Greg van Inwegen is in the catastrophe avoidance business. As chief investment risk officer at Ivy Asset Management, a $15 billion fund of hedge funds, van Inwegen patrols his firm’s portfolios with an eye toward potential financial peril. His approach is informed by a hard-earned appreciation for the unexpected: On September 11, 2001, van Inwegen was working for Deutsche Bank in one of the buildings across from the World Trade Center. Hence the rock-climbing gear he has stowed away in a credenza in his office at Ivy’s Manhattan location: harnesses, carabiners and two 60-meter ropes — enough to rappel down from his perch on the 30th floor of the Midtown high-rise. “I don’t expect to ever use it,” says van Inwegen, a recreational climber. “It sits in there gathering dust. It’s tail-risk management.”
As van Inwegen recognizes, today’s financial risk managers face an uncertain world that is subject to market tail events — those statistically unlikely occurrences that can have an outsize impact on returns. Tail events are like Murphy’s Law on steroids: If anything can go wrong, it will — and in a very big and messy way.
Divining where or when such events might occur, determining how they might affect investments and trying to defend against them are the risk manager’s ultimate test. An alarming number of hedge funds have failed at these tasks during the past year, raising questions about the application of risk management methods across the industry. Hundreds of billions of dollars in wealth have been destroyed by an especially strong and persistent series of tail events: the subprime market meltdown that began in the spring of 2007, the winter credit crunch, and this summer’s sharp and sudden fall in the prices of commodities-related stocks. The punishment for failing to properly prepare for and protect against these events has been swift and merciless.
Peloton Partners, a London-based hedge fund company run by a pair of Goldman, Sachs & Co. veterans, illustrates the problem. Its $1.8 billion, heavily leveraged asset-backed-securities fund went from being a star performer in 2007, with an 87 percent return, to collapse in early 2008, after the firm bet that mortgage bonds would rebound. They didn’t. Instead, the credit crunch undermined the value of the firm’s long bond positions so badly that Peloton was unable to raise enough collateral to meet margin calls.
The growing list of tail event casualties — which just this month claimed mortgage giants Fannie Mae and Freddie Mac and 158-year-old investment bank Lehman Brothers — has sent hedge fund managers scurrying for protection. They are hiring more qualified risk professionals at higher pay (see related story, click here) and giving them increased authority over investment decisions, including the power to order more hedging or even to unwind positions deemed too speculative.
Hedge funds are also investing more in software and services. New York–based RiskMetrics Group, a leading financial risk management software and services company, witnessed a 31 percent jump in revenue during the first six months of 2008 over the same period in 2007. Damian Handzy, chairman and CEO of Investor Analytics, which also supplies such systems and services, says demand for his company’s products has been soaring. “I have been in this industry for ten years, and I have never seen so many fund managers interested in risk,” he adds.
The solutions offered by firms like RiskMetrics and Investor Analytics revolve around more-extensive testing and analysis of investments, using techniques like stress testing and correlation modeling to try to estimate how different scenarios might affect portfolios. Comprehensive historical databases offered by these companies provide information for more-robust risk tests.
Still, most of the changes in risk management during the past year have been more incremental than breakthrough. Many hedge funds, including some run by large and sophisticated banks, were already beefing up risk management efforts more than a year ago yet got swamped by the flood of tail events. As van Inwegen says, “Over the past year these tail events just came up and hit you.”
Magnifying the problem is the herd mentality of most asset managers. The phenomenon is the opposite of what has come to be known as the wisdom of crowds. Financial tail events are all about the lunacy of herds.
“Until we start to model herd mentality, we are never going to get a good enough handle on tail risk,” says Andrew Lo, a professor of finance at the MIT Sloan School of Management, as well as chairman and chief scientific officer of the $600 million Cambridge, Massachusetts–based hedge fund firm AlphaSimplex Group. “When you have certain types of financial market dislocations, all hell breaks loose at once. This is the source of tail events.”
