The algorithm is not the enemy of the trader. At least that's what one trading veteran says.
"I really don't think that my brethren out there on the buyside should fear algorithms," says Larry Billimeck, a senior equity trader at Nicholas-Applegate Capital Management, a $14-billion subsidiary of Allianz Global Investors. Effective traders can't let these mathematical marvels intimidate them.
"If you're afraid of them, you'll be left behind." Billimeck points out that this smart-trading tool has its place. Still, he doubts algorithms will replace skilled buyside traders.
Part of that logic rests in the fact that roughly 10 to 15 percent of the firm's trading is done via algorithmic machines. Indeed, that percentage has declined a bit, he adds. Billimeck says that algorithmic trading at his firm is about equal to ECN usage. That means essentially 20 percent to 30 percent of the firm's flow is executed electronically. "I really don't see those numbers going higher," he adds.
"Our PMs realize that we have a lot of caveats as it relates to algorithmic trades," Billimeck says. Indeed, liquidity and readable trading patterns are concerns. Lehman Brothers, Morgan Stanley and Credit Suisse First Boston are his three main machines of choice, he says.
These trading formulas are invaluable, Billimeck explains. That's because the desk can use them to control the order. "I want to be able to tweak the order on the fly," he adds.
Then buyside traders still have the opportunity to add value in a low-return environment, Billimeck says. "I see our role today as even more important," he adds.
Just as the markets have evolved, along with technological advances, so has the role of the buyside trader, he says.
To remain an effective desk, he adds, the firm must watch broker performance. The firm uses three trade-cost analysis vendors Plexus, Elkins-McSherry & ITG in its broker ranking. In fact, broker performance is part of its pre-trade analysis and may help to dictate where an order goes, he says.
Through some internal tweaking of software, the firm is able to put each broker's TCA numbers side-by-side for comparison. "Even if a PM says to pay a certain broker, we still look at their numbers to date." Billimeck says. But always, Billimeck says, it is a people business.
The firm still must balance its relationships on both the research and trading side. The annual research list is updated quarterly, which takes on more importance as commission rates and trading volume decrease and the amount of "free business" becomes scarcer. That's primarily because soft dollar "chits" remain a constant: It takes longer to pay those bills.
To counteract that, the firm has carved its core list of brokers down to 80, which may still sound like a lot, but that number was about 320 before the bubble burst, Billimeck says. "We've got a smaller core, but we haven't noticed any cut back on their service," he adds. "We still maintain good relations with our brokers."
Billimeck says that despite a rise in algorithmic trading, there is a willingness to trade blocks. Some of that may have to do with the fact that "we are not in an environment of momentum," when a trader might be more concerned about being "steamrolled."
"When there's low volatility, it's easier to trade blocks," Billimeck says. Since few of its orders require capital commitment, it's common to have an order with a broker, while the desk also has an order in Liquidnet, he adds. Back in the fat 1990s, when firms were risking capital, that would have been taboo. "I think everyone is comfortable with that now," he adds.
