“We’ve had a pretty good crisis,” says Barry Bausano, smiling as he surveys Deutsche Bank AG’s sprawling 35,000-square-foot equities trading floor in downtown Manhattan. Behind him, on bookshelves in his glass-walled office, sit two copies of “Too Big to Fail,” along with “Liar’s Poker” and “Triumph of the Optimists,” which tracks 101 years of worldwide investment returns.
Bausano, 47, who heads the Frankfurt-based bank’s equities division in the Americas, has reasons to be upbeat, Bloomberg Markets magazine reports in its April issue. Deutsche Bank got the best stock prices for institutional clients both globally and in North America during the 12 months ended on Sept. 30, 2010, according to data compiled for Bloomberg Markets’ worldwide ranking of equities brokers.
Germany’s biggest bank displaced Goldman Sachs Group Inc., which had swept all four categories in the previous ranking. This time, New York-based Goldman Sachs was ninth in the world, fifth in Europe and third in Asia. The data were compiled by Ancerno Ltd., a spinoff of brokerage Abel/Noser Corp. that measures transaction costs for $7.5 trillion of trades in more than 70 countries every year.
While U.S. banks retrenched, merged, took government bailout money and -- in the case of Lehman Brothers Holdings Inc. -- declared bankruptcy during the financial meltdown of 2008 and 2009, Deutsche Bank overhauled its electronic stock trading and analytics.
It hired computer trading veterans, including Jose Marques from Credit Suisse Group AG, and rolled out algorithms that customers can adapt to their needs.
One program, called SuperX, responds to instructions about how aggressive to be in filling an order while tapping up to 20 dark pools -- private trading venues that don’t display bids and offers in advance. The bank also boosted the historical and real-time data analytics that feed its algorithms and expanded trading in its own dark pool, called DBA, Bausano says.
In a world obsessed with speed, stealth and sniffing out the best platform for a trade at any given millisecond, customers are looking for brokers with strong electronic offerings, says Peter Weiler, executive vice president of global sales at New York-based Abel/Noser.
“Deutsche Bank and the other top brokers have invested a tremendous amount in their electronic execution, and they early on recognized the need to service clients globally,” Weiler says. “Deutsche Bank has built a really comprehensive offering.”
Less Volume, Volatility
Since the beginning of 2010, brokers have had to navigate a marketplace with lower average volume and less volatility than during the crisis, which erased $37 trillion from global equity markets from the peak in October 2007 to the low in March 2009. While a rough, plunging market was good for the biggest brokers, a calmer environment showcased the ability of specialized firms such as Liquidnet Holdings Inc., a New York-based operator of dark-pool platforms. It was No. 4 in the global ranking.
An average of 8.4 billion shares a day changed hands in U.S. markets in 2010, down from 9.8 billion in 2009 and 8.8 billion the previous year, according to data from Credit Suisse Securities. The Chicago Board Options Exchange Volatility Index, a gauge of investor uncertainty, averaged 22.6 in 2010, dropping from 31.5 in 2009. Volatility is expected to be even lower this year: The VIX averaged 17.37 through Feb. 28.
“In 2008, asset managers and their clients were getting crushed and there was so much volatility, which brokers benefited from,” Weiler says. “Two years later, you have a situation where people are trading less, money flows are going out of equities and everything is moving in lock step in the market.”
Hunt for Liquidity
This year’s best performers found ways to piece together the liquidity investors needed even with trading splintered across public exchanges, dark pools and private platforms.
“We’re looking for brokers that can adapt to market conditions as liquidity becomes fragmented,” says Ryan Larson, head of U.S. equity trading at Toronto-based RBC Global Asset Management Inc. His firm, which manages about $200 billion, uses about 60 securities firms, including Deutsche Bank, to execute its trades.
Larson says it’s crucial for brokers to offer a mix of electronic and traditional strategies to get the best trades without causing speculators to jump in and exacerbate price moves.
“The ability to trade stock with the least amount of price impact is first and foremost,” he says.
Larson says smaller brokers that handle fewer orders than big banks are important in specialized areas such as trading financial stocks. These companies may give individual orders more attention than a larger firm.
“Some banks may not hand-hold my trade in the same way,” he says. He singles out Liquidnet and Investment Technology Group Inc. for creating what he says are good venues to trade blocks of stocks while reducing the market impact and costs.