Hubris nurtures and sustains tail events and undercuts risk management, notes Patrick Welton, founder and CEO of Welton Investment Corp. of Carmel, California, a $500 million hedge fund specializing in managed futures and global macro strategies. “Human beings have a behavioral problem: They believe they control more things than they do,” he explains. “But it is impossible to understand all the changes that can take place. There will always be a new case in the future that we didn’t anticipate in the past. If you don’t have humility and include that in your risk assessment, you won’t be prepared.”
In August, New York–based Ospraie Management’s $2.8 billion flagship Ospraie Fund fell 26.7 percent, battered by the sell-off in the energy, mining and natural-resources sectors. The fall in these stocks was so swift and so steep that it caught Ospraie off guard, leaving the fund vulnerable to such heavy redemptions that founder Dwight Anderson had little choice but to announce that he would be shutting it down.
Ospraie joined a long line of hedge funds that have fallen during the past 16 months, starting with the spectacular collapse of Dillon Read Capital Management, shuttered in May 2007 by parent UBS following the fallout from the Swiss bank’s $37 billion in subprime-related losses. That spring two Bear Stearns Cos. credit funds lost $1.6 billion, hastening that venerable firm’s demise; JPMorgan Chase & Co. bought Bear this May for $10 a share, a fraction of the $150 price its stock commanded a year earlier. May also saw New York–based Drake Management shut down its $2.5 billion Global Opportunities fund after it got crushed by the credit crunch.
Lo says the fact that widespread losses piled up with armies of highly trained risk managers on the watch “points to one ineluctable conclusion: Something is broken.” The calls for change from investors, fund managers, consultants and academics revolve around two distinct issues: analysis and management. Put another way, the challenge is first to understand how much risk a particular investment might face, particularly during a tail event, and then to do something about it to protect the integrity of a firm’s finances.
Of course, one of the biggest problems is that risk managers often lack sufficient authority to unilaterally order corrective action when they see problems; too many are outranked by portfolio managers who can simply ignore their counsel. Yet even if risk managers were given more authority, it is far from certain that they could have protected their firms from the recent spate of events. Predicting the future during turbulent markets is still a guessing game.
“Nobody has a crystal ball,” muses Oleg Movchan, director of risk management at Alexandra Investment Management, a New York–based hedge fund firm with $1 billion in assets. “Nobody can forecast what is going to happen. What you can do is prepare the firm for a situation when a tail event does happen.”
Campbell Harvey, a professor of international business at Duke University, places the blame for the recent problems on practitioners. He says that university researchers have been steadily improving risk modeling and measuring methods to deal with tail events, creating models that can provide a better assessment of liquidity and leverage, but that financial firms have been slow to adopt them.
“In the practice of risk, there is not that much that has changed over the past ten years,” says Harvey, who points to Long-Term Capital Management, the hedge fund that famously collapsed in 1998. He notes that many of the firms that got caught out this time not only employed concentrated leverage like LTCM did but also had similar risk modeling and management methods.
According to Harvey, one of the key problems today is that too many decisions are based on analysis using normal return distributions. The practice of calculating the standard deviation of bell-shaped curves using historical returns doesn’t take into account the remote chance that a big, disruptive and unexpected event might occur and skew results.
“The standard deviation approach doesn’t really capture the tail,” Harvey explains. “What you need to capture is the forward-looking skew.”
Figuring out how to measure the likelihood and impact of future tail events has been a preoccupation of Harvey’s for decades. He published his first paper on analyzing and measuring the skewness factor back in 1976 and refined it in another paper in 2000. Other academics have offered theories and proposed methods for incorporating possible future tail events in risk analysis as well as better ways of predicting the potential impact of leverage and liquidity on a portfolio.
The lapses in risk practices were evident in a 2006 survey conducted by accounting firm Deloitte & Touche that polled 60 managers who ran 244 hedge funds. Most were in the U.S. and about a quarter had assets of $1 billion or more. Because the survey was taken two years ago, it offers a snapshot of risk practices before the subprime meltdown.
