Asia-Pacific Financial Information Conference: Asia Weighs Benefits, Costs of Low Latency

Asia-Pacific firms must carefully evaluate the cost of investments in low-latency technologies, since such a small portion of the region supports high-frequency trading, but should consider where these investments can yield other benefits, according to panelists on an FISD panel at the Asia-Pacific Financial Information Conference.

“A focus on latency is unavoidable. But the rate at which people will implement the technologies we see in Europe and the US will be very different,” said Peter Tierney, regional chief operating officer for NYSE Technologies in Asia. Between 2007 and 2009, firms only cared about speed, Tierney says, whereas a second wave of firms buying low-latency infrastructure are more concerned with its ability to deliver cost-effective trading across different regional markets than with latency alone. “The rate at which markets adopt this… will depend a lot on return on investment,” he said.

However, proving ROI is harder in Asia than elsewhere, because to take advantage of proximity to local markets, firms must co-locate in each market center, said Denny Honess, head of global banking and markets dedicated products at Bank of America Merrill Lynch Japan—rather than, say, accessing the majority of US markets from a single co-location center.

“The market here in Asia is completely different to the US or Europe in terms of geographical spread and regulation. And there are only a handful of organizations in the low-latency space,” said Eugen Tjong, head of Enterprise Solutions for real time at Thomson Reuters in Asia-Pacific. “There are exchanges in the region that have invested a lot in co-location space but are 75 percent empty because most of the market is a tier below… and some markets are just not ready. If low latency is your only differentiator from competitors, then you have a problem.”

“Most of us in the region may not have talked about high-frequency trading yet, but we have all spent two or three years getting latency down… to get the best price for clients,” Honess said, adding that latency-reducing technologies such as field-programmable gate arrays (FPGAs) are still largely unknown locally, so anyone looking to deploy them must fully understand them.

However, firms must be free to make decisions in the region that benefit local operations, regardless of a central technology strategy, said Alan Ross, regional head of market data services for Asia-Pacific at UBS. “In any global bank, there’s an amount of strategy that has to be global—that’s how you get firm-wide economies of scale—but then we have to evaluate aspects of specific markets… and the change we’ve seen in the last couple of years is towards managed services and the idea of data as a utility,” Ross said.

Thomson Reuters’ Tjong echoed this, highlighting the importance of making it easier to consume and distribute data. “Data consumption has gone up dramatically over five—or even two—years ago… and pressure on institutions is growing… so there are great opportunities for vendors who can effectively deliver this to users in the right way at the right time,” he said.