Volatility Spurs Aggressive Algo Strategies
Equity traders are no strangers to volatile markets, and in the last several months they have responded similarly to how they have during other turbulent times, by trading more aggressively.
"We see a definite shift in what people choose when you hit the very high volatility days," says Dan Mathisson, managing director and head of Advanced Execution Services, Credit Suisse. When the market gets volatile, people shift from patient algos to aggressive algorithms. He explains, "There is much less use of full day VWAP and our Guerilla algorithm and a shift to Sniper and Blast, our most aggressive tools."
Mathisson adds, "And we see shifts within algos, when people have a choice as to whether they want to be patient or aggressive with a particular algorithm, they choose aggressive."
The reason? Traders are more worried about not getting the trade done than getting it done at a bad price.
[For more on how the volatility is impacting the buy side, see "How Buy-Side Traders Are Adapting to the Volatility"]
"In a non-volatile market, if you're looking to sell 100,000 shares of Intel, for example, a trader will sit on the trade and try to squeeze out that extra penny by trading well. But in a volatile market they are looking for the maximum liquidity and are more concerned with getting the trade done, than saving a few pennies," says Mathisson. "The big fear is that the market plummets and they can't get the thing done."
For Credit Suisse customers that generally means a shift to algos like Blast and Sniper. Blast is the most aggressive in the firm's suite of tools, according to the firm. Blast routes an order simultaneously everywhere; the order goes out to all the exchanges and ECNs that are displaying a particular name. The algorithm sends an Intermarket Sweep Order to all venues with the goal of getting as close as possible to simultaneous arrival.
Mathisson explains, "You're not doing it serially -- if you do it that way, there's a chance that the others will fade once they see the print. Blast doesn't give anyone time to change their positions."
Meanwhile the Credit Suiise Sniper algorithm is almost the opposite. "Imagine a sniper hiding in the bushes pulling the trigger one at a time. The goal is to get price improvement. It's less aggressive than Blast, but still aggressive, and is a good tool to take if you want to get price improvement or size improvement," Mathisson says.
Scott DePetris, chief operating officer of Portware, agrees. "During the period of volatility which has still not gone away, there was a definitive trend toward less durations scheduling strategies." He explains in a similar way to Mathisson that the name of the game was to "Get it done with these constraints." He adds, "Traders are not being passive as they risk market movement."
THE HUMAN TOUCH
One trend that DePetris has noticed is that there has been greater human interaction in trading with algos. "Instead of putting the trade in the black box and letting it go, firms that offer algorithms have created oversight groups to watch the algos and tweak them."
Algorithms behavior is based on historical data, and during volatile markets the norms are not the norm. These oversight groups are looking for anomalies and outliers. They are watching how algorithms are performing in real time.
There is a much more rigorous review process of how the algorithms are behaving, DePetris explains. "Traders and their algo providers are doing daily reviews. They are looking at how the algorithms performed globally and in the United States, they are tweaking the algorithms overnight," he says.
DePetris offers an example, "If you're given something pre-market, and historically 8 percent has gotten done on the open, but yesterday it was 12 percent and today it was 14 percent, and I don't do anything, I will participate at 8 percent. It's better to tweak the algorithms and move it to 11 percent to get a better clip done." These are the kinds of things being done during the oversight process.
Firms are also setting up alerts that look for anomalies and change algos for clients. DePetris says, "You have this new sell side human interaction. The buy side has more tools at their disposal, but there is more customization, and alerting buy-siders. They are looking for outliers, and the stocks they need to be concerned about today." DePetris explains that this type of thing used to happen in release cycles, but now it's happening on the fly.
The volatility has opened up a much more detailed dialogue between brokers and buy side clients, he says. The greater complexity comes more discussion with the sell side and is giving them an opportunity to prove their value.
DePetris says algos were commoditized at one point, but now there is a value added benefit of consultation by the sell side that is being integrated into the workflow process.
Kerry Massaro Bowbliss is the former editor of Advanced Trading.
Tweet This