Rarely have two hedge fund managers taken opposite sides of a bet and each made hundreds of millions. But Paul Marshall and Ian Wace, the founders of London-based Marshall Wace, have done just that. When they launched their business in 1997, the two partners held fundamentally different views of investment research: Marshall, former head of Mercury Asset Management’s pan-European equity business, thought that the investment calls made by stockbrokers were worthless. Wace, former global head of equity and derivatives trading at Deutsche Morgan Grenfell, maintained that the banks wouldn’t pay for research if it didn’t have value.
The partners decided to settle their long-running debate in 2001 by designing a system that, in a highly disciplined, quantitative way, would measure the accuracy of the sell-side investment calls they received. The result was TOPS, short for “trade optimized portfolio system,” a filtering tool that pools stockbrokers’ top picks into one vast, interactive database of sell-side tips. Although those calls are certainly informed by the intelligence coming from the banks’ research teams, Marshall Wace does not solicit direct input from research analysts -- but rather from the brokers paid to make sales calls to clients.
Initially, Marshall Wace used TOPS merely to track recommendations for its flagship Eureka Fund. The firm launched a stand-alone fund strategy based on the system in 2005. Since then TOPS has evolved into an increasingly powerful device for harvesting investment intelligence from markets around the world, with ideas from 5,000 individual sell-side brokers operating in 53 countries.
“If you have the bandwidth to absorb information from a vast number of individuals, you quickly realize that there is an awful lot of idea generation out there -- far beyond anything you’d be able to capture by conventional means,” says Wace, a brusque, intense 44-year-old who serves as the firm’s CEO. “A vast pool of alpha is just sitting out there, beyond the reach of most investors.”
The impact of TOPS on Marshall Wace has been profound. The system, which uses algorithms to analyze and rank brokers’ recommendations, has transformed the firm’s business model and driven its profits to new heights. The $3.67 billion Opportunistic TOPS Fund, which relies on investment ideas based on market themes and stock-specific events from sell-side generalists and country specialists, has delivered an annualized return of 19.23 percent from its January 2005 inception through the end of July. The $2.84 billion Fundamental TOPS Fund, which was also launched in January 2005 and which culls company-specific ideas from industry and sector specialists, delivered an annualized return of 16.20 percent during the same period.
By the end of July, the Marshall Wace team -- concerned about the increased volatility in the global equity markets caused by credit fears and the possibility of the subprime contagion spreading -- had already begun to reduce its exposure to the financial sector across all of its TOPS products. Although the market upheavals in August did cause a spike in the funds’ volatility, the two biggest TOPS funds ended the month down 1 percent or less, and all of the TOPS funds are still in positive territory for the year.
TOPS also powers Marshall Wace’s first-ever closed-end investment fund, MW TOPS, which was launched on Amsterdam’s Euronext exchange in December 2006, raising E1.5 billion ($1.96 billion). With $10.9 billion in assets under management, Marshall Wace was the ninth-largest European hedge fund firm when this year began, according to Alpha’s Europe 50 survey. As of July 31 it employed 120 people and had $13.8 billion in assets under management -- the vast majority of which was invested in TOPS-driven funds.
“Marshall Wace has taken a fascinating approach to gathering information, and it has really had an impact on the way in which people think about handling the equity markets,” says Sir Andrew Large, chairman of MW TOPS and a former Bank of England deputy governor. “I think it’s very exciting.”
In TOPS, Marshall Wace’s founders have devised a scalable, sustainable system they can use to build whatever investment products they want: market neutral, long-only, geographically focused, even dedicated-sector funds. The most innovative aspect of TOPS may be its fusion of stock recommendations with data about the brokers themselves -- which ones, for example, perform better in a rising or falling market -- creating what is in essence a whole new investment universe, where the securities are brokers’ stock picks rather than the stocks themselves.
