IRVIN GOLDMAN Building a team for Cantor Fitzgerald

Cantor Fitzgerald has earned much of its reputation as a fixed-income inter-dealer brokerage, but many are surprised to learn that the institutional equity side of the business is its most profitable. Now the firm wants to bring fixed-income sales and trading up to the same level as equities. And Irvin Goldman is the man hired to do the job. Dalia Fahmy reports

Cantor Fitzgerald, which made its name as Wall Street’s largest inter-dealer broker of US Treasuries, wants to turn itself into a diversified institutional brokerage and has placed its bets on Irvin Goldman to lead the way.

Goldman joined Cantor in July of 2003 as CEO of debt capital markets and asset management, with a mission to build up Cantor’s institutional fixed-income business to service small and medium-sized clients who have been neglected by the behemoth banks created by the consolidation on Wall Street over the past five years.

A former Wall Street dropout and a self-described artist with a passion for “quality of life”, Goldman seems an unlikely CEO. The native Long Islander spent his childhood summers working the counter at his father’s drugstore in the hopes of becoming a pharmacist, and only set his eyes on Wall Street after he saw how much money his older brother earned when he joined Lehman Brothers fresh out of college.

After gaining his MBA from New York University in 1983, Goldman became a repo trader at Salomon Brothers, and eventually climbed the industry ladder to become head of interest rate sales and trading at Credit Suisse First Boston. Then he quit.

For three years, Goldman took art classes, traded securities on the internet and went to the beach with his family. He even ended up as executive producer on a documentary film with a grant from the Sundance Film Festival.

In his office, the only reminders of his extended holiday are a gigantic photograph of his two sons playing in the sand that greets anyone who enters, and a couple of his drawings that hang almost out of view. “It was spectacular,” he says. “I thought I was never going to come back to Wall Street.” Goldman did return in 2003, ending his three-year hiatus, to head up Cantor’s new debt initiative.

Most people know Cantor as the fixed-income inter-dealer brokerage that helped pioneer electronic bond trading with eSpeed, a firm that now earns $26 million annually acting as a middleman between large banks that buy and sell bonds. But fewer people are aware that Cantor has had an institutional equity business since the 1960s, and that this side of the firm has always been its most profitable, earning more than the fixed-income brokerage. Goldman’s job is to bring the fixed-income sales and trading business on a par with equities, and start an asset management division along the way.

“Debt capital markets has a big sibling to keep up with,” says Cantor Fitzgerald CEO Howard Lutnick, referring to the equity group.

Although Goldman says he “shed a tear” when he decided to come back to Wall Street instead of running a technology firm as he had originally intended, he says Lutnick’s offer was one he couldn’t refuse. “The idea of being able to build your own company within a company, and be in such an entrepreneurial environment where you can bring in a private partnership and create an old-model firm…was very exciting.”

Close ties

Lutnick and Goldman have known each other for over a decade, from the days when Goldman traded bonds at CSFB and used Cantor as a broker. Their sons are in the same class at school, and on the morning of September 11 when Lutnick escaped the terrorist attacks that killed 658 Cantor employees while he was dropping off his son, Goldman’s wife was taking their son to school too.

Their lives crossed again while Goldman was still on leave in the summer of 2003. Goldman was trying to help some former colleagues sell their interest rate sales and trading operation and he approached Lutnick about a deal. The two men met in Lutnick’s office around seven o’clock in the evening to discuss it, but the conversation soon turned to Cantor.

Lutnick told Goldman he had been rethinking the firm’s future. He had already spun off eSpeed with an initial public offering in 1999 after realizing that Cantor’s inter-dealer fixed-income business was quickly being dominated by electronic trading. (A few years later, Lutnick would spin off the last remnants of the inter-dealer voice brokerage into a separate unit called BGC Partners.) On that evening, searching for a new direction for the firm, he asked Goldman for advice. Goldman suggested that he build an institutional fixed-income business that mirrored the business model on the equity side. “We talked until two in the morning,” recalls Goldman, and the two men planned the new division over take-out sushi.

Goldman’s vision made sense. Over the past five years, he said, mergers had dramatically reduced the number of banks and brokerages with which institutional investors could trade. Sales and trading desks, as small divisions within gargantuan corporations, were not allowed to deal with clients who did not bring in a certain amount of business to the rest of the firm.

