Sunday, 26 March 2017
Last updated 2 days ago
Jun 30 2006 | 6:28pm ET
It may not be possible to tell which hedge funds will perform the best, but it is possible to see which ones are riskier than others, according to experts.
"Certain performance characteristics are predictable, however returns aren't predictable," said Franklin Neubauer, principal of Core Metrics, a quantitative research and consulting firm.
Earlier this week statisticians, portfolio managers and risk management experts gathered in midtown Manhattan at the 5th Annual Hedge Fund Analytics conference to discuss just what performance characteristics were possible to predict, and what tools were the most useful when analyzing hedge fund data.
"It could take around 15 years to statistically validate that a manager with an information ratio of 0.5 is successful or just lucky by looking at the information ratio alone, which is often more time than many hedge fund managers have been around," said Jordan Alexiev, a vice president in the portfolio and risk management group at State Street Associates.
Alexiev suggested that a better way to evaluate whether a manager is adding value is to test the ability of the manager to control the downside of her portfolio. He introduced tools such as within-horizon risk management to account for the probability of breaching a loss threshold at any given point in time, not only at the horizon's conclusion. He also demonstrated how the block bootstrap methodology can be used to evaluate if the riskiness of the actual portfolio exceeds what one would expect in the presence of no investment skill.
"On average, hedge funds [as described by Hedge Fund Research indices] are riskier than what we would expect when no skill is involved," he said.
When it comes to measuring risk, Neubauer said there are no hard and fast rules. "There are as many ways to measure risk as there are letters in the Greek alphabet," he said, adding that one that is often talked about is Omega. "Some people believe that Omega, which is return per unit risk, is an improvement over the Sharpe ratio."
Other presenters at the conference included Erin Roye, a former rocket scientist for NASA(literally) who now does research at Merrill Lynch.
"History does not predict the future," said Roye, who has been researching portfolio diversification, specifically in the fund-of-hedge funds space. She made some observations about fund-of-funds that she believes apply across the board when investing.
"What we found is that when you put together a portfolio of 40 hedge funds that you get much better risk adjusted returns when the hedge funds are not correlated." Conversely, "you don't gain very much by putting together a highly correlated portfolio of hedge funds." While Roye's research was focused on funds-of-funds, she theorizes that you could apply the same scenario to single-manager hedge funds. "We think this is a fundamental behavior, be it hedge funds ors specific securities," she said. In other words, if one were to create a portfolio of uncorrelated securities, it would outperform on a risk adjusted basis."
Andrew Weisman, managing director at Merrill Lynch, also had a few key observations about risk, as well as a few tongue-in-cheek comments about the industry. For example, he laid out one of Andy's Laws: "Losses are inversely proportional to the probability of their occurrence," Weisman said. "Just look a tLong Term Capital Management," he chuckled.
On a more serious note, he said that he too does not believe in looking at historical performance to gauge a fund's risk, rather he thinks that it is important to create a behavior model that can characterize a manager's betting style.
"Risk is a process that develops as you hold different assets and have different portfolio exposures," he said.