One of the most widely used risk testing methods cited by hedge funds in the survey was Value at Risk, which estimates the maximum potential loss in a portfolio at a given probability (usually set at 99 percent) over a set time period. But standard VaR tests were designed for so-called normal markets rather than turbulent ones and tend to be ineffective during tail events. Deloitte recommends that risk managers overlay additional metrics like stress and correlation tests to get a better portrait of exposure. Sixty percent of respondents who were using VaR also did stress and correlation testing, leaving the portfolios of the 40 percent who used only VaR dangerously vulnerable to tail events.
The survey also found problems in valuation practices for instruments like credit default swaps. Accurate valuation data is critical to running effective stress and correlation tests. The danger of inaccurate valuation in those types of lapses is that funds can underestimate their risks. Low risk estimates can serve as a green light to portfolio managers for boosting leverage — a tactic that can dramatically magnify the effect of a tail event. That scenario played out at Dillon Read, where an overreliance on VaR was partly blamed by UBS’s own management for the inaccurate estimate of the firm’s risk exposure.
“I think the biggest lesson from recent events is that the whole hedge fund industry needs to be much better at identifying, managing and pricing risk, especially liquidity risk,” says Alexandra Investment’s Movchan.
Several industry groups, research organizations and government panels have called for improvements in practices to help dampen the impact of tail events. In July, Celent, a Boston-based financial research and consulting firm, issued an extensive report on liquidity risk, which it identified as the root cause of the latest crises. Among its recommendations: Use stress testing and modeling to more accurately measure risk exposure from complex products, then set aside enough capital to cover a liquidity crisis in those investments for a fixed period of time.
At some hedge fund firms, risk management is woven into the entire investment process. New York–based D.E. Shaw & Co., for one, has put in place a framework designed to keep it from being swallowed up by tail events. Guiding the effort is a risk management committee whose four members are all part of the six-person executive team that runs D.E. Shaw. The committee sets the firm’s risk policy and parameters and meets at least every two weeks to review positions and strategies, market movements and risk profiles. Portfolio managers are responsible for monitoring and measuring their own risk profiles and adhering to the firm’s risk outlines. Their compensation is based on performance, which includes the risk-adjusted returns they generate.
D.E. Shaw also has a chief risk officer, who monitors the firm’s various portfolio positions as well as its overall risk profile. Since 2006 that position has been filled by Peter Bernard, who previously served as president of RiskMetrics. Bernard reports directly to the risk committee, putting him on equal footing with portfolio managers.
Eric Wepsic, a D.E. Shaw managing director and member of the risk management committee, refers to the system as a combined bottom-up, top-down model designed to keep several sets of eyes watching the firm’s risk profile. “We want to generate good returns in a sustainable and responsible way,” Wepsic says. “Avoiding catastrophic risk is a fundamental part of that.”
For van Inwegen, an unexpected event two years ago provided a wake-up call that led to reforms in Ivy’s risk management methods — and to a boost in his authority. Ivy had major investments with Greenwich, Connecticut–based hedge fund Amaranth Advisors, a play that backfired when a trader at that firm placed a huge bullish bet on natural gas — and lost. At least three Ivy funds were invested in Amaranth, which was forced to close after dropping
$6.5 billion in less than two weeks in September 2006. Spooked investors withdrew $1 billion from Ivy, prompting its management to restructure. Although Ivy’s risk managers had spotted Amaranth’s concentrated bet on energy by spring 2006 and sounded an alarm, its investment managers, who had the last word, waited too long to act.
In the post-Amaranth shake-up, Ivy elevated risk management to a position equal to that of the firm’s investment managers. Van Inwegen now chairs a risk committee that reviews all investments and has greater authority to demand changes in investment positions if he spots problems. Ivy also beefed up the system it uses to profile the individual fund managers with whom it invests, rating them not just on performance and operational standards but also on leverage, liquidity and the likelihood that they might get caught by a drawdown similar to what happened at Amaranth.
