The buyside must become more efficient or else. Blame this tougher environment on decimalization. Blame it on the increasing complexity of best execution.
For Tim Christiansen, that means he can't put all his execution eggs in one basket. No single ECN, liquidity pool or algorithm is enough to work an order. "We can't afford to miss some pockets of liquidity," he says.
Christiansen, one of three equity traders at Sawgrass Asset Management, has been with the firm for two-and-a-half years. Sawgrass is a growth equity and fixed-income money manager with $1 billion in assets. The traders operate out of the firm's Jacksonville, Florida, headquarters.
The Sawgrass trading desk pursues best execution with a variety of trading tools and metrics, including pre-trade cost analytics. The traders use pre-trade analytics to get a quick look at expected trading costs across a spectrum of positions, trade sizes, and names. "We don't use program trades, so the pre-trade analytics help us to prioritize how much we want to load up into crossing networks and how much we want to work ourselves in the ECNs," says Christiansen.
The trading desk uses Instinet's intraday crosses, and ITG's TriAct and intraday crosses. The desk also uses Liquidnet, which reads the order management system. "We also look at broker flow indications to try to find a natural," says Christiansen. "If there's a match with something we have to do, we'll immediately prioritize that." He emphasizes that trading in the current ECN environment requires understanding the subtle differences in the functionalities of various ECNs and ECN aggregators.
Christiansen and his colleagues tend to work the more difficult names through ECNs. In some cases, they pass them on to Themis Trading or Pulse Trading, two agency-only brokerages that specialize in ECN trading. "We view them as an extension of our desk," says Christiansen.
The money manager is also beginning to focus carefully on algorithmic trading tools. "It doesn't do any good to load an order into an algorithm and have wrong expectations about how the algorithm will act under different scenarios," Christiansen says. Traders need to understand clearly whether the algorithm will conform to their expectations across market conditions.
Christiansen also notes that "there's still room for improvement in terms of algorithms trying to discover hidden pockets of liquidity." A trader can easily accumulate a position through a lot of 200-share prints and "lessen the chance of leaving a footprint." But there's an opportunity cost if that potential hidden liquidity disappears before the trader has finished buying, according to Christiansen. With the right tools a trader can find bigger pieces of liquidity.
Sawgrass recently began working with Miletus Trading, a brokerage and algorithmic trading firm, to custom-build algorithmic trading strategies. There are four algorithms in the works. "When you're on the desk working a small number of orders, you have the ability to pick apart what's happening on ECNs," Christiansen says. "You hunt for hidden discretion and try to be smart about going after someone's reserve order." The algorithms being developed will try to replicate the way that Sawgrass traders think.
"The goal is that we can eventually send orders off into those algorithms and we'll understand the logic of how they're working," says Christiansen. It will then be up to the algorithm to probe that liquidity in a way that's smart and that "hopefully won't hurt the overall picture," he adds.