At the same time, New York was flooded with mid-career professionals between the ages of 35 and 60, who had either lost their jobs in the financial sector consolidation or were too old and experienced to slave away anonymously at huge companies. Many of these people were looking for a second, more fulfilling career, and complained about the way they were getting paid: senior managers balked at receiving a large percentage of their pay in options when the stock market had become so unpredictable, and savvy managers didn’t like being at the mercy of the booms and busts of other divisions over which they had no control. “There was this sense that if they could, they would rather be in a pay-for-performance environment,” says Goldman.

Some of these veterans tried their luck at smaller platforms, such as start-up hedge funds, but soon they realized the difficulty of launching a profitable business without a critical mass of capital and clients. “We thought there was a tremendous opportunity to build a business focusing on specialized products in debt markets, products where customer relationships and value were the hallmarks,” says Lutnick.

By the end of the conversation, Goldman and Lutnick had decided to create a fresh fixed-income business that would fill the gaps left by the financial sector consolidation, with a bit of old-fashioned partnership similar to that of Salomon Brothers or Donaldson, Lufkin & Jenrette, circa 1983.

Growth plans

Since Goldman joined Cantor, he has hired 150 people and plans to add another 150 in the next year and a half. While his group generates less than a fifth of the income that the equity side earns, Goldman says the two groups should be on par in the next few years. “I’d be very happy if in five years we had $1 billion in revenues,” says Goldman about debt capital markets and asset management combined, predicting that in five years debt capital markets will generate about 75% of the income while asset management should contribute 25%. Eventually, this distribution should even out or could even tip slightly in the favor of asset management, he says. More tangibly, Cantor recently moved into new offices on East 59th Street—its first permanent headquarters since its old base atop One World Trade Center was destroyed. The firm is expanding geographically as well, and more than 10 new offices have sprung up in the past two years from Los Angeles to Darien, Connecticut.

Goldman eventually hopes to have about 400–500 employees, with the biggest locations expected to be in New York, Chicago and, surprisingly, Charlotte. “There’s a lot of opportunity in Charlotte,” he says, explaining Cantor’s decision to compete in the home base of banking giants Bank of America and Wachovia. “There are a lot of high-quality people working down there.”

Regional and middle-market institutions have been a good source of business for Cantor’s debt capital markets group, but Goldman says he has been able to attract a wide range of clients from high net worth individuals to insurance companies, hedge funds, thrifts, real-estate investment trusts and mutual funds. The idea, he says, is to focus on clients and not on business lines, so the firm can grow with the market. “We pride ourselves on having a very flexible structure. We’re not overly committed to one business or one sector, and we have the flexibility to move with our clients,” says Goldman.

To build elasticity into the firm’s structure, Cantor pays employees mainly in commissions. That way, the staff is forced to adapt to the evolving needs of the clients, says Goldman. “You want people to be flexible. It’s very demand-driven since people are getting paid based on the business they’re doing and the clients they have,” he says. When demand changes, employees have no choice but to shift too.

Who’s on the team

Right now, it’s easy for Goldman to tout his group’s ability to adapt, given that most of its desks are still in start-up mode. The repo and securities lending desk co-headed by Jeffrey Kidwell, former head of repos at Morgan Stanley, is the most profitable and complete with about $40 billion in daily trading volume. Cantor has always had a big repo presence as an inter-dealer broker, and the new business contributes about 25% of the debt capital market group’s revenues and 50% of the profits, although Goldman expects this to even out as the other desks mature.

Kidwell co-manages a 20-member team and expects to hire three to five more people this year, making Cantor’s one of the largest specialized repo teams on Wall Street. Growth has been strong, with a matched book that is four to five times larger than it was a year ago, and Goldman hopes trading volumes will double in the next 18 months.

Most of the current business is matched book trading for middle-market customers, along with some securities lending, municipal bond reinvestment, and financing for the expansion, but down the road Kidwell expects to trade more structured repos, derivatives and extend the curve on the repo-matched books from the current maximum of one year to five years.

“Our business model is to be focused on the client relationship, and find the middle-market business, or the niche business that’s been disenfranchised by all the larger broker-dealers,” says Kidwell, pointing out that the financial sector consolidation has halved the number of primary dealers to 20. “We come at an unusual time when the Street needs balance sheet, and the Street needs client relationship focus which they don’t feel they’re getting from the larger players.”

