Add your name to the FMW e-mail list:

As with many an occurrence in the marketplace that persists and eventually deepens to a trend, so have electronic trading strategies formed their own niche within the realm of trading. While the popularity of algorithms continues to increase, the larger brokerage houses are buying the electronic trading strategies vendors in order to keep their market share of the client base. However, although algorithmic trades are becoming more common, they still account for only approximately 5% of total buy-side trades, leaving the flesh-and-blood traders with a firm standing in the marketplace as it currently functions. Traders are, however, left with a plethora of questions surrounding how to differentiate between algorithmic trading strategies in order to make informed and appropriate trade decisions.

When considering performance, it’s difficult to compare trade strategies on an across-the-board basis due to the unique nature the processes allow each trade. With the development and increased use of algorithmic trades, they have become a process that is increasingly more commoditized and, as a result, it is sending brokerage houses back to the basics as they are left with little more than their reputations to maintain the edge on their client relationships. Clients are looking to the brokers that will offer the best perks for their business, and on the client’s terms.  Michael Boyd, head of institutional sales and trading at CJM Securities, comments that “funds only want the execution aspect of the relationship – they don’t need the whole shebang anymore, and as a result they only want to pay for execution. They want lower rates, and the funds are really the ones calling the shots right now.” Gavin Little-Gill, TowerGroup analyst, claims “the market leaders in 2005 will be those able to execute…and gain mindshare from a fickle buy-side client base,” further extending the idea that best execution and existing relationships serve as the basis for buy-side traders when considering which brokerage houses they will use to execute their algorithmic trade orders.

Still, with the vast majority of trades that can not be sent through a fully automated system, traders need to be able to best direct those trades that need to be hand-held through the process, whether they be entirely manual or partially automated and monitored. David Cushing, managing director at Lehman Brothers, sees the possibility of advancement, however, with the expectation that in the future, traders will be able to string algorithmic trades so that if the conditions of one are not met within a specified period of time, the order is automatically routed to an alternate algorithmic trading strategy to complete the transaction. Known as conditional auto trading, “it is how algorithmic trading will evolve to challenge traditional trading techniques,” says Cushing. “The most rapidly diminishing request is that of the easy orders sent for manual processing.”

Measuring the performance of individual algorithmic trades is not so much the quandary, as they are passably measured against one or more benchmarks; rather, it is the inability to effectively compare a multitude of different algorithmic trading strategies to be able to better choose from among them the appropriate method for future trades. With the customizability of each trade, one is hard pressed to locate similar trades across two institutions at the same time and under the same market conditions. However, it is for this very reason that algorithmic trades are so effective, especially for the hedge funds that are known for their incessant desire for customizable products and services.

Cushing emphasizes the point that “performance measurement is crucial. Good performance is the ante to get into the game, and most techniques that are used to measure performance entail comparing the average executed price to one or more benchmarking prices.” Ultimately, the success rate of individual trades is adequately measurable and, because of the methodological documentation of the trade process, algorithmic trades also offer a solution to a regulatory matter; each trade’s thorough documentation provides a trail to support the transaction, unlike manual trades carried out on the floor.

As algorithmic trades become more complex and capable, they also will become more commonplace in spite of the inability to draw cross-comparisons between trade strategies. Brokers will find ways to show clients the advantages they have over other brokerage houses. Just as hedge funds are sought after investment vehicles due in part to their unique and inherently secretive nature, so will algorithms become more the standard because of their inimitability. That is, after all, the very quality that makes them such a useful method.