What the Literature Says: Private Equity Fund Valuations
As a result of investing in illiquid assets that lack secondary market prices, private equity funds are a natural target of regulatory concern. Economic researchers have focused on whether there are any systematic biases in the net asset values (NAVs) reported by private equity managers.
A 2013 working paper examined private equity reporting and found that reported NAVs of buyout funds were consistently better during the fourth quarter than any other time of year. Tim Jenkinson, Miguel Sousa & Rüdiger Stucke, How Fair Are the Valuations of Private Equity Funds? (Private Equity Inst. Feb. 17, 2013). One possible explanation is that funds’ assets are being valued more conservatively in the rest of the year. But another explanation is that fundraising for follow-on funds may affect NAVs. Research has shown that when general partners are promoting follow-on funds, the existing funds’ NAVs are higher on average, with the increase gradually unwinding over subsequent years.
There are various explanations for this performance pattern. One is self-selection bias: All else being equal, general partners are more likely to begin follow-on fundraising when the existing funds’ valuations are higher. Investors may have caught on to this possibility, as research has shown that general partners have greater difficulty raising funds for follow-on funds after increases in the reported NAV of the existing funds. Gregory W. Brown, Oleg R. Gredil & Steven N. Kaplan, Do Private Equity Funds Game Returns? (Fama-Miller Working Paper, Mar. 2015).
What the Literature Says: Hedge Fund Performance
Research on hedge fund performance has been more extensive due to greater data availability through hedge fund reporting databases. Recent findings have focused on statistical patterns in reported returns and whether statistical anomalies relate to potential wrongdoing. In a February 2015 speech, Mark J. Flannery, chief economist and director of the SEC’s Division of Economic and Risk Analysis (DERA), noted that the Office of Risk Assessment “combine[s] the analytic and financial modeling capabilities of DERA’s quantitative staff with the institutional market expertise of other Commission staff.”
For example, in SEC v. Yorkville Advisors, LLC (2012), the SEC charged a hedge fund advisory firm with inflating the value of assets under management and exaggerating the reported returns of its hedge funds in order to increase the fees collected from investors and to solicit additional investors for their existing funds and new funds. Also, in 2015, a case was brought against an investment advisor in which the SEC described its first use of a statistical analysis simulation for trade allocation and identified instances where it appeared advisors were inappropriately allocating trades, showing that
trades allocated to the principal’s accounts had an average one-day return of 6.28%, while trades allocated to other accounts had an average one-day return of 5.05%. . . . This analysis, involving one million random simulations, allegedly demonstrated that the likelihood of the principal’s accounts receiving such a high proportion of profitable trades by pure random change was less than one trillion.
Harry Frischer & Rachel Wolkinson, Proskauer, “SEC’s Data-Driven Analysis Identifies Allegedly Improper Trade Allocations by Investment Advisor,” Nat’l Law Rev., July 1, 2015.
A discontinuity in the distribution of monthly hedge fund returns was noted in research by Nicolas P.B. Bollen & Veronika K. Pool, “Do Hedge Fund Managers Misreport Returns? Evidence from the Pooled Distribution,” 64 J. Fin. 2257–88 (2009). Specifically, Bollen and Pool observed that the likelihood of reporting a slightly negative return appears to be significantly lower than the likelihood of reporting a slightly positive return. They hypothesize that this statistical pattern relates to manipulation of reported investment returns. As possible corroboration of their hypothesis, they found that the discontinuity is absent from audited three-month results. Similar research has found that hedge fund returns tend to be higher at the end of year. Vikas Agarwal, Naveen D. Daniel & Narayan Y. Naik, “Do Hedge Funds Manage Their Reported Returns?,” 24 Rev. Fin. Stud. 3281–20 (2011).
Other research, however, has offered explanations for those statistical anomalies that are unrelated to potential misreporting. One such explanation relates to how hedge funds accrue for performance fees with high-water marks. According to Philippe Jorion & Christopher Schwarz, “Are Hedge Fund Managers Systematically Misreporting? Or Not?” (111 J. Fin. Econ. 311–27 (2014)), because performance fees are assessed only on positive returns, and high-water marks affect fee accrual after losses are incurred, there is a tendency for net returns to exhibit discontinuity in the distribution of returns around zero.
