Momtchil Pojarliev
Traders Magazine Online News

Some Like It Hedged

BNP Asset Management's Pojarliev discusses a variety of options to address foreign currency exposures. Although there is no single best-practice solution for addressing foreign currency exposures, institutional investors have three main choices, he says.

Traders Poll

Amid changes in builder, do you think the CAT project will be completed by 2020?

Free Site Registration

May 26, 2005

Could More Volatility Sabotage Algorithms?

By Peter Chapman

The surge in algorithmic trading will likely come to an end if the market becomes more volatile. That's what some top brokerage execs said at a trading and technology conference in New York City. Deustche Bank's Greg Sharenow and Dan Mathisson of Credit Suisse First Boston are expecting capital commitment to replace algorithmic trades if volatility levels rise.

"When volatility picks up," Sharenow told the TradeTech USA conference, "algorithms as they are today will be less effective." Sharenow said he would prefer more volatility. Volatility, as measured by the Chicago Board Options Exchange's Volatility Index, or VIX, is at its lowest level in eight years. It hit 15.38 recently, a level not seen since 1996. Low levels of volatility mean a more predictable trading environment. That permits computer-generated trading strategies to perform more effectively. With a choppier trading environment, buyside traders are more likely to pay up for brokers' capital rather than risk bad trades.

Mathisson agreed that higher levels of volatility will make his firm's algorithms less desirable. He suggested, however, that market volatility may have become permanently dampened. "Part of the reason we are seeing low volatility," Mathisson said, "is because of the algorithms." An algorithmic trade sends out only a few hundred shares each time into the market, Mathisson noted. Thus, the strategy is less likely to push stock prices dramatically up or down. CSFB says it processes about half its 200 million shares daily via algorithms.