Industry observers agree that over the past 12 months, algorithmic trading has taken on a new identity, maturing into more adaptive strategies with better routing features. Algorithms are no longer just another innovative trading tool available to the marketplace. Algorithmic trading has become a necessity for savvy traders seeking to gain access to dark liquidity pools. It is no longer just about volume-weighted average price (VWAP).
Looking back on 2007, algorithmic trading has become a vital tool in any trader’s electronic tool chest, says Brad Bailey, senior analyst with the Boston-based research firm Aite Group, explaining that the rapid change in the U.S. equity market structure has driven and added complexity to how the average institutional trader gets a trade done. “You have to think about where all of this liquidity is,” he says. “That drove a lot of people to look at aggregation tools, such as dark algorithms, and algos looking to aggregate this liquidity.” That’s been one of the big themes in 2007–trying to gain access to the dark liquidity and access to this fragmented market structure.
Along with this has come the migration away from simple benchmark algorithms, such as VWAP or arrival price, to algorithms that provide more flexibility with specific tactics used to target those benchmarks, said Carl Carrie, global head of algorithmic products and head of Neovest at JPMorgan. “It’s no longer good enough just to say, ‘Trade according to arrival as a benchmark,'” Carrie explains. “A trader typically wants to have a lot more control over how that’s done. With more fragmentation, there’s more granularity and richer texture to trading then there ever was before.”
Another major trend in 2007 was the push toward using algorithms in different product sets, such as foreign exchange, says Tom Price, senior analyst of securities and capital markets at the Needham, Mass.-based consulting firm TowerGroup. “What I think is the trend or has been the trend is the migration of successful equity strategies into alternate asset classes such as foreign exchange,” he says, noting that the equity model is already fairly mature.
When discussing algorithmic trading trends in 2007, one term jumped out over all the rest, and it isn’t a four-letter acronym–it’s customization. And in 2008, most agree, it will continue to prevail as the industry’s favorite buzzword. The customization of algorithms is far from becoming yesterday’s news anytime soon.
“Customization isn’t even a trend–it’s the driving force now, and fewer and fewer clients want a one-size-fits-all solution,” says Carrie. “Nearly every client we talk to wants variation on what we offered them the previous day. And even within the same client institution, there are traders that trade differently, and they want special parameters exposed. Very often traders want an algorithm to behave differently under different conditions, and when they do that often enough, it becomes a customized algorithm for them.”
Over the last year, the sellside has been developing its strategy around customization, agrees Adam Sussman, senior analyst at the Westborough, Mass.-based TABB Group, noting that there were one or two firms that formally launched marketing and sales efforts around the concept in 2007.
“I think we’re really going to see [customization] come to the forefront in ’08,” he says. “It’s going to be one of the top trends moving into ’08, and there are a couple of different reasons for that. One reason is that it’s difficult for the sellside to get the ear of the buyside these days, as more and more trading becomes electronic. Customization has been a pretty powerful tool for a number of banks to re-engage the client.”
One aspect of that will be the shift to clients self-customizing their own algorithms, says Carrie. “Another emerging trend will be products that tie in algorithms much closer to the alpha-generating process and the emergence of algorithmic management systems to improve workflow for trading,” he explains.
Aite’s Bailey asserts that the industry is still only in the early days of customization. “People are going to demand more and more out of the algorithms,” he says. “They’re going to demand more metrics around the trading and more integration of pre-trade and post-trade analysis.”
