Is IRR a reliable measure for private equity performance or does it display intrinsic weakness?
If you use the Internal Rate of Return (IRR) to measure the performance of your private equity fund you may be surprised to learn why it is inherently flawed, how it can create skewed interpretations not to mention the problems it causes with benchmarking. With many limited partners drilling down on very precise details of fund performance to make their allocation decisions, private equity will have to adopt a broader palette of performance metrics, which should prove to be more robust than the questionable IRR.
Traditional private equity performance measurement
Private equity fund performance has traditionally been measured by using the Internal Rate of Return (IRR) coupled with a combination of other multiples – such as Distributed to Paid-in Capital (DPI), Residual Value to Paid-In Capital (RVPI) and TVPI (Total Value to Paid-in Capital) – plus Paid-in to Committed Capital (PIC). These metrics have somehow become the industry standard endorsed by all the official industry bodies including the European Private Equity and Venture Capital Association (EVCA) and the British Private Equity & Venture Capital Association (BVCA) in their respective reporting guidelines. As well as The International Private Equity and Venture Capital Valuations Guidelines Board (IPEV Board) in its draft Investor Reporting Guidelines (IRG) for Private Equity and Venture Capital Fund, and Global Investment Performance Standards (GIPS). One of the most compelling reasons to choose IRR is that given the nature of private equity funds and their particular lifecycle, their performance cannot be measured with annualised returns, but should rather be measured on a since-inception basis. Second, IRR takes into account the timing of the cash flows unlike the multiples which neglect the time dimension. Third, IRR is relatively easy to calculate with the assistance of a computer and can be relatively straightforward to interpret, but can also be complex as we’re about to discover.
Controversy over using IRR to measure fund performance
The use of IRR has both ardent fans and critics alike. Not surprisingly, the fans are typically practitioners representing private equity investment managers or general partners (GP) while the critics are usually constituents of the institutional investor community.
A Swiss fund of funds limited partner (LP) comments: “PE firms have got to wake up to the fact that IRR is not an adequate performance indicator and that we are not going to buy into their fund simply because they claim a top-quartile IRR ranking – everyone always says they’re top-quartile.”
But it is not only some LPs that criticise IRR. Some leading academics and other private equity practitioners are also sceptical of IRR as a robust performance indicator.
Josh Lerner, Jacob H. Schiff Professor of Investment Banking at Harvard Business School says: “When you look at how people report performance there’s often a lot of gaming taking place in terms of how they manipulate the IRR.”
Another recognised expert in private equity mathematics, Professor Oliver Gottschalg of HEC, Paris and Co-founder and Head of Research at Peracs Due Diligence Services, said at a recent PE Accounting Insights seminar in London: “The IRR is such a bad measure of private equity performance that I do not know why the industry still uses it.”
Despite the controversy surrounding the use of IRR to measure fund performance has been ongoing for years, it still remains the industry-recognised performance indicator advocated by all the industry bodies.
So, does IRR have major flaws?
In fact, there are four potential major pitfalls when using IRR to measure performance private equity:
- Computational difficulties – multiple IRRs and no defined solution
Computation of an IRR involves an iterative search procedure that may not always converge. The IRR can cope with symmetrical cash flows, but as soon as these start to be variable (with alternate positive and negative signs – in-flows and out-flows), the cracks begin to show in its computations. The trouble is that virtually every single private equity fund has asymmetrical cash flows, which means that both timing and amounts are hard to predict with the added complication of changing the direction of the cash flow more than once (e.g. out-flow, in-flow and then again out-flow and in-flow). When the direction of the cash flows varies, there is a high probability of generating a multiple/unstable IRR – a situation where there is more than one solution to satisfy the equation, or in other words, for one set of cash flows, the calculation will produce two or more IRRs, each of which is the right answer so this outcome can be tremendously confusing. There are also cases in which the IRR for a given stream of cash flows is mathematically not defined, that is, there is no solution.
- An inherent flaw in the assumption underlying the IRR calculation: the reinvestment assumption
The standard IRR formula always makes the implicit assumption that at any given time excess cash is reinvested at the IRR rate generated up to this point in time. It is known that the effective rate of return experienced by investors differs from IRR. They are equal only if investors can reinvest intermediary distributions at the IRR rate and borrow at the IRR rate to finance intermediary payments, but this is virtually impossible in practice.
- Gaming/manipulation of the second inherent flaw
This second inherent flaw just explained is dangerous enough on its own without intentionally using it to manipulate the results. On the top of that, some ‘skilful’ GPs that know how to use it to their advantage actually ‘plan’ their cash flows. Since the cash-flow timing can significantly distort a fund’s IRR, early wins (quick returns of significant amounts early in the life of the fund or investment) can disproportionately boost the IRR. Not surprisingly, there is a lot of directly linked gaming and manipulation happening.
The private equity asset class benchmarking standard of comparing funds with the same vintage years is arguably an inadequate measure, according to some experts, and one which creates a number of challenges. By only comparing the vintage years there is the risk of making comparison between apples and oranges as the grouping criterion to identify comparable funds has significant limitations. For example, how do you compare a $25 million fund to a $5 billion fund? Or a single-industry specialist fund investing in cleantech companies to a generalist fund investing across industries as diverse as food and retail to financial services? The second challenge is the ambiguity of the classification of funds. It is not always clear if a distinction is made between a first close and a final close, or whether funds are in fact counted in different vintage years such as when one fund closes on December 31 and another closes on January 1. Another concern surrounding like-for-like comparison is which database is used as the reference point – some databases show vastly different results, resulting in problem-ridden benchmarking.
What is it with all these ‘top-quartile’ claims?
Recall the comments made earlier by the Swiss fund of funds investor that “everyone always says they’re top quartile”. In support of this, Oliver Gottschalg’s research states: “Using a representative sample of 500 PE funds, we found that 66% of all funds can claim to be 'top-quartile' according to at least one data source for their actual vintage year. Giving PE funds an additional flexibility with respect to the choice of the vintage year (one year before or after), this percentage increases to 77%.”
What actions do we need to take next?
Now that we know what is wrong with IRR it would seem doubtful that we should continue to deploy this measurement tool. Perhaps we should continue to use IRR at least for official reporting purposes until the industry bodies figure out (provided that they are trying to) more advanced metrics. Although not perfect, modified IRR (MIRR) or the equivalent Net Present Value (NPV) largely tackle the well-known pitfalls of traditional IRR. There are a number of other metrics and approaches to consider: Public Market Equivalent (PME) and PME-Plus; time-zero IRR; annualised Rate of Return, and Alpha Performance Benchmarking against other private equity investments or risk-adjusted private equity returns.
More and more LPs are appointing expert consultants to advise them on how to pick the real winners. It is likely some time soon there will be a collective realisation in the private equity asset class that the existing performance metrics are just not adequate. But until then it is your choice whether you want to keep your head buried in the sand or if you want to beat your competition and stay one step ahead of the LPs before they start asking uncomfortable questions about your performance.
If you want to learn more about this subject, be sure to attend our Private Equity Performance Measurement & Portfolio Mathematics Master Class in New York City November 20, 2012.
For more information, visit: http://pemathematics.peaccountinginsights.com