“TOPS is almost akin to a disruptive technology,” says chairman Marshall, 48, a reticent, blond, bespectacled manager known for his fundamental investing style. “And the development of the system presents an enormous global opportunity for us because the data we gather, which is massive, can be analyzed and assembled in an almost infinite number of ways. We’ve only just begun to scratch the surface of what is possible.”
Although Marshall Wace holds first-mover advantage, other hedge fund firms, investment banks and traditional money managers are also developing proprietary electronic data filters, known in the industry as alpha-capture systems. In London two hedge fund powerhouses, GLG Partners and Gartmore Investment, have recently created their own self-styled alpha-capture funds to skim the cream of their brokers’ best ideas. And investment bank Merrill Lynch International has created an in-house system, called Trade Ideas Monitor, to highlight its teams’ high-conviction ideas.
In addition to local threats emerging in London, global changes in the hedge fund industry itself are driving Marshall Wace’s evolution. Long-short equity investing may be among the most liquid of strategies, but it is also the most crowded: Since 2002, when TOPS was launched, the number of hedge funds worldwide (including funds of funds) has nearly doubled, to 9,832, according to Chicago-based Hedge Fund Research. Assets in long-short equity have increased by 280 percent in those five years, to $489.6 billion -- more than a quarter of the $1.76 trillion managed by hedge funds worldwide.
“The industry is becoming more crowded, and the best managers are those who are really pushing to evolve their strategies,” says Michael Burton, head of EAFE equity long-short research with London-based Financial Risk Management, which manages
$13 billion in funds of hedge funds. His firm began investing in Marshall Wace’s value-oriented Eureka Fund in 1999 and now has money spread across a range of TOPS funds.
“Some managers have gone from being pure value investors to adopting strategies that are much more akin to private equity,” says Burton. “Others, like Marshall Wace’s TOPS team, have become more and more systematic in the way that they differentiate themselves.”
MARSHALL AND WACE NEVER EXPECTED TOPS TO BECOME such a powerful and profitable tool. Following the bursting of the tech-stock bubble in 2000, they were frankly skeptical that the sell side’s recommendations had any value whatsoever. In the press, research analysts were being excoriated for losing their objectivity in favor of reaping investment banking profits, and sell-side brokers were seen as lackeys, paid to simply distill their analysts’ reports. What that overriding view didn’t take into account, however, were the material differences between individual brokers’ recommendations -- which is exactly what TOPS was designed to determine.
In 2001, Marshall and Wace tapped a recently minted University of Oxford graduate, Anthony Clake, who has degrees in philosophy, politics and economics, to design a database that would assimilate the flood of broker picks and pans that crossed their doorstep. The first version of TOPS was little more than an Excel spreadsheet.
“When we undertook this project, we didn’t think that the results were going to be particularly good,” says the affable, boyish-
looking Clake, 27, now a partner at the firm and the TOPS project manager. “Our working assumption was that brokers would be totally bullish, despite the bear market, and cost us lots of money. The real discovery we made in those early years was that -- even as the markets were still going down -- some of the advice we were getting was truly valuable.”
The system that Clake, Marshall and Wace developed does far more now than simply automate and systematize classic phone-call interactions between the buy and sell sides. It captures, sorts and prioritizes tens of thousands of stock recommendations by geography, industry sector, market capitalization and contributor accuracy, among other metrics. The result is a multidimensional database of investment opportunities: In 2006, for example, the Marshall Wace TOPS (Europe) fund polled 1,796 individuals from 343 teams at 77 brokerage houses and collected 86,000 ideas in real time. That information was then processed and subjected to a multifactor optimization model that uses complex algorithms to filter, evaluate and weight investment ideas.
No single investment idea is allowed to dominate the TOPS funds, and single-stock positions are restricted to 3 percent of net asset value in any given portfolio (external risk factors, such as sector or country exposures, are also tightly constrained so as not to exceed 15 percent of the total risk). Although Marshall Wace trades heavily -- usually in $1 million blocks, making about 800 to 1,000 trades a day, according to Clake -- its traders are not chasing short-term momentum. In 2006 the average holding period for a single stock in the Opportunistic TOPS pool was 89 days; it was 173 days in the Fundamental TOPS pool. Through the first seven months of 2007, the holding period in Opportunistic TOPS had grown to 117 days; for Fundamental TOPS it was 206 days.
