One of the most significant trends on Wall Street over the past few years has been the empowerment of the buy side, as emerging technologies have enabled portfolio managers and their in-house traders to handle more and more order flow without phoning a sell-side counterpart. But, as everyone knows, with new power comes new responsibility.
And while the buy side technically can handle a portion of its order flow, the question then becomes, how well is it doing so? Enter transaction cost analysis (TCA) and the burgeoning business of predicting what a trade should cost and where to send it (pre-trade analytics), and then examining how well that strategy was implemented by comparing the results to various benchmarks (post-trade analysis).
"As the buy side takes over more of the executions, are the traders on the buy side outperforming the sell-side traders?" asks Adam Sussman, a senior consultant with Westborough, Mass.-based TABB Group and author of a recent report on TCA. "Are the buy-side traders outperforming the broker algorithms? The buy side needs to figure out the best execution strategies for reducing slippage."
Douglas Cote, senior portfolio manager with Hartford, Conn.-based ING Investment Management, describes TCA as a highly integrated process that starts with acquiring a solid understanding of the day's highest cost and/or "most difficult" trades, then developing a strategy to get those trades executed with minimal market impact. At ING, Cote - who trades baskets of stocks - says he sends the most difficult trades to his block desk and the remainder of the orders to one of the firm's algorithmic providers. According to Cote, the firm works with New York-based agency brokers ITG and Miletus, as well as a third firm that he declines to name. Additionally, ING works with one sell-side broker for algorithms that Cote also declines to name.
Customization Is Key
The key to keeping transaction costs down when working with algorithmic providers is using those that can customize a strategy for each trade's specific needs, says Cote. "All these providers have these tools, and everyone says they are a commodity," he relates. But, "They are not a commodity because they customize these exception strategies for me, to match my portfolio objectives," Cote continues. "We are trying to land on the efficient trading frontier - the sweet spot where the minimum impact cost meets your risk tolerance." TCA providers then determine the success or failure of the chosen trading strategy, he adds.
Cote notes that he uses ITG for pre- and post-trade analytics. ING, he explains, has ITG software on its desktops in the form of a Microsoft Excel add-in. "I send in trades, and it automatically calculates them on the pre-trade and on the post-trade," Cote says, adding, "It's a Web-based product."
From there, ING works closely with the agency broker of choice to define an optimal trading strategy. "So, based upon that algorithmic provider's capabilities, we will say, 'OK, it will be some version of an implementation shortfall strategy.' And, at the end of the day, we will do a detailed post-trade analysis and see how well they compare against the three benchmarks we use," Cote says.
Those three benchmarks involve looking at how well traders did against the open - the rival price in sync with the implementation shortfall strategy - and how well they did against the close, as well as evaluating the overnight slippage. "It's not just transaction-cost analytics - it's more an integrated process of pre-trade, execution strategy and then post-trade analytics on a day-to-day basis," asserts Cote.