Securities Technology Monitor 9/16/10
Earlier this week, the UBS Investment Bank took a major stake in a nearly unknown company called Mantara.
Its claim to fame: a “superfast” method of moving billions of messages through trading networks. This is critical to high-frequency trading, where strategies often depend on flooding the system with orders, analyzing market reactions instantly and then flooding the market with cancellations.
Mantara’s product is known as expressWay. Until 2003, it was a technology in search of a solution, in Australia.
Then known as Elvin, the idea was to use create, in a way, a sophisticated twitter system. Users would create expressions using a set of mathematical expressions that detailed the numbers and character strings they were looking for. This in turn would create a event notification system that trading algorithms can act on, quickly. You could call it a “distributed event routing service.”
This particular approach to publish-and-subscribe content delivery was originally developed at the Distributed Systems Technology Centre at the University of Queensland in Australia (think Brisbane).
In 2003, the research group that created Elvin left to start up Mantara. Eventually, folks like Mohr Davidow Partners and CM Capital recognized the potential use in high-frequency trading.
And UBS, which as recently as November of 2009, said it was not in the high-frequency trading business, now is adding the technology to what it’s offering the clients it serves from what is possibly the world’s largest trading floor, housed in Stamford, CT.
Titan Trading Analytics is a similar tale. The firm got started in Edmonton, Canada, as a neural networking firm. Neural networks try to resemble the neural processes of the brain. Where basic computing is about processing series of digits linearly, as fast as possible, neural networks try to process many streams of information at the same time – and recognize patterns in the streams as well as specific “sensory” inputs.
It’s this last aspect, the search for patterns, that yielded the application to Wall Street.
Titan is now focused on modeling financial behavior. Specifically, the herd mentality in stock markets. It’s idea: Figure out what in the past has reliably shown emotion to be taking over in the trading of a stock or stocks and send out a trading signal, when there’s a clear indication that emotion is about to take over again.
For this task, it’s compiled its own pile of tick-by-tick data from all U.S. equities exchanges, including the New York Stock Exchange, NYSE Arca, NYSE Amex, the Nasdaq Stock Market, BATS Exchange and the Direct Edge venues.
Besides collecting market data, it mixes in streams of economic indicators and news that shows sentiment about stocks in general or a stock specifically.
Then, it looks for instances where a certain set of conditions has reliably – like five or six times – produced the same emotional response in the past.
This occurred last year, with oil, contends president John Coulter. Prices in the oil rig sector spiked 25 percent in about a three-month period. Offshore drilling looked like the next big thing (until the Deepwater Horizon). Money was pouring in.
But the system had seen this phenomenon before, recognized that there were alternative energy sources that the spike would encourage development of, and sent out trade signals, by Coulter’s account, four days before a correction sent in.
Timing is everything, of course. And Titan is totally unproven. Its first product comes to market October 1. That’s when it’s “in the lab” record of winning 73 percent of the time that it sends out trade signals really starts to get tested.
The idea that math can be used to check outbursts of “irrational exuberance” is comforting of course. It could easily be applied, in theory, to popping any market bubble, as the air is reaching capacity inside the balloon.
It’s show time now, however, for the talent in question. Michael Chin, who announced the deal with UBS, was only announced as CEO at Mantara on September 7. Coulter was only announced as CEO at Titan on Monday.
It’s on their shoulders to show that it’s worth hunting for obscure technical talent in Australia, in Asia, in Africa, in the Americas. That developing technologies that let the minds of brilliant traders work faster is the talent that matters.
Or a technology that can suss out the difference between the minds and hearts of buyers and sellers. Before they figure it out themselves.