Commentary

Brijesh Malkan
Traders Magazine Online News

Solving the Last Mile Problem in Investment Research

One executive takes a forward look at how will the research value proposition change over the short to medium term, and what are the products and strategies managers will turn to?

Traders Poll

Do you think it's a good idea to conduct an access fee pilot to assess the pricing models used by many trading venues?

Yes
No
Should have had a pilot program a long time ago.

Free Site Registration

August 1, 2010

Deutsche Bank Shakes up Algos

Deutsche Bank Shakes up Algos

By By James Ramage

Jose Marques, Deutsche Bank's new head of equity electronic trading, wants the firm to build trading tools that take into account what algorithmic trading will look like tomorrow, how it will evolve.

Today's buyside has been using algos that think like traders. But Marques says clients need to have algos that think like automated traders-like machines.

During his first six months on the job at Deutsche Bank, Marques' guiding directive has been to transform the firm's algos into trading tools that more resemble the ones high-frequency traders use. The first algorithm Deutsche Bank rolled off of its assembly line this spring, a dark pool aggregator called SuperX, reflected this thinking.

"At Deutsche Bank, this is the key problem we're trying to solve in electronic trading," Marques said. "We're taking the best tools that statistical and low-latency traders use for providing liquidity and turning them into buyside tools for taking liquidity."

Marques joined Deutsche Bank from Credit Suisse, where he ran the firm's CrossFinder dark pool. He is a 15-year veteran who has also held senior electronic trading roles at Morgan Stanley and Goldman Sachs.

As the marketplace has changed, today's liquidity providers have evolved, he said. To his thinking, these aren't the same brokers and market makers with whom buyside traders have golfed or shared information about their kids. Often, they're quantitative statistical machines that view market complexity differently than a human mind does.

"They're looking at it cross-sectionally and in milliseconds," Marques said. "They're looking for statistical patterns in a very multidimensional, complex and dynamic data space."

But amid these changes, the buyside trader faces the same challenges. He still has a blotter full of large, high-impact orders that must be executed efficiently, without dislocating or impacting the market, Marques said. "But now his liquidity providers are a totally foreign thing, and there's a real mismatch between the two," he added.

SuperX uses a statistical ranking model that analyzes the execution traits of each venue. It takes an order and determines the best place to trade, given the trader's constraints and where the algo has found the least leakage in the stock. And if the order is creating too much market impact, the algo automatically backs off of the trade, Marques said.

But Marques' approach to algorithms goes beyond one new algo. He wants to ingrain similar logic into all of the firm's algos-present and future. This will entail some retrofitting, he said. Currently, his group is working on rebuilding its liquidity-seeking algo, Stealth, with this new technology. Marques declined to say when the retooled algo would be rolled out.

Furthermore, Deutsche Bank provides direct market access to around 80 markets worldwide. And its algos access roughly 66 of those markets, Marques said. Ultimately, the goal is to align Deutsche Bank's market-access and algo products around the globe with those in the U.S. and Europe.

"This is important for the global-oriented clients," Marques said. "We've made a push for our algo behavior to be consistent and take into account the idiosyncrasies of local markets, to have a true global platform."

Marques said he also wants to make Autobahn Equity, the firm's electronic trading product, a better-integrated business.

To do this, Autobahn's entire staff now reports to him. That includes IT, trading, sales and product development.