In October, Cantor rolled out an electronic repo trading system (CantorFITS) which is currently available to clients on a view-only basis but will be upgraded into a trading tool soon.

Cantor’s second-largest desk is the mortgage-backed and asset-backed securities and agencies group, under trading chief Steve Bischoff and sales head Scott Golden. With 30% of the debt capital market group’s revenues and 25% of profit, the MBS, ABS and agency group has the largest staff with 60 salespeople and traders, and another 40 expected to join by the end of the year. Revenue has grown six-fold from January 2004 to 2005, says Bischoff. Currently, about 60% of revenues comes from mortgages, 30% from asset backeds and 10% from agencies, but Bischoff expects this distribution to even out as the group becomes more established.

One of Cantor’s least profitable units, but expected to grow rapidly in the coming year, is its corporate bond business. In March, Goldman hired Carmine Urciuoli, formerly with Bank of New York, to head up credit and sales trading. The group currently has a staff of about 18, which is expected to grow to 30 by the end of the year. Urciuoli says despite a tougher business environment overall, Cantor has an opportunity to further build out its investment-grade and high-yield desks, and should eventually also fold credit default swaps into the mix.

“What’s working to our advantage in this phase is there seems to be some turmoil on the Street,” says Urciuoli, referring to the flurry of personnel changes on Wall Street earlier this year which has made it easier for Cantor to hire.

Urciuoli is Cantor’s second shot at building up the credit group. Last year, Goldman brought in a credit market veteran—Robert Jackson, formerly a bond trader with Advest Securities—who ended up steering Cantor’s business toward odd-lot trading of small and unevenly sized blocks, which Goldman says was a mistake given that many of these transactions now go through electronic platforms.Instead, the group now plans to focus on medium-sized and large blocks, often traded by larger, relative-value oriented clients. At the same time, Cantor is developing a technology platform that will carry much of the odd-lot volume in the future.

Once Cantor has sufficiently expanded its corporate bond distribution, Goldman plans to begin underwriting debt. To do this, he plans to sign alliances with investment banking boutiques which can offer their deal-making expertise in return for Cantor’s sales network. “We’re not going to go out and hire a bunch of investment bankers and go pitching our ability to underwrite debt. What we’re going to do is continue to service clients the way we do; we want to be able to provide that service to them if they want it.”

Other desks that Goldman has built up and intends to grow further include interest rate products, municipal bonds and futures.

Asset management

Goldman tends to express his goals by referring to client needs, and as a way of servicing customers and making money for the firm, he plans to build an asset management business with a prime brokerage. In fact, Cantor already lets its employees manage their wealth at the firm using in-house technology, research and investment recommendations, in keeping with the old private partnership model.

Goldman says, “We believe that two people should always come to work: you and your money. It’s very, very difficult to work hard at a job and manage your personal wealth outside. To the extent that we can help our employees do that, great. And to the extent we can help our clients do that, it’s going to be a very valuable business.”

Now that Cantor has successfully tested its asset management capabilities on employees, the firm is working on building the technology to launch its services to clients. At first, the service will consist mainly of the prime brokerage business and private wealth management for individuals with over $50 million, but eventually Goldman hopes to launch proprietary funds, a fund-to-fund business, and someday even index funds.

As part of this effort, Cantor has also started producing research for the first time, using an unusual model pioneered by Goldman: instead of offering research as a free service to clients who trade with the firm, Cantor is selling its research for money.

So far, the debt group only has two analysts, one who covers market technicals and another who covers high yield and distressed debt. They sell individual reports to Cantor salespeople and traders, but also to clients. The idea is to keep analysts competitive by making them earn their living directly, like other commission-dependent bankers.

“A lot of big firms have these huge institutional infrastructures for research,” says Goldman, explaining that he wanted to avoid building a research department that writes reports that just collect dust on clients’ shelves. “We end up turning the research analyst into a business person.”

This business model for research is one of Goldman’s many attempts to create a unique culture at Cantor, through innovations based on lessons he says he learned from working on Wall Street for 20 years. “There’s something very special about being part of a private partnership, about being a person who can impact the growth, being excited about your job, and having much more of a family-oriented culture,” says Goldman. “I think that is absent on Wall Street.”

