Cost-cutting by banks and doubts about the integrity of some of their analysts have got to have an upside.
Ethan Berman, chief executive of RiskMetrics, has found it. A New York-based quantitative analysis firm that was spun off from JPMorgan in 1998 by 25 people, RiskMetrics now employs just over 200 and provides more than 450 clients with risk analysis tools.
“I’d be lying if I said I saw how everything would develop,” says Berman. “Like I say, better lucky than smart. But I do think we did things well.”
Berman’s career switch in his mid-thirties followed 11 years in investment banking. He and his colleagues – risk managers and computer programmers – could not see much future at JP Morgan. “They looked at the senior people there and saw that they were not risk managers and programmers,” he says.
On top of that, their work, focusing on data management, daily risk reporting and cross-product volatility analysis, was not a core part of JP Morgan’s business.
Meanwhile, the Long-Term Capital Management disaster hit the markets, making risk analysis a function no bank could afford to be without, and forcing JPMorgan to cut costs. It made sense for all concerned for this division to be run separately. The spin-off was amicable, with JPMorgan providing part of the start-up capital.
RiskMetrics is three times larger now than Berman had predicted. It is comfortably turning a profit and Berman expects revenue to grow by 70% this year.
“I would love to say I predicted this and I saw how this market would explode,” says Berman, who insists he never entered the business for the IPO potential it held – the company is still privately owned, with all staff holding equity. “We just thought this would be an interesting thing to do. It has been more lucrative than we expected.”
Stress-test opportunities
Hedge funds and CTAs provide the biggest areas of growth. The RiskMetrics products offer them ways to stress test their positions and examine the interrelationships between, say, bond prices and equity prices, forex and commodities, or credit derivatives and BBB credits.
The firm also offers data cleaning services. In effect, this means that it is running bank indices, but the client bank can still market the index as its own and develop products from it. Berman reckons if he sells data cleaning to 20 banks at 10% of the charge they would usually pay in-house, he more than makes his money back.
With quantitative analysis already in full swing, Berman is considering moving into credit research and equity research. “I don’t believe quant research is the only way,” says Berman. “There is a role for qualitative research too.”
The ideal model for distributing this research, he says, is to sell it directly to investors. That rewards the top-ranked analysts that Berman considers truly earn their pay, and bypasses lower-ranked ones.
If RiskMetrics does move into qualitative research, it will probably be through an acquisition, partly because that would give a convenient point of entry to new clients, partly because, as Berman puts it: “It is hard hiring about 300 people. We’ve already been through that.”
RiskMetrics hires probably experienced a culture shock. Almost all of the original founding group of 25 are still at the firm. And Berman describes overall staff turnover as “absurdly low” – less than 3% per year.
For former bankers, fresh from title-obsessed, highly-structured bulge-bracket firms, it must be unsettling to arrive at RiskMetrics, where no-one has titles and the management structure is flat.
He concedes that this does not suit everyone. Without job titles, it is difficult to make people feel that their career is moving forward. He recalls the remarks of a member of staff who joined RiskMetrics from a large US firm.
In a job review, Berman asked him how he was getting on. “Well, it is very small,” said the new recruit. “And I have a lot more responsibility. I own my projects and if something doesn’t work it’s, like, my fault.” Whether that is a point of praise for RiskMetrics or a point of concern for the big investment banks is a matter for debate.