November 17, 2010

FX Algo Adoption Soars

Meanwhile, buy-side firms are adopting FX algo trading strategies supplied by the large banks -- including Barclays Capital, Citi, Credit Suisse, Deutsche Bank and JPMorgan -- to access multiple liquidity pools and obtain executions with the least market impact, according to a recent report by Celent. In fact, the rise of FX algos is causing volume spikes as they slice and dice trades.

"We're seeing generally a trend that mirrors other asset classes: smaller ticket sizes while the actual number of tickets is increasing," eExchange's Lowry reports. "For example, if I'm an investment manager and I have to do a 20 million euro/U.S. dollar trade, ... I might use an algorithm provided by a large bank. These algos will be seen in the market as small trades of a half million to a million dollars. But taken together this could be a $400 million order."

Algorithmic trading is well suited for the forex market, with its fragmentation and relatively simple back-office demands. In its September 2010 report, Celent found that algos were used by 50 percent of active FX traders, up from a third of FX traders in 2008-2009, and were expanding to include third-party customized suites -- the so called "white-box" algos that are less proprietary than their black-box counterparts.

But, "Electronic trading is a doubled-edged sword for the banks," says Jim Downs, CEO and founder of Chicago-based Connamara Systems. "Spreads get narrower because users can look across the universe of quotes and pick the best. Electronic trading will allow more order flow across their desks, but at the same time electronic trading requires the dissemination of market data electronically," which would drive the banks further toward automated systems.

As the universe of FX destinations expands, market data and price discovery mechanisms that were once considered optional are becoming essential for buy-side traders -- providing venues with an opportunity to distinguish themselves from one another. ECN consolidators, such as Connamara Systems, could streamline the liquidity search, Downs suggests, noting that the firm's FX Market Data Adapter Suite links to four FX venues: Currenex; LavaFX, which was bought by FXall; ICAP/EBS; and Hotspot. The product was designed to normalize and integrate low-latency FX market data into both the buy side's and sell side's trading applications.

According to the Celent report, however, to succeed in the increasingly fragmented FX market, buy-side traders need to adopt new technology to handle a flood of information from multiple trading venues. Most of the solutions, Celent says, will involve upgrades to electronic dealing platforms and technologies such as complex event processing (CEP) to aggregate and filter data. In the end, asset managers swimming in the same pools as the banks, prop desks and stat arb funds will need to decide how quickly and how much they will have to embrace new technologies to stay competitive.