This is not dissimilar from past research offering benign explanations for the distribution of corporate earnings around zero: Taxes affect positive earnings more than they affect negative earnings. E.g., William H. Beaver, Maureen F. McNichols & Karen K. Nelson, “An Alternative Interpretation of the Discontinuity in Earnings Distributions,” 12 Rev. Acct. Stud. 525–56 (2007). Consistent with this finding, Jorion and Schwarz showed that the discontinuity in hedge fund returns is much less pronounced in funds with lower incentive fees.
Bollen and Pool extended their research into possible hedge fund return manipulation by comparing information on the set of funds subject to SEC enforcement actions and lawsuits with information on those funds’ characteristics to obtain statistical indicators of potential reporting violations. Bollen & Pool, “Suspicious Patterns in Hedge Fund Returns and the Risk of Fraud,” 25 Rev. Fin. Stud. 2673–2702 (2014). The authors categorized allegations related to misappropriation, overvaluation, misrepresentation, and Ponzi schemes as performance violations. Because this research is based on allegations, not admissions of wrongdoing, it may highlight indicators that the SEC has used to screen funds for further scrutiny rather than provide indicators of wrongdoing itself.
Bollen and Pool identified several indicators that are more pronounced among the sample of funds with alleged wrongdoing as compared to the remaining sample, including repeated returns, infrequent negative returns, and performance that does not correlate with that of other hedge funds with similar investment styles. Jorion and Schwarz, however, cautioned that such indicators are not necessarily related to wrongdoing. For example, repeated positive returns may be associated with infrequent updates to the prices of illiquid assets or with how the value of a held-to-maturity fixed-income instrument is booked.
Investments in fixed-income securities, whose yields do not go below zero, may explain infrequent observations of small losses. Where a fund may have low correlation with benchmark indices, individual analysis is required to discern whether such low correlation simply relates to fundamental differences in fund investment strategies.
Hedge fund performance also can be temporarily affected by actual transactions. Research by Itzhak Ben-David, Francesco A. Franzoni, Augustin Landier, and Rabih Moussawi found that stocks that are more heavily owned by hedge funds exhibit abnormally higher prices during the last trading date of quarter ends. However, these price increases reverse on the next quarter’s first trading day. Ben-David et al., “Do Hedge Funds Manipulate Stock Prices?,” 68 J. Fin. 2383–34 (2013). Such unusual price changes are more pronounced for less liquid stocks.
Because such price changes are relatively small and short-lived, they are unlikely to have an effect on fund manager performance compensation. Also, even if a particular hedge fund invests in an illiquid stock with such unusual quarter-end price spikes, the fund manager may not necessarily be causing the increase, as many other funds often invest in the same stock.
Conclusions
The emerging literature on private equity and hedge fund returns provides potential quantitative signals that may attract a regulator’s focus. A recently launched SEC program called the Quantitative Research Analytical Data Support Program—which is currently processing market data from the Financial Industry Regulatory Authority, over-the-counter security-based swap data, and mutual fund flow data—supports these findings. Flannery, supra. It is worth noting that economic research has focused on large samples across thousands of observations. One should be careful, therefore, in applying such research to a single fund.
In our experience, inferences about potential reporting issues do not easily translate from large samples to an individual fund, where understanding case-specific facts and circumstances is required to explain patterns in observed fund performance. As noted above, the literature on private investment manager reporting issues is in its infancy, and debate is ongoing regarding whether statistical findings consistent with manipulation can be explained by reasons unrelated to wrongdoing.
Even taking the academic findings at face value, it is always easier to find statistical anomalies in large samples than to prove them for any individual fund. Identifying such benign reasons may require an individualized analysis by a fund, not only examining any patterns in its reported performance but also reconciling those patterns to that fund’s specific investment strategies.
Keywords: commercial and business, litigation, SEC, private investment funds, economic research, performance