RBC Capital Markets is rolling out its first new product since hiring a team of high-frequency experts and staff last year. The new tool is a smart order-routing technology designed to prevent institutional orders from getting beaten to liquidity by rapid-fire traders.
The smart order-routing technology, named THOR, after the Norse god of thunder, looks to nullify the time advantage HFTs have over others. It does this by adjusting the routing speed of orders, so that each order in a stock reaches its destination at the same time as the others. The advantage to this strategy, at least on paper, is that multiple executions occur before HFTs are able to cancel their posted orders and run ahead.
THOR is said to be in use in the U.S. and is being beta tested in Canada, according to sources. The product’s status could not be confirmed because RBC Capital Markets declined to comment for this story. However, RBC discussed the new product in two general-circulation publications: The Globe and Mail of Toronto and WSJ.com.
Order routers work in two ways. First, they can work sequentially—by sending orders first to the trading venue providing the best rebate, and then to the one with the second best and so on. They can also “spray orders,” as THOR does, by sending orders to multiple exchanges at the same time, regardless of rebate.
Several buyside traders gave the concept a thumbs-up, including Kelly Reynolds, director of trading at Toronto-based Hillsdale Investment Management, which manages $C500 million in equities. Reynolds’ desk has been beta testing THOR for two months. So far, she’s a believer in the product. “It’s real-time latency adjustment for orders,” Reynolds said. “For what it does, I am very satisfied.”
She said THOR’s smart order-routing technology provides an alternative to other brokers that route their orders based on economics—they send orders first to the exchanges or venues with the best rebates and then to others. This type of order routing often gets the attention of HFTs looking for trading patterns. RBC’s spray routing product looks to avoid a trading footprint. After all the orders are executed, THOR cancels all of the outstanding bids or offers immediately.
The problem, as RBC and firms like Hillsdale see it, is that HFTs’ trading strategies do not help investors. Using hyper-fast computers and trading strategies, HFTs have long been accused of sniffing out large institutional orders and executing small trades ahead of them, exploiting price discrepancies over fractions of a second that might occur between exchanges. The HFTs then pull the remainder of their orders, taking liquidity and the best execution price with them.
Although RBC declined interview requests, that didn’t stop others from discussing THOR.
Dennis Fox, head trader at Birmingham, Mich.-based Munder Capital Management, also thinks THOR is a good idea for combating HFT strategies. Although he doesn’t use the product himself, Fox said the RBC program trading desk does use it for his programs, and he’s a fan. “It works well for my orders,” Fox said, because it helps hide his trading footprint.
According to RBC’s Web site, THOR works like this: It measures the time it takes for orders to reach all exchanges and then slows each order’s speed to the pace of the slowest slice of the overall order. So, all child orders of the parent order arrive at all their exchanges or trading venues nearly at the same time—or, as RBC claims, within a few hundred microseconds.
THOR is apparently part of RBC’s bid to grow its electronic desk and offerings, which began in 2009. The firm is looking to earn a bigger slice of the electronic trading revenue pie, Brad Katsuyama, head of electronic sales and trading at RBC Capital Markets, told Traders Magazine last June.
RBC said it hired HFT professionals, including infrastructure specialists, product developers and software engineers—some of whom likely had a hand in developing THOR. Given their background, these professionals understand the intricacies and tactics of HFTs and can help counter them. It’s similar to casinos hiring card counters to spot other gamblers who count cards.
RBC told Traders Magazine last year that its goal is to give clients more ways to pay for its research and, ultimately, grow its commission levels with clients.
And RBC isn’t alone in trying to quash the effects of HFT strategies. Credit Suisse recently announced it would debut in March a new ECN, LightPool, that will filter out users of short-term trading strategies that disadvantage long-term investors. Those most likely to be filtered out are HFTs.
Credit Suisse will use a proprietary formula to assess the intent of those who use LightPool. Once a customer logs 500 trades, the broker will measure the short-term alpha-a calculation of excess return—from every fill. Clients found to have short-term alpha will be labeled “opportunistic” and expelled from the ECN.
However, the high-frequency crowd isn’t rewriting code or breaking a sweat just yet. Manoj Narang, chief executive at Tradeworx, a Red Bank, N.J.-based HFT, said firms like RBC are playing off the buyside’s misconception of HFTs.
“It [THOR] is a semi-clever idea, but I think it is mostly a marketing gimmick,” Narang said. “I doubt it will add real value, because it is based on a faulty premise—namely, that firms that post orders on the exchanges engage in manipulative practices to inflate the prices paid by buyside investors.”
Last April, Narang filed a brief with the Securities and Exchange Commission, disputing the buyside’s assertion that HFTs engage in manipulative practices. In it, he said, Tradeworx provided empirical evidence showing that the prices posted by liquidity providers on the national exchanges are fair.
If HFTs were engaging in price manipulation, he said, THOR—which in his opinion more closely resembles an algorithm—is more likely to hurt investors than help them.
“The reason is, a slowed-down VWAP algo will be unable to capitalize on the same fast-moving microstructure signals that HFTs trade on,” Narang said. “That will decrease the performance of the algorithm. A better approach is to run the algorithm on a world-class market-access infrastructure with super-fast feeds.”
One knowledgeable source agreed with Narang that there is a bit of marketing spin to the new product, but said that if the orders can actually hit the destinations simultaneously with zero latency, then RBC’s THOR could be something special. Still, this source wondered how the product could accurately calibrate the changing latency of destinations throughout the day.
Gimmick or not, Cheryl Cargie, head trader at Ariel Capital Management in Chicago, said THOR sounds promising if it lives up to its billing.
“On the surface, this [THOR] sounds really good,” Cargie said. “The buyside will try to do business with whatever it can if it minimizes the likelihood we’ll be picked off.”
