Booms, 'bots, And Bogus Flash Boys

Booms, 'bots, And Bogus Flash Boys

The arrest of the trader who allegedly triggered the 2010 Flash Crash has a lot of people talking about algorithms, the computer programs that do most of the trading for us now. Those of us traders and investors who are not robots should know more about them.   



Booms, 'bots, And Bogus Flash Boys
Booms, 'bots, And Bogus Flash Boys

In fact, since algos are everywhere, not just in trading but in how Pandora picks songs and Amazon picks "Recommended for you" items, everyone should know a little about them.

It's only fair. They know a lot about you: your likes and dislikes, what disease you may have, how to do your job, and maybe even how to crash your economy.

THE FLASH BOY WHO CRIED WOLF

Navinder Singh Sarao, the London trader who put on thousands of "spoof" or fake sell orders only to yank them off when the market moved close to them, was arrested yesterday. At first, he did the orders manually, from his home PC. When he realized he was too slow, he had his FCM write a special "cancel if close" function into his trading software. That algorithm could place and remove thousands of sell orders in seconds, making it look to other traders as though there was a large would-be seller always lurking just above the market.

By placing all these fake sell orders, Sarao would artificially drive down the price of the E-mini futures. It's classic spoofing: He'd place a lot of big orders to sell, everyone else would say, "Ooh look at all those big sell orders, I'd better sell too," they'd sell, the market would go down, he'd buy, he'd turn off his algorithm, everyone else would say, "Oh hey never mind, things are great again, there are no more big sell orders," they'd buy, the price would go back up, and Sarao would sell the futures he'd bought at a lower price a moment ago.

The Bloomberg article quoted above raises some good questions about the arrest. For one, even at the peak, Sarao's algo only accounted for 20% of the volume of the flash-crashing market.

I would add two more questions: how did the trade in the S&P futures have such a powerful effect on the equity markets themselves? The futures are supposed to be a derivative, a reflection of the underlying index. This is the tail wagging the dog. My other question is simpler. His FCM knew what he was doing (heck, they wrote him the algo that helped him do it) and didn't question him even after the CFTC warnings. What was his FCM thinking?

Okay, one more obvious question: is he really the only one playing this game?

ALGOS ARE NEITHER GOOD NOR BAD

Algorithms, since they were invented by Euclid in Greece and Brahmagupta in India, are decision making sequences. They were further developed by the Persian Al-Khwarizmi, whose writings were the source for Europeans, who called them algorithms. All algorithms begin with the first decision: what do you want this mathematical robot to do? Therein lies the ethical choice.

Some algorithms do marvelous things, like processing millions of symptom profiles to diagnose rare diseases in patients. Others write a growing portion of daily news articles. They manage increasingly large investments, often beating human financial advisors. And they do work that humans need not and should not do.

One good example of robots taking jobs from humans: for a long time, camel races in Arab countries used child jockeys, often slaves from SE Asia, India, and Africa. Abu Dhabi's camel races have now banned the use of children and replaced them with robots. Their algos are simple: when the camel slows down, a motor turns faster and whips the camel harder. That particular algo is good for everyone, except possibly the camel.

Yale lecturer Vikram Mansharamani makes the point that more robots in factories may actually be good, because while they would take away some jobs from factory workers, they would also make labor costs a non-factor. So instead of moving factories to countries with cheap labor, companies could keep them at home.

ILLUSION AND COLLUSION

Algorithmic trading now makes up 60-80% of volume in many markets, including the e-mini S&P futures that Sarao was trading (or pretending to trade). Sarao used his algos to project a fake image of a large seller waiting in the wings. It's like he projected a big, scary hologram for the entire market to see.

And when they saw it on May 6, 2010, the market sold, all together, all in. Sarao made less than $900,000 shorting the S&P that day. In about 15 minutes, over $1 trillion of stock value temporarily disappeared. The stocks of eight major S&P firms went to one cent per share. At the same time, other stocks went to over $100,000 a share. Sarao had nothing to do with it; his algo was off during the actual crash. It was other traders and other algos.

What that says is, the Flash Crash didn't happen because of a rogue trader in London with a simple gimmicky algo that was barely an algo. It happened because of what a lot of other people did when he spooked them. It was their collusion, not his illusion, that drove the price down so far so fast.

There are other illusions at work in major markets right now: ghost cities in China built during a housing boom financed by over-leveraged debt, an average price-to-earnings ratio that is unprecedented and difficult to explain. Dr. Mansharamani, in his popular Yale course on Booms and Busts, mentions two other precursors to crashes: skyscrapers and overpriced art. In 2006, Gustav Klimt's Portrait of Adele Bloch-Bauer I sold for a then-record $130 million, one of three paintings to sell for that much that year, all to Americans. We know what happened in the US the next year.

Now, Sotheby's reports that its biggest buyers are wealthy Chinese. A Picasso recently went for $107 million, a record. And a vase expected to fetch $800 at a Christie's auction went for $18 million. Meanwhile, China is about to take five of the top ten slots in the tallest skyscraper ranks, even as it faces a massive debt crisis.

The danger is not so much that another "rogue" trader at a home PC will unleash a tricky algo to spoof the global markets. The real danger is what everybody else's algos will do in response.


This information has been prepared by Nadex, a trading name of North American Derivatives Exchange, Inc., prepared by independent third parties contracted by Nadex or reproduced form third party news agencies. In addition to the disclaimer below, the material on this page does not contain an offer of, or solicitation for, a transaction in any financial instrument. Nadex accepts no responsibility for any use that may be made of these comments and for any consequences that result. No representation or warranty is given as to the accuracy or completeness of this information. Consequently any person acting on it does so entirely at their own risk. Any research provided does not have regard to the specific investment objectives, financial situation and needs of any specific person who may receive it. It has not been prepared in accordance with legal requirements designed to promote the independence of investment research and as such is considered to be a marketing communication.