Automated electronic trading powered by sophisticated algorithms is not just for equities anymore. The tools, which use complex mathematical equations to determine market-optimal and cost-effective trading strategies, are set to aid more in the trading of options, futures, foreign exchange and even fixed income.
The moves result from multiple factors impacting these sectors, including the outsize growth in their trading volumes, market fragmentation from boosts in electronic transacting, increased demand for multi-asset class or "side-by-side" trading, and rising industry and regulatory efforts aimed at improving their competitive market structures in line with the digital age.
This convergence has spurred demand among major brokerage firms and institutions for algorithmic trading across products, and set the stage for further adoption of cutting-edge, multi-asset class trading methods.
Firms have specifically called for an ability to simultaneously execute trades in options and their underlying securities in combination with futures transactions. Such a capability would allow them to apply state-of-the-art hedging strategies in managing risk.
While the larger brokers already offer algorithmic trading for multiple security types, many are working at improving their offerings in specific products, such as options and foreign exchange (FX), and in transitioning proprietary trading strategies to their customers. "Firms' dealing rooms are using [multi-asset class algorithmic] strategies today, they're just not offering them as agency strategies to their clients," says Rob Flatley, managing director of Electronic Trading Services (ETS) at Banc of America Securities.
While he would not describe ETS' specific plans to upgrade algorithmic trading for customers in options and futures, Flatley says he is "investing" in improving quantitative e-trading capabilities across these and other asset classes, including in FX and fixed income. "You can get the same type of value from automated trading through algorithms in the futures market, and for that matter in the fixed income market and FX market, as you can from algorithms in the equity market."
Goldman Sachs in February added futures algorithms along with FX and options spread trading to version 5.0 of its REDIPlus direct access trading platform-the latest move in an ongoing algorithmic effort originally boosted from the acquisition of Hull Trading in 1998.
The initial step of algorithmic trading-to find otherwise undiscovered pockets of liquidity among multiple trading venues quickly-remains the same across products. The more fragmented and electronic a market is, the more it calls out for algorithmic trading solutions. "The base technology can be used from the equity marketplace to define electronic trading in these spaces," Flatley says. "I think FX is the most interesting because it is truly fragmented, where you have people representing liquidity as independent providers of liquidity, [but] there's tons of liquidity providers for FX and for futures and for options, and for that matter for Treasuries, [though] it's a little bit different market because it's an inter-dealer market. It's fragmented in a different way."
He cites the growth of electronic trading venues like Currenex, HotSpot FX, FXall and FX Connect as signs that the FX market is ripe for the application of automated and quantitative trading techniques. Similar venues, though fewer, are available for bonds, such as TradeWeb and MarketAxess; but new possibilities are emerging.
Paris-based Euronext said last month it is partnering with Borsa Italiana, the Milan stock exchange, to offer to buy MTS, the pan-European bond trading platform. Euronext also owns London-based derivatives exchange Liffe, and is second in size in Europe only to the London Stock Exchange.
The connectivity among the dealers and their customers and these systems, though evolving and growing, already exists. Like other brokers, BofA's ETS offers electronic algorithmic trading services to customers directly or through partnerships with buy-side vendors of order management and other systems. For instance, users of Reuters Station and Reuters 3000 Xtra desktop systems can connect to ETS' algorithmic and block trading tools, following a link the firms unveiled in February.
Despite diversification into new instruments, many brokers still tout equities algorithms. "We actually have a different profile for Lucent than we do for IBM and we have a different profile for Red Hat and Intel," Flatley says. "So we understand how that stock trades the day after Thanksgiving, we understand when the spread is 10 percent beyond the typical mean spread, and what that means, and whether there's an opportunity or whether we should decelerate. We understand what happens on a stock-by-stock basis when the largest print in the day hits the tape for that stock."
Other improvements can be more technical. Data latency, for instance, is a constant worry for managers of algorithmic offerings, particularly for stock trading. "Usually the first mile is the slowest," Flatley says. Thus, electronic pipes or links must be constantly tested with clients.