At its core TOPS is a massive investment-optimization engine. From a broker’s perspective, however, it simply looks like a personalized Web page: Each broker is granted a certain amount of notional capital for virtual investment purposes and then asked to allocate it to his or her most compelling investment ideas -- long or short -- and provide a 200-word rationale for each. Some small, single-sector portfolios may have as few as six positions; larger, generalist pan-European portfolios may have as many as 25. Total
notional capital can range from E10 million to E250 million.
Although none of these individual totals have any connection to Marshall Wace’s capital base or real asset allocations, they do allow participants to track how well their portfolios are performing relative to a given benchmark, such as the MSCI pan-European index.
“TOPS is a bit like fantasy football for salespeople,” says John Millar, head of the hedge fund advisory business in London for Merrill Lynch, whose private clients have invested more than $100 million in MW TOPS. “The most clever thing about the system is that Marshall Wace made it exciting -- they set up various inducements to participate and turned it into a competition. Having a chance to compete -- and get compensated for winning -- is what every alpha-oriented individual on Wall Street and in the City of London wants to do.”
Every month, Marshall Wace sends its contributors a report that reflects how much money their virtual investments would have made and ranks them according to their absolute and relative performance. Based on that cumulative data, Marshall Wace then allocates commissions to its contributors’ firms on a time-lagged basis.
Contributors to the TOPS system get compensated for the quality of their ideas regardless of whether Marshall Wace has actually traded on them, effectively reducing the pressure on brokers to do anything but generate winning suggestions. For those ideas Marshall Wace does adopt, the firm parcels out trade orders based on its assessment of the banks’ trade-execution quality and services, regardless of where the idea came from, further reinforcing the culture of performance-based compensation.
“I think TOPS taps into the psyches of many frustrated salespeople who would secretly like to be hedge fund managers themselves,” says Daniel Squires, a director of equity markets at Merrill Lynch who has contributed to TOPS for the past three years and is responsible for the bank’s European hedge fund coverage. “We’d all like to think that we could do a great job if someone would just give us several million to run. For many brokers this is as close as they’ll get to managing a real portfolio.”
Currently, TOPS receives recommendations from roughly 5,000 brokers, half of whom are based in Europe. Anyone on a sales team can contribute, but Marshall Wace looks for those who provide the best information: The firm adds about 200 contributors a year. The average broker has been in the system for more than two years and some as long as five, says Clake. The Marshall Wace team performs its own assessment and optimization of the investment ideas -- looking at more than 30 performance metrics, including duration, turnover and alpha generation -- before allocating any of the firm’s capital.
The detail that TOPS captures about its participants is mind-boggling. Clake can tap into the system and find out, for example, which brokers shine in a bear market or at what hour of the day they make the most accurate calls. Marshall Wace looks at the combinations of stock pick and broker to identify the best stocks to buy long or short.
“The contributors are their raw material,” says Squires. “And these guys put a lot of time and effort into managing us. They are constantly refreshing the pool.”
Marshall Wace’s focus on self-improvement is driven in part by its London rivals. In September 2006, GLG Partners began marketing its own dedicated alpha-capture strategy -- called the Esprit fund -- which the firm seeded with $400 million of its own capital and listed on the Dublin Stock Exchange. Just as at Marshall Wace, GLG’s system was first developed as an internal research tool and was run for more than three years before the creation of the Esprit fund.
GLG, however, gathers investment recommendations from a much more select group of brokers and imposes a very different filtering system. Instead of capturing as much data as possible on brokers’ picks, GLG asks its counterparties to select one broker from within their organization -- currently they have 51 total contributors -- and deputize that person to gather the firm’s best long-term, fundamental ideas. Just as with TOPS, those ideas are delivered to GLG through a Web-based system, and the results are then processed using algorithms to optimize the buy and sell signals for the portfolio.