One big change in the post-financial-crisis world: Clients look at whether a firm will be around for the long haul. Before Lehman declared bankruptcy in September 2008, many asset managers didn’t worry about the risk that a big broker-dealer could fail or not settle trades, says RBC’s Larson, who says his firm did consider the risk. These days, everyone has to think about it, he says.
Newcomers on List
Only five of the top 10 global brokers remained on the current list from the previous ranking, which covered the 12 months ended on June 30, 2009.
Right behind Deutsche Bank in Bloomberg’s world ranking is Credit Suisse, Switzerland’s second-biggest bank, which moved up from 10th. Macquarie Group Ltd., Australia’s largest investment bank, was a new entry in third place. No. 4 Liquidnet operates a platform on which the average U.S. trade size exceeds 45,000 shares, compared with just a few hundred shares on most U.S. markets.
UBS AG, Switzerland’s biggest bank, followed at fifth. Citigroup Inc., Nomura Holdings Inc., Jefferies Group Inc. and BNY ConvergEx Group LLC all joined the top 10. Goldman Sachs was ninth. Bank of America Corp., Morgan Stanley, Barclays Plc, JPMorgan Chase & Co. and Investment Technology Group dropped off the top 10 global list.
For the current ranking, Ancerno changed its methods to account for the difficulty of a trade and assess how much value a broker was providing. Ancerno first calculates the difference between the price of a share when the broker gets the order and the average price a client winds up with, called the implementation shortfall.
Then it subtracts that figure, in basis points, from the median result for trades with similar difficulty based on such factors as the stock’s momentum while the broker was executing the trade. (A basis point is 0.01 percentage point.) The previous ranking assessed brokers only on the implementation shortfall.
Credit Suisse, based in Zurich, did well on its home turf, coming in at No. 1 in Europe. Citigroup was next, followed by Deutsche Bank and UBS. Macquarie proved best in Asia, with Hong Kong-based CLSA Asia-Pacific Markets, a unit of Paris-based Credit Agricole SA, in second place. Since securities regulations differ sharply across the continent, brokers have to master a range of approaches in Asia, says Stevan Vrcelj, Sydney-based head of global cash equities trading at Macquarie.
“An execution is not just something you throw into a machine,” he says. “You need to customize the technology to the nuance of each market, visit clients to understand what they want to achieve from the execution and assess whether the right algo and trading strategies are being used at the appropriate time.”
Like Deutsche Bank, Jefferies, No. 2 in North America, has revamped its offerings as clients demand more tools.
The 49-year-old firm, which has relied on human traders to quietly locate counterparties for customers who want to dispatch blocks of shares, built an electronic equities-trading business from scratch in the past few years, says Dan Charney, who heads both electronic trading and U.S. equity sales trading. Brokers now need algorithms, as well as research, derivatives and salespeople who can compete globally, he says.
“We went from a relatively low footprint in the electronic trading business to it being a meaningful part of what we do,” he says.
James Angel, a finance professor at Georgetown University in Washington, says brokerages have to balance people and computers.
“You can’t be a good institutional broker without the old-fashioned smartness of the guy who watches the market,” he says. “You also need the technology and you need the understanding of market structure to get really good executions.”
This year is the first that money managers’ trading with algorithms will claim the same portion of their orders as human brokers, or “high-touch desks” -- 35 percent each, New York-based research firm Tabb Group LLC estimated in July.
Deutsche Bank began building up its electronic trading two years ago. It focused on algorithms that continually rank and reassess dark pools so clients and its own staff can trade with the least amount of detection.
“We’ve been very nimble and aggressive in developing products that help customers improve their execution quality,” says Bausano, who adds that big customers will no longer work with brokers whose electronic offerings and human traders aren’t -- in his word -- “superb.”
Marques, 46, who has a doctorate in particle physics from the University of California, Irvine, joined in January 2010. Today, he oversees 45 people around the world from his New York office -- half of whom arrived since mid-2009.
On a January day, Marques’s team is analyzing statistical differences in how trading occurs across more than 50 venues in the U.S. and which types of firms are likely to be in those markets. They’ve adopted some tactics of high-frequency traders.
If an algorithm discovers that the relationship between a stock price and a basket of shares is veering from its historical level, the program may alter how it trades or it may not transact at all.
“We monitor every trade in real time,” Marques says. “We’re choosing when not to trade, and we’re minimizing information leakage while maximizing liquidity.”
As markets splinter and rivals rebound, Marques’s computers will have to work even harder.