be done. Trading is all about balls and whos are biggest. You would be surprised what one trader can do to move a market without even taking a very large position. If one senses the market is very long, you sell by offering lower and lower. Yes, initially some will take your price. But if you stay there and keep going lower, panic sets in. WTF is going on they say? Is that Soros? Is there news? Etc, etc. Eventually you are the bid and get given their longs well lower.
Sorry for being a bit rambling and long winded!