Many of his innovations echo the older days of white-shoe partnerships where commissions ruled and Wall Street lived by the axiom: you eat what you kill. “It’s a complete meritocracy, and it’s direct pay-for-performance. The salesmen and traders know every day when they walk out how much money they’ve made for themselves and the firm. And they know they can’t make money without servicing their clients,” says Goldman.

With the partnership model comes a flatter organizational structure, and a looser hierarchy, which in theory allows beginners to climb the ladder more quickly. “I don’t know anyone else in my graduating class who sits next to someone as senior as I do,” says Jane Ginsburg, who graduated from Tuck School of Business at Dartmouth last summer and now sells MBS and ABS from her desk next to Scott Golden.

Ginsburg is one of 10 new hires who went through debt capital markets’ first training program last year, which Goldman says he modeled on Salomon’s famous 1980s classes.

“You could tell that the management really made a commitment,” says Ginsburg, pointing out that senior partners gave detailed presentations and made an effort to become acquainted with the trainees. “In comparison to my colleagues who are at different firms, I think our program is far more useful.” At other firms, she says, training programs are often a mere formality.

New hires get their first indication that working for Goldman is different when they open their welcome packages and find a book called The Art of Happiness at Work by the Dalai Lama. Goldman is also keen to foster a family-friendly culture and spends a lot of time talking about quality of life. Some of this emphasis on family life is natural, given the central role that the families of the victims of September 11 play at Cantor. (The victims’ families will continue to receive 25% of Cantor’s profits until next year.)

“One thing I learned from taking three years off is that when you’re working for 20 years, 18 hours a day, you can’t get the balance between family life and work,” says Goldman. He describes his home as the richest part of his life. He eats breakfast with his two sons and wife every day at 6:30am, makes it home for dinner at least two times a week, and rarely works on weekends. As part of his efforts to make Cantor a better place to work, Goldman plans to start a monthly lecture series in a classroom he had built on the trading floor, where employees can learn about anything from nutrition to photography without having to leave the building.

In the firing line

Perks like these make working for Goldman sound pleasant, but it’s also clear that his sights are squarely set on the bottom line. Goldman is not shy to admit that he fires non-producing employees, and since he joined in 2003, he has let go more than 20 staffers. “If you hire 150 people, it isn’t unusual that 10% or 20% won’t fit in,” he says.

Goldman calls non-producers “carnation people”, alluding to relatives at weddings who are not part of the wedding party but are given flowers to wear to make them feel important. “They just walk around like they are somebody but they don’t do anything,” he says. “We don’t have a lot of those people.”

Goldman’s singularity of purpose seems to have served him well so far. Given how much his group has grown, and that he has a history of working at private partnerships that end up going public—first at Salomon Brothers and then at Credit Suisse First Boston—one question that springs to mind is whether his group intends to go public.

Goldman says, “I don’t think we plan on doing anything. But opportunities always come up, and you evaluate them at the point they do.”

For now, Goldman says he only wants to serve clients, give employees a good place to work and making a living in the process. “It’s always been our philosophy that it’s not about being the biggest, the baddest. It’s about being the best,” he says.

Caution flag for eSpeed Electronic inter-dealer brokerage eSpeed was spun off with an initial public offering in 1999, but Cantor Fitzgerald still owns about 40% of the business. With Howard Lutnick as chairman and CEO and Kevin Foley as president, eSpeed transacts trades worth an annual $45 trillion in markets such as corporate, municipal and government bonds, futures, options and energy instruments. Although eSpeed helped pioneer electronic bond trading and business boomed initially after the IPO, the company has recently been hurting. Competition from rivals such as Icap subsidiary Garban and BrokerTec helped slash 2004 earnings almost by a third, to $26 million. The stock is down more than 50% on the year. ESpeed’s recent string of bad legal luck hasn’t helped. In February, a jury declared invalid eSpeed’s claims that Icap had infringed on its electronic bond trading patent. ESpeed says it plans to appeal. And in March, a shareholder class action suit charged that eSpeed touted itself as an “unmitigated success story, but knowingly or recklessly mispresented and failed to disclose material adverse facts”. Espeed says the suit is without merit.