The Japanese attack on Pearl Harbor might have had a very different outcome had the United States made better use of a new and unique tool in its possession. The tool — called radar — could identify, track and anticipate the movements of specific targets using only position data and special filtering algorithms. Unfortunately, the technique was so new its users could not trust the results to avoid disaster. Radar eventually became a trusted cornerstone of civilian aviation and air defense, with advances like sophisticated filtering techniques, which helped to guide the descent of the Apollo 11 lunar module to the moon, among other key uses. In light of the Madoff scandal, in addition to other recent hedge fund fraud cases, it seems appropriate to ask why we can’t develop tools to warn us of these kinds of financial threats as well.
As it happens, such "financial radar" is available to investment professionals. However, it has only been with recent advances in technology and modeling that the approach has been able to address the needs of hedge fund analysts and gain their trust. The disciplined use of such tools, along with skeptical qualitative analysis, can help the industry better identify and avoid its next threats.
In his groundbreaking work on returns-based style analysis in 1992, Nobel laureate William F. Sharpe demonstrated that one doesn't have to know the historical holdings of a portfolio to understand its investment style and anticipate its performance and risk; rather all one needed was a stream of monthly or quarterly portfolio returns and a generic quadratic optimization technique. The adoption of returns-based style analysis tools over the past 17 years has made the industry more transparent, secure and efficient. The applications ranged from manager searches, due diligence and monitoring, to risk management and competitive intelligence.
Institutional investors, among the first to embrace returns-based style analysis, found that they no longer had to rely solely on a consultant's or manager's assessment of strategy and skill — they could easily vet both with the click of a button. Today, recent advances in returns-based style analysis have incorporated some of the techniques used in actual radar and hold much promise in improving hedge fund risk analysis.
Hedge funds are like fighter jets
Historically, returns-based methods were first applied to traditional long-only portfolios, which, similar to commercial and cargo jets, are subject to certain regulations and restrictions, move slowly and do not often dramatically change their course on a whim. For such portfolios, Sharpe's returns-based style analysis approach can effectively monitor changes in style exposures. Hedge funds, in contrast, are akin to fighter jets — fast-moving, exotic and employing an arsenal of shorting, leverage, derivatives and illiquid instruments. Moreover, hedge fund data is noisier, position information is usually unavailable and, at times, the funds even use "jamming techniques." Unsurprisingly, the use of traditional regression analysis is typically not successful in analyzing hedge funds. The tried and trusted filtering techniques that helped to guide Apollo to the moon, however, have proved to be very capable of handling the data noise and extreme dynamics of hedge funds. And, like radar, which was not specifically designed to detect UFOs but can still alert controllers if an object is defying the laws of gravity, advanced returns-based tools can help alert institutional investors to potential fraud.
So why wasn't the industry able to use such advanced techniques to identify the Madoff threat? First, although the traditional investment world embraced returns-based style analysis long ago as a central part of their analysis process, the hedge fund world still remains rooted in qualitative analysis and non-predictive ratios or statistics. Additionally, many of those that pursued a returns-based style analysis approach to hedge fund analysis in the past were unaware of the more advanced techniques that have moved to the forefront in recent years. Lastly, even the best tools and techniques ultimately require adequately trained analysts to properly apply them, interpret results and, finally, make investment decisions. The investment decision-making process of hedge fund investors can often defy comprehension. For instance, the Madoff anomaly was easy to spot because of the unusual smoothness of the returns. The Madoff funds were outliers in almost every financial ratio, including the Sharpe ratio, a measure of excess return for amount of risk. While statisticians are usually focused on finding good justifications to keep outliers in an analysis (and not to remove them), investors, on the contrary, are simply investing in outliers. Statisticians would generally agree that the greater the outlier, the more it has to be investigated. Investors often find such outliers to be irresistible.
It's not the first time that the hedge fund investors have faced threats. There's no doubt that the next case will be more sophisticated and deceptive. How can investors know that the returns reported by hedge funds are real? The appropriate tools and techniques, the same ones that once helped to guide Apollo 11, have been available for some time, and it is now up to investment professionals to upgrade their radar systems and diligently monitor the results.