Bloomberg Tradebook recently expanded its suite of algorithms that give buyside traders working orders greater control over how their limit orders are posted at different execution venues. What’s driving the need for more smart-order-placement strategies, according to Tradebook, is market fragmentation.
Tradebook’s B-smart, a smart-order venue-placement algo for single stocks, rolled out last October. Now, Tradebook has added REACT, a limit-price placement strategy that mimics how traders seek to respond to short-term price trends and executions patterns.
“In the old days, traders would be in one venue at one price,” said Gary Stone, a Tradebook executive. “For limit orders, they’ll now smart-order-quote and split the order into different venues to minimize the risk of missing liquidity due to market fragmentation.” Smart posting is increasingly seen as the counterpart to smart order routing, which helps traders take liquidity from the market efficiently.
The majority of Tradebook’s non-marketable limit orders are already using smart posting to split orders and pack them off to different venues, Stone said. However, some buyside traders using direct-market-access tools wanted the strategy to have more of a human touch, especially when the market was coming toward the trader, noted Sabrina Rovelli, a colleague of Stone’s. In response to that demand, Tradebook expanded the B-smart options to include REACT.
REACT enables customers to minimize the likelihood of “being traded around and maximize price improvement opportunities,” Stone said. “It’s essentially a laddering strategy. If I’m a 23 bid and am getting hit hard, should I stay at that price or drop down?” The engine behind REACT, he said, “looks at how the stock is trading, the marketplace, the short-term price trends, and sees if the market is coming down. Instead of being at 23, a trader might drop to a 20 bid, wait for the market to come down, take some stock, and then reassess again.” He noted that the strategy isn’t like pegging, “where someone else is making the decision for the trader.”
REACT’s decisions, Rovelli said, are based on a survey of the volume and volatility of the stock being worked, the size of prints in a particular venue versus what’s displayed, the bid-offer spread, and the speed of executions to try to anticipate the aggression level of traders on the other side of the market. B-smart’s posting tactics come in three other “modes,” from passive to aggressive.
Although REACT launched only a month ago, Rovelli said, most clients using the B-smart passive strategy have already shifted to REACT. “For a lot of our traders, passive wasn’t passive enough,” she said. B-smart’s existing passive strategy focuses on posting, but buys stock if there’s size available. REACT adjusts its pricing decisions to market conditions.
Algorithmic trading is clearly on the rise. But while some traders want to be able to push orders into algos, Rovelli observed, many still work orders themselves or turn to DMA tools when the market is volatile. At those times, she said, “they lose faith in the execution levels of algos automating the trading for them.”
Bloomberg isn’t the only broker focusing on how non-marketable limit orders are placed into the market. Many big brokers, having developed smart order routing for a fast-moving fragmented market, also focus on order placement as part of their DMA and algorithmic offerings.
The smart placement of limit orders based on execution-related and other statistical factors may get even more attention in the coming months. At the Securities and Financial Markets Association market structure conference in May, Erik Sirri, the Securities and Exchange Commission’s director of the Division of Trading and Markets, raised the issue of best execution for non-marketable limit orders. He discussed several venue-related factors that brokers should consider in their “order routing calculus.”
Bloomberg Tradebook’s Rovelli said the broker expected the B-smart order placement algos to be used mainly for illiquid stocks, but that traders also turned to smart posting “in large liquid names where they no longer need to pay the penny spread.” B-smart algos execute in displayed venues and dark pools.
Tradebook’s AlphaPro suite of benchmark algo strategies also uses the REACT technology to send orders into the market based on shifting pricing trends and market conditions. AlphPro includes volume-weighted-average-price, time-weighted-average-price, go-along and arrival price algos, among others.
