Opening Cross: Getting a Head Start

Whereas some look for the early-warning price movements that might grow into an upswing, the smart players are looking for the price-moving indicators before they have begun to impact the market. It sounds so easy in theory, yet it takes complex models and monitoring of quoting versus trading trends over small time periods, and having huge volumes of historical data at your fingertips-not to mention the ability to analyze it-to predict when a price will change, and in response to what.

So while technology providers such as Redline Trading Solutions are focusing on developing hardware-based ticker plants that can process large volumes of data at low latencies-benchmarked by the Securities Technology Analysis Center at being able to process 5.6 million messages per second with latency of around 20 microseconds-others are focusing on finding the next wave of leading indicators that will power the strategies that require those low-latency solutions.

Recently, we wrote about Selerity, a start-up focused on sourcing structured, text-based releases and rulings and turning them into a machine-readable signal that traders can incorporate as an input to algorithmic trading engines-for example, to buy or sell a pharmaceutical company's stock based on whether or not it receives FDA approval for its latest drug or based on a ruling in a court case.

Now online business information provider Alacra is also launching services that consolidate commentary from analysts and company insiders, which can potentially be used as an input to automated trading strategies. However, with these less structured types of information, the vendor needed to build and apply a proprietary sentiment analysis model that creates a reliable trading signal from, say, an analyst's or executive's comments. In fact, not satisfied with only using comments on stocks as an input, the vendor is already developing a version that analyzes macroeconomic commentary to derive inputs for equities trading.

"I think this sort of trading and information is becoming more mainstream," says Steven Goldstein, chief executive of Alacra. "People came and talked to us about StreetPulse in Q4 2009, asking if we had thought about packaging it for algorithmic trading. I said ‘no' because I thought algo trading had to be high frequency and low latency, whereas StreetPulse is the opposite. But increasingly, algo traders are looking for signals with longer holding periods.... So, we started taking the data and trying to create a sentiment score."

Goldstein says that some banks are using the vendor's solutions to monitor their proprietary trading groups to identify outlier trades not clearly prompted by publicly-available information, which might signal market abuse-something that UK-based consulting body Avenues says has become harder to spot in Europe since the introduction of the Markets in Financial Instruments Directive, and as a result is calling for the creation of a consolidated tape of pan-European market data and the use of sophisticated "eavesdropping" technologies to maintain fair and orderly markets.

And this is the point where speed becomes really important: not just for traders to react the moment they receive their chosen indicator and get a head-start before everyone else spots the rising wave and plunges in, but also for compliance staff and regulators to play their part effectively in an ultra-low-latency market.

Max Bowie
[email protected]