Knight Capital Group on Monday unveiled a new version of its Sumo algorithm that aims to help institutions camouflage their trading activity so they can safely interact with high-frequency traders.
The algo has been redesigned to use real-time and historical liquidity measures to give priority to large orders and avoid information leakage.
EdgeTrade, which Knight acquired in 2007, first released Sumo more than half a decade ago. Joseph Wald, managing director at Knight, said the algo has been completely overhauled to reflect the changes in the market over the past few years.
“This is a radical departure from anything that Sumo had done in the past,” Wald said. “It really is a different approach.”
The revamped Sumo uses a size-biased routing strategy to find larger blocks of liquidity at fewer venues. Rather than simply slicing and spreading orders across multiple venues over time, the algo seeks the largest order at each chosen venue so it can complete as much of the trade as possible while minimizing information leakage.
Wald said the algo looks to interact with high-frequency flow so long as that flow is benign. He said there are a number of HFT strategies that provide important liquidity that can be helpful in filling large orders.
“One of the things that the strategy does extremely well is tell the difference between what’s good high-frequency and what is high-frequency that we want to avoid,” Wald said.
The strategy is similar to Knight’s Oasis algo, but Oasis is designed to trade small-cap and mid-cap stocks, while Sumo is designed to handle larger names.
Institutions can access Sumo through the Knight Direct broker-neutral electronic trading platform and through various order management systems and trading system providers.
