In a recent paper titled “Yale's Endowment Returns: Case Study in GIPS Interpretation Difficulties” in The Journal of Alternative Investments, author Ludovic Phalippou analyzed Yale's endowment private equity returns highlighting a flaw in the widely used and GIPS-recommended internal rate-of-return measure. The author's chief criticism pertained to Yale's private equity portfolio's since-inception IRR, which is published yearly in the Yale University Investments Office's annual review of the Yale endowment annual review of investments.
The first “oddity” the author points out is that an investor earning 30.4% over a 38-year period would have 24,000 times the initial investment. In Yale's case, if they the endowmenthad allocated $1 million to private equity in 1973, the value of the investment would be $24 billion. As of June 30, 2012, the Yale endowment stood at $19.3 billion.
A second oddity noted in the paper was that the return since inception “has basically never changed over time” while 10-year performance has changed substantially. Since-inception (“SI-IRR”) return has barely budged to 30% in 2010 from 34.1% in 2000, although the 10-year rate of return dropped to 6.2% in 2010 from 37.9% in 2000.
The last oddity Mr. Phalippou points out is that the numbers published in the report “simply do not add up.” The author calculates that if the returns were equally weighted by year or weighted by capital allocated, the returns since inception would be lower.
The oddities relate to the “implicit re-investment assumption” in IRR calculations. Successful investments early on in an investment program continue to compound at their return rate in future periods though subsequent investments might not have as high of a return.
Ludovic Phalippou constructed both a simple and complex cash flow model to illustrate the point. The complex model looked at a theoretical institution that invested a series of cash flows in private equity and earned an average return as measured by Cambridge Associates. The highly successful investments in the 1990s led to SI-IRRs not budging in the 2000s.
The author's suggestions for better performance reporting of an institution's private equity investments include a breakdown of return between types of private equity (e.g. venture capital vs. buyout), use a modified IRR approach or use a net-present value analysis.