The firm also instituted a metric to ensure that no individual hedge fund gains too large a position within a portfolio, using green, yellow and red flags to rate funds. Under the yellow caution flag, an Ivy portfolio manager has to justify the level of investment held in a particular hedge fund. Under a red flag the position must be reduced immediately — no questions asked.
“In general, the process will maintain a discipline, so that the overall portfolio is not going to get hit by that kind of tail event,” says van Inwegen, who has a Ph.D. in finance from the Wharton School of the University of Pennsylvania. “That is what you are looking to avoid.”
For Welton, the Carmel, California–based manager, the two key investment considerations in risk analysis are leverage and concentration. Avoiding potential traps in those two areas, he says, can shield a fund from many of the types of problems that surfaced over the past year.
“Temporary heroes are made from leverage and concentration,” says Welton, who got his start in finance in the late 1980s trading futures and options for Commodities Corp., the Princeton, New Jersey, company that launched the careers of many renowned traders, including Paul Tudor Jones II and Bruce Kovner.
The firm Welton founded in 1989 maintains four levels of risk monitoring and review. Portfolio managers and traders are required to report risk considerations in all investment decisions. A portfolio review manager examines those decisions and reports on risk matters to the firm’s senior management. A separate compliance and risk manager independently inspects positions and also reports findings directly to senior management. The final stop is Welton himself, who evaluates risk profiles and reports and, together with his top managers, sets risk parameters.
But even firms with risk controls and multiple levels of review can’t totally immunize themselves. “You can’t eliminate all the risk,” Welton says. “So the discussion becomes, Are the risks understood? Is there compensation for their assumption? and, Can we absorb them so we don’t have what we might call catastrophic risk?”
For risk managers, understanding the issues and knowing what questions to ask is only half the battle. The investment world is populated by characters with notoriously oversize egos and a voracious appetite for profit. An impressionable risk manager with an academic rather than financial background parachuted into that environment can easily be overwhelmed. Indeed, one of the criticisms raised in the recent round of losses is that risk managers didn’t stand up to traders.
Satyajit Das, a former investment banker who now works as a risk consultant in Australia and is the author of a book on derivatives and risk called Traders, Guns & Money , says the personality and character of the risk manager plays a central role in the effectiveness of a fund’s risk management. Only strong-willed types with inquisitive minds will do.
“What’s important now is getting the right type of risk manager,” Das says. “He has to be the devil incarnate: internally skeptical, externally suspicious, not someone who is a patsy and rolls over and wants his stomach tickled. Today the patsies are winning.”
MIT finance professor Lo agrees that risk managers need to be strong enough to stare down traders within their own firms. He also subscribes to the principle that risk managers must rank high enough to be able to challenge or overrule portfolio managers.
At Lo’s hedge fund firm, AlphaSimplex, chief risk officer Arnout Eikeboom not only reports directly to the CEO but also has the authority to readjust investments by executing trades. Eikeboom has a significant incentive for taking a hard line on risk: His employment contract stipulates that he must ensure that investments stay within the firm’s established risk guidelines. If risk levels get out of whack, he can be terminated immediately.
“He has unilateral authority to shut down any strategy or to tell traders to unwind strategies,” Lo says. “He doesn’t have to ask me if a trade should be taken off. A lot of portfolio managers don’t like this, but that is the point.”
Lo argues that still more needs to be done to improve hedge fund risk management and that the missing link is some form of government or quasigovernmental oversight. The paucity of reporting requirements for hedge funds makes it difficult to understand and learn from fund failures, he says. Without that solid data, risk managers remain hard-pressed to make truly accurate projections of risk and to recommend appropriate hedges.
Ultimately, the challenge facing hedge funds is how to structure investment strategies that can produce better-than-average profits without endangering the firm’s capital in a tail event. In many cases, these risk strategies are, like van Inwegen’s climbing ropes, the only way out in an emergency.
As van Inwegen says, “I always want to have exit options.”
