Estimating Operational Risk for Hedge Funds: The ω-Score

Using a complete set of U.S. SEC filing information on hedge funds (Form ADV) and data from the Lipper TASS Hedge Fund Database, the study reported here developed a quantitative model called the ω-score to measure hedge fund operational risk. The ω-score is related to conflict-of-interest issues, concentrated ownership, and reduced leverage in the Form ADV data. With a statistical methodology, the study further related the ω-score to such readily available information as fund performance, volatility, size, age, and fee structures. Finally, the study demonstrated that although operational risk is more significant than financial risk in explaining fund failure, a significant and positive interaction exists between operational risk and financial risk. The hedge fund industry has experienced tremendous growth in the past decade. An estimated 9,000 hedge funds exist worldwide, with more than $1.8 trillion under management, compared with only $39 billion in 1990. In particular, institutional investors have increased their presence in hedge funds. The hedge fund industry, however, is also known for its high attrition rate. Selecting a successful manager can be very challenging. The increasing demand for hedge funds, together with potential failures as a result of operational risk, calls for due diligence in selecting high-quality managers, which is commonly practiced by many prudent investors before they invest.Because the assessment of operational risk necessarily relies on such intangible variables as historical manager behavior and unethical or illegal acts, due diligence in the hedge fund industry is primarily concerned with qualitative rather than quantitative matters. As the number of funds increases and the fixed cost of evaluating them remains constant, however, numerical scoring models are needed. Although a quantitative model can never fully replace human judgment, the analysis of “soft information” can help prioritize the due-diligence process. Indeed, with the increasing flow of available information about managers, a reliable model that reduces the dimensionality of the due-diligence process is essential to better assess hedge fund exposure to operational risk.In analyzing data from the U.S. SEC Form ADV filings, which some U.S.-based hedge funds were formerly required to submit, we found that operational risk, as measured by the past legal or regulatory problems of investment advisers or fund managers, is strongly related to such Form ADV variables as conflict of interest, ownership, and leverage. Hence, developing an instrument for assessment of operational risk, based on Form ADV data, is feasible. Given that future Form ADV filings will be limited and thus complete information on operational-risk cofactors may not be observable in the future, alternative models based on available information are warranted. In our study, we used variables in the Lipper TASS Hedge Fund Database to develop such an instrument. Through a statistical mapping technology, we were able to link the Form ADV variables with the TASS variables and then use the TASS variables to develop a risk instrument, the ω-score, which is a function of fund performance, volatility, fund age and size, and fee structure.We then turned to the crucial question of whether the ω-score could be used to predict fund failure. The main contribution of our study is a scoring model for detecting operational risk in the hedge fund industry. We also examined the interaction between operational risk and financial risk, especially the marginal impact of operational risk on predicting fund failure after controlling for financial risk. Although we have little doubt that more sophisticated models will be developed, our study demonstrates the feasibility of scoring funds according to their potential for operational-risk events.