Low Latency Spending Moves Full Speed Ahead

Why It's Important: In the past, Wall Street's reliance on technology to reduce latency was focused mainly on algorithms and the trading of equities. While algorithmic trading of equities remains the most sensitive to data latency, other asset classes -- such as foreign exchange and certain types of derivatives, especially exchange-traded options -- are also moving more and more toward algorithmic trading and are increasingly sensitive to latency.

"Low latency is critically important in the options market, and in the coming years it will only become more so," writes Kevin McPartland, senior analyst, TABB Group, in his report "Low-Latency Options Trading: Unraveling the True Meaning of Speed." "Latency is already being reduced at each stage of the trading process but at increments and levels of priority that vary by firm."

Why is it so important in the options market? "Options is the epicenter of market data, with 2 billion messages a day," McPartland tells WS&T. "There are a lot of people executing high-speed options strategies." With the Options Price Reporting Authority (OPRA) recommending that market participants have the capacity to handle nearly 2 million messages a second (10 billion a day) by January 2009, simply handling all of that data is going to require firms to use technology that reduces latency in all parts of the process.

Where the Industry Is Now: Technologies that enable lower latency -- such as the newest quad-core chip sets, complex event processing (CEP), field-programmable gate arrays (FPGAs) and hardware acceleration -- have been used to speed the trading process. But with the industry's increased focus on reducing risk in 2009 through intraday or even real-time risk calculations, certain latency-reducing technologies will find their way into the risk management organization. This will be especially true if financial firms look to integrate real-time risk monitoring with the trading desks because risk management calculations tend to require large amounts of "compute farms," says Peter Lankford, director at STAC Research. "The latency on the compute farms has a big impact on understanding the firm's positions during the day. The speed of the processing and analysis and latency within grids can be a big deal."

Focus in 2009: A lot of the low-hanging fruit has been picked when it comes to data latency, so firms are looking for new ways to decrease latency. Hardware acceleration is gaining traction, and in the current climate of reduced budgets, firms are looking for efficiency gains that save money and also reduce latency, such as "collapsing the stack" (reducing the hops between systems to streamline processing) and improving resource efficiency in the data center, says Lankford. "Reducing latency sometimes goes hand in hand with better efficiency," he says. For instance, FPGAs are a low-power technology that can greatly reduce latency.

Industry Leaders: All of the brokers have some trading operations and algorithms that are latency-sensitive. They have all spent time and resources over the past few years aimed at reducing data latency. This trend will continue into 2009 and beyond, as for "any firm that is positioned as a trading firm, milliseconds will matter," says Robert Iati, partner, TABB Group.

Technology Providers: CEP providers include Aleri, BEA Software, Vhayu, Coral8, Streambase, Kx, Progress Apama, Oracle Fusion and KnowNow. Hardware acceleration vendors include graphics processing units (GPUs) providers AMD ATI and NVIDIA, among others.

Price Tag: Considering that TABB Group estimates that "if an agency-broker's electronic trading platform is five milli-seconds behind the competition, it could cost the firm up to $4 million," most firms will be willing to make investments in latency-reducing technology. The ultimate price tag varies, depending on the size of the firm.