These days, a lot of buying and selling of stocks is done by computers—it’s called algorithmic trading. Computers can do it much faster than people. Watch how they’ve been going wild!
The date is at lower left. In 2000 it took several seconds for computers to make a trade. By 2010 the time had dropped to milliseconds… or even microseconds. And around this year, market activity started becoming much more intense.
I can’t even see the Flash Crash on May 6 of 2010—also known as The Crash of 2:45. The Dow Jones plummeted 9% in 5 minutes, then quickly bounced back. For fifteen minutes, the economy lost a trillion dollars. Then it reappeared.
But on August 5, 2011, when the credit rating of the US got downgraded, you’ll see the activity explode! And it’s been crazy ever since.
The movie above was created by Nanex, a company that provides market data to traders. The x axis shows the time of day, from 9:30 to 16:00. The y axis… well, it’s the amount of some activity per unit time, but they don’t say what. Do you know?
The folks at Nanex have something very interesting to say about this. It’s not high frequency trading or ‘HFT’ that they’re worried about—that’s actually gone down slightly from 2008 to 2012. What’s gone up is ‘high frequency quoting’, also known as ‘quote spam’ or ‘quote stuffing’.
Over on Google+, Sergey Ten explained the idea to me:
Quote spam is a well-known tactic. It used by high-frequency traders to get competitive advantage over other high-frequency traders. HF traders generate high-frequency quote spam using a pseudorandom (or otherwise structured) algorithm, with his computers coded to ignore it. His competitors don’t know the generating algorithm and have to process each quote, thus increasing their load, consuming bandwidth and getting a slight delay in processing.
A quote is an offer to buy or sell stock at a given price. For a clear and entertaining of how this works and why traders are locked into a race for speed, try:
I don’t know a great introduction to quote spam, but this paper isn’t bad:
• Jared F. Egginton, Bonnie F. Van Ness, and Robert A. Van Ness, Quote stuffing, 15 March 2012.
Toward the physical limits of speed
In fact, the battle for speed is so intense that trading has run up against the speed of light.
For example, by 2013 there will be a new transatlantic cable at the bottom of the ocean, the first in a decade. Why? Just to cut the communication time between US and UK traders by 5 milliseconds. The new fiber optic line will be straighter than existing ones:
“As a rule of thumb, each 62 miles that the light has to travel takes about 1 millisecond,” Thorvardarson says. “So by straightening the route between Halifax and London, we have actually shortened the cable by 310 miles, or 5 milliseconds.”
Meanwhile, a London-based company called Fixnetix has developed a special computer chip that can prepare a trade in just 740 nanoseconds. But why stop at nanoseconds?
With the race for the lowest “latency” continuing, some market participants are even talking about picoseconds––trillionths of a second. At first the realm of physics and math and then computer science, the picosecond looms as the next time barrier.
Actions that take place in nanoseconds and picoseconds in some cases run up against the sheer limitations of physics, said Mark Palmer, chief executive of Lexington, Mass.-based StreamBase Systems.
Black swans and the ultrafast machine ecology
As high-frequency trading and high-frequency quoting leave slow-paced human reaction times in the dust, markets start to behave differently. Here’s a great paper about that:
• Neil Johnson, Guannan Zhao, Eric Hunsader, Jing Meng, Amith Ravindar, Spencer Carran amd Brian Tivnan, Financial black swans driven by ultrafast machine ecology.
A black swan is an unexpectedly dramatic event, like a market crash or a stock bubble that bursts. But according to this paper, such events are now happening all the time at speeds beyond our perception!
It’s a price spike in the stock of a company called Super Micro Computer, Inc.. On October 1st, 2010, it shot up 26% and then crashed back down. But this all happened in 25 milliseconds!
These ultrafast black swans happen at least once a day. And they happen most of all to financial institutions.
Here’s a great blog article about this stuff:
• Mark Buchanan, Approaching the singularity—in global finance, The Physics of Finance, 13 February 2012.
