Anthony Malakian: What I Learned at Sifma 2011

My dad always told me that every day that I learn something new is a good day, the moral being that I should broaden my horizons even on the most mundane days. Days spent at Sifma are not mundane; therefore, I should have learned a lot—and I did.

But first, I’d like to contradict myself. On the biggest theme of the year—regulation—I didn’t learn much at all. Or, perhaps, I learned a ton. I liken it to Brad Pitt in the movie Seven right after he shoots Kevin Spacey—staggering around not sure of what just happened.

The problem is that lots of people say lots of different things when it comes to the buy side and how firms are gearing up for the regulatory overhaul. Opinions varied wildly during the numerous discussions I had at the Sifma event on the subject. So here’s what I choose to believe until I am made to think otherwise: In Europe, the buy side is further along in its technology planning as a result of new regulations, whereas in the US, buy-side firms are taking more of a wait-and-see approach.

It is logical for firms to sit back and wait—plan strategically, but hesitate before pulling any triggers—since there is a belief that as the implementation of the Dodd–Frank Act gets drawn out (weren’t we supposed to have some answers this month?) the bite of some of its proposals will be lessened.

While new regulations have imbued the industry’s collective conscience with a sense of anxiety, it doesn’t strike me as a full-scale panic.

Sneak Peek
But enough about regulation—there will be plenty of time for that in the coming months and years ahead. Let me offer you a peek into some of the more interesting conversations I had from this year’s Sifma event—a sort of “Best of Sifma,” if you will.

Topping the list is the dying lifecycle of an algorithm. According to Kx Systems’ Simon Garland, in the high-frequency trading world, the typical lifespan of an algorithm before it gets gamed has fallen from a minimum of seven weeks down to just two weeks in many instances.

“We were speaking with a high-frequency trader a couple of days ago and he said that the typical life of an algorithm, these days, is about two weeks, because as soon as the people are on to you, then of course it’s over and you have to do something else,” he says. “Two weeks is the minimum now, because that’s someone who’s really good. So you can’t send your good people on holiday because they’ve got to have a new algorithm in two weeks.”

I also had a chance to sit down with Portware’s Eric Goldberg. We discussed the lack of automation in the fixed-income market, and how long it will take for the seas to change. The gist was this: It ain’t coming this year.

“I think we’ll start seeing it in 2012, but it will really start to happen in 2013,” Goldberg says. Nobody has the resources right now to internally develop it, as everyone has their IT staff working on Dodd–Frank, and, he says, it would have to be a large bank that doesn’t have a big business—that has nothing to lose—that will kick it all off. “It’s going to be like Credit Suisse was in equities with Advanced Execution Services (AES),” Goldberg says. “It is going to catch on so quickly that everyone is going to start jumping on.”

I also enjoyed chatting with FinAnalytica’s David Merrill, who told me the profile of the chief risk officer is evolving, even if we are at the very beginning of this change. “We’ll start to see the profile—the persona—of the chief risk officer change,” he says. “People talk about this, but I think it stops too short: If you want to have effective and influential risk management practices in a firm, they must be embraced by the head of trading, the chief investment officer, the CEO, and the head of sales—the key players in the firm.”

Finally, I learned that it’s never a good idea to stay out until 5 a.m. on a work night (day). Wait, shouldn’t I have already known that from my college days? I guess some lessons just can’t be learned after all.