“The best fund managers know that if you ask someone for their top 100 ideas, you’re likely to get diversified nothing,” says Merrill Lynch’s Millar. “But if you ask them for the five or six ideas that they’re most passionate about, then you tend to get better results and alpha capture.”
GLG’s Esprit fund is by far the biggest and most public rival for Marshall Wace, but Gartmore is also keen on capturing alpha. Last year the firm set up a broker-driven fund, AlphaGen Acamar, which is even more concentrated than GLG’s Esprit. Acamar traces its origins to the work of several members of fund manager Roger Guy’s team, including Guillaume Rambourg, who is a senior manager of European equities for Gartmore. They invited about a dozen of their favorite, most-trusted brokers to submit their highest-conviction ideas for a new hedge fund.
“Part of the impetus for developing this strategy was the realization that only about 40 percent of the ideas coming from our best brokers were making their way into our existing funds,” says Rambourg. He cites various reasons for those omissions: The funds primarily focus on large-cap companies, and some of the brokers’ suggestions were small-to-midcap names; other suggestions were focused on companies in peripheral European countries, like Portugal and Greece, which have never been central to their fund mandates.
Unlike Marshall Wace, which typically has 750 to 1,000 live positions in any one of its TOPS funds, Gartmore has constructed a concentrated portfolio of just 70 to 80 positions. Every idea comes from its chosen brokers, who are allowed to submit four to six recommendations per quarter. Participants are encouraged to present only those ideas they believe will have a price movement of at least 15 percent, up or down, during that three-month period.
Although all of the picks are equally weighted in the portfolio, Gartmore does have a few rules in place to prevent the fund from having a disproportionate amount of risk: No single stock can be more than triple-weighted, even if several brokers recommend it. The fund also has country and sector limits of 25 percent, to maintain diversification, and a rolling stop loss of 10 percent on any single position before it gets cut. Gartmore determines the gross and net exposures of the fund, but beyond those parameters it truly is a broker-driven product. Although AlphaGen Acamar delivered a 14 percent net return from its February 2006 inception through August 2007, the fund has yet to find many adherents. Currently, it has just $200 million under management, with 80 percent of that coming from Gartmore itself.
While GLG and Gartmore pride themselves on deputizing only a few brokers, and allowing them to have input into their customized alpha-capture funds, Marshall Wace is moving rapidly in the opposite direction -- becoming ever more inclusive and voracious in its search for information. In the past few months, the firm has added brokers in the Philippines, Russia, Turkey and Vietnam. No market is off-limits, provided that it offers sufficient liquidity and a well-established network of stockbrokers.
Thanks to TOPS, what was once an arbitrary assessment process has, in the span of just five years, become increasingly quantifiable and standardized. More important, the surprise that TOPS delivered -- the proof that many brokers do add value, if one knows how to listen to them -- has already transformed sales desks at investment banks all over the world. As TOPS reaches maturity, the system could begin to drive fund development in new and unanticipated directions. Wace, who speaks of the possibilities with an almost evangelical fervor, offers examples of new and unusual arbitrage opportunities across countries and industry sectors -- opportunities that few investors have seized because they require gathering and interpreting vast amounts of information, including behavioral data, on the brokers with whom they work.
“The key drivers of TOPS are Wace and Clake’s passion for data and for precision in their analysis,” says Merrill Lynch’s Millar. “They were the first to turn this type of research into a semi-industrial process -- they have got dozens of quants and engineers working to enhance this system every day.”
What began as a bet between friends has since transformed the firm: Marshall Wace now has 28 mathematicians working full-time on optimizing TOPS, 15 dedicated traders, eight risk managers and 12 local-language speakers in far-flung locales to recruit brokers. Some 50 other people work on its Internet technology and back-office systems at centers in India and Vietnam, making sure that everything operates smoothly. In a world where fund managers have long questioned the worth of brokers’ recommendations, Marshall Wace has simply turned the use of that information on its head -- and built a highly successful business.