I won’t try to outdo Buchanan’s analysis. I’ll just quote the abstract of the original paper:
Society’s drive toward ever faster socio-technical systems, means that there is an urgent need to understand the threat from ‘black swan’ extreme events that might emerge. On 6 May 2010, it took just five minutes for a spontaneous mix of human and machine interactions in the global trading cyberspace to generate an unprecedented system-wide Flash Crash. However, little is known about what lies ahead in the crucial sub-second regime where humans become unable to respond or intervene sufficiently quickly. Here we analyze a set of 18,520 ultrafast black swan events that we have uncovered in stock-price movements between 2006 and 2011. We provide empirical evidence for, and an accompanying theory of, an abrupt system-wide transition from a mixed human-machine phase to a new all-machine phase characterized by frequent black swan events with ultrafast durations (<650ms for crashes, <950ms for spikes). Our theory quantifies the systemic fluctuations in these two distinct phases in terms of the diversity of the system's internal ecology and the amount of global information being processed. Our finding that the ten most susceptible entities are major international banks, hints at a hidden relationship between these ultrafast 'fractures' and the slow 'breaking' of the global financial system post-2006. More generally, our work provides tools to help predict and mitigate the systemic risk developing in any complex socio-technical system that attempts to operate at, or beyond, the limits of human response times.
When you get into an arms race of trying to write algorithms whose behavior other algorithms can’t predict, the math involved gets very tricky. Over on Google+, F. Lengvel pointed out something strange. In May 2010, Christian Marks claimed that financiers were hiring experts on large ordinals—crudely speaking, big infinite numbers!—to design algorithms that were hard to outwit.
I can’t confirm his account, but I can’t resist quoting it:
In an unexpected development for the depressed market for mathematical logicians, Wall Street has begun quietly and aggressively recruiting proof theorists and recursion theorists for their expertise in applying ordinal notations and ordinal collapsing functions to high-frequency algorithmic trading. Ordinal notations, which specify sequences of ordinal numbers of ever increasing complexity, are being used by elite trading operations to parameterize families of trading strategies of breathtaking sophistication.
The monetary advantage of the current strategy is rapidly exhausted after a lifetime of approximately four seconds — an eternity for a machine, but barely enough time for a human to begin to comprehend what happened. The algorithm then switches to another trading strategy of higher ordinal rank, and uses this for a few seconds on one or more electronic exchanges, and so on, while opponent algorithms attempt the same maneuvers, risking billions of dollars in the process.
The elusive and highly coveted positions for proof theorists on Wall Street, where they are known as trans-quantitative analysts, have not been advertised, to the chagrin of executive recruiters who work on commission. Elite hedge funds and bank holding companies have been discreetly approaching mathematical logicians who have programming experience and who are familiar with arcane software such as the ordinal calculator. A few logicians were offered seven figure salaries, according to a source who was not authorized to speak on the matter.
Is this for real? I like the idea of ‘trans-quantitative analysts’: it reminds me of ‘transfinite numbers’, which is another name for infinities. But it sounds a bit like a joke, and I haven’t been able to track down any references to trans-quantitative analysts, except people talking about Christian Marks’ blog article.
I understand a bit about ordinal notations, but I don’t think this is the time to go into that—not before I’m sure this stuff is for real. Instead, I’d rather reflect on a comment of Boris Borcic over on Google+:
Last week it occurred to me that LessWrong and OvercomingBias together might play a role to explain why Singularists don’t seem to worry about High Frequency Robot Trading as a possible pathway for Singularity-like developments. I mean IMO they should, the Singularity is about machines taking control, ownership is control, HFT involves slicing ownership in time-slices too narrow for humans to know themselves owners and possibly control.
The ensuing discussion got diverted to the question of whether algorithmic trading involved ‘intelligence’, but maybe intelligence is not the point. Perhaps algorithmic traders have become successful parasites on the financial system without themselves being intelligent, simply by virtue of their speed. And the financial system, in turn, seems to be a successful parasite on our economy. Regulatory capture—the control of the regulating agencies by the industry they’re supposed to be regulating—seems almost complete. Where will this lead?