Daniel Patrick Moynihan
I’ll confess a bit of surprise when I saw that my friend and colleague, Carl Bacon, CIPM, was speaking at the annual GIPS(R) conference this month on the subject of the internal rate of return. My reaction was that it would be like having a hooker give a talk on the merits of chastity. He and I have discussed and debated this topic ad nauseam, and so his views are well known to me. Carl sees extremely limited applicability of this very important measure, which, as you can guess, is totally opposite of my position.
And so, this post will attempt to set a few things straight.
The internal rate of return and liquidity
Carl tends to see the measure as one that’s sole use is for portfolios holding securities with limited liquidity; e.g., private equities. We know that for the general partner of a private equity investment, the internal rate of return is the only return measure used. Is Carl right, then? Does it have to do with the illiquidity of private equities?
To conclude this is to miss the point that everyone in the private equity space seems to be aware of: private equity managers control the cash flows. And so, unlike time-weighting, where we want to remove or reduce the effect of flows (since they’re typically controlled by the client, a point Peter Dietz brought to the industry’s attention nearly 50 years ago), for private equity we want these flows to impact our return.
Private equity managers obtain commitments from their clients (their limited investors, who commit the capital (“committed capital”) necessary for the anticipated investments); but until the money’s needed, it’s kept outside of the partnership. The manager will draw it down, via capital calls, which are, as you probably know, cash flows.
Liquidity isn’t the factor that drives the requirement for the IRR in this space: it’s the control of cash flows.
Comparing the internal rate of return
My colleague, John D. Simpson, CIPM, who attended the conference, informed me that “among [Carl’s] criticisms of the IRR was that it is not comparable with benchmarks and across managers.” Please let me respond to both points:
- As for comparing to benchmarks, the IRR fails because it’s money-weighted while the benchmark will be time-weighted. However, there are methods available to “money-weight” the benchmark, so that we have an applies-to-apples comparison. Carl apparently did make the point that the benchmarks could be adjusted for cash flows (or Public Market Equivalents) and that it would be okay to compare a money-weighted return against a money weighted benchmark.
- And as for “not being comparable … across managers,” my response is that this is untrue. Of course it is! Private equity managers all make their own cash flow decisions. And so, they’re being treated exactly the same. We want each to have these decisions included in their returns, as it demonstrates both their timing and investment skills. A pension fund, however, would not benefit from having their IRR compared with that of another fund’s, since each will have their own cash flow amounts and timings. An individual investor who received his/her “personal rate of return” for their mutual fund holdings (something that many funds now offer) might find some benefit in comparing with others, as a way to gauge their timing decisions relative to another’s. I’m sure there’s much more that can be said on this topic.
Applicability to liquid assets
John reported to me that Carl “also said IRR should never be used for liquid assets, even if the manager controls the cash flows.”
There’s a well known understanding regarding time-weighting: we use time-weighted returns to eliminate the impact of cash flows. Surely a corollary to this would be that if the manager does control the flows, we wouldn’t use time-weighting.
Carl’s dismissal of this is known to me, so his comments in San Diego aren’t surprising. However, if his statement holds, that we wouldn’t use the IRR even if the manager controls the flows, then this earlier rule (“we use time-weighted returns to eliminate the impact of cash flows”) can’t hold, and we must conclude that we just always, under all situations (except the liquidity of the asset) use time-weighting. That makes things very simple for us, doesn’t it? But, it’s incorrect.
We would definitely want to use the IRR when the manager controls the flows, whether the assets are liquid or illiquid.
In addition, we use the IRR when the manager doesn’t control the flows, in order to let the client know how they’re doing.
Where’s the evidence?
Carl does have a habit of making his claims with no real evidence to back them up. On occasion he will cite references, but they’re all ones he’s written, so they have no real objective value. His arguments tend to be his views which, of course, have value, but are not based upon independent evidence in our extensive literature.
However, we can find items in the performance literature dating back nearly 50 years ago that clearly identify the role of the internal rate of return in measuring performance. Industry luminaries such as Dugald Eadie and even Peter Dietz wrote that time-weighting was to be used to evaluate managers, while the IRR is to be used to tell us how the fund or portfolio has done (i.e., how the client has done).
The Global Investment Performance Standards even recommend today that asset owners report the IRR. Carl’s statements run in total conflict with these provisions.
The internal rate of return: the more important measure
Two of my colleagues, Stefan Illmer, PhD and Steve Campisi, CFA, have joined me in celebrating the value of the IRR for more than 10 years, and we’ve met with considerable progress in educating the industry on its value. We’re seeing its use take root and expand. In the United States, the Government Accounting Standards Board (GASB) now requires public pension funds to report the IRR on an annual basis; in Canada, the CRM2 regulation also requires reporting the internal rate of return. These are important initiatives that recognize the role this measure has in evaluating performance.
While Carl and I have our disagreements, we generally agree more often than we disagree. But on this subject, the differences in opinion are significant. Industry groups and performance measurement practitioners have recognized the value of the IRR for 50 years. I’m hoping that Carl will, too.
Will we hear from Mr. Bacon?
As you might expect, I wouldn’t post something like this without first running it by Carl. I sent him a draft yesterday, and asked him to review it. Had he objected to its publication, I would have, of course, refrained from doing so, but he did not. He made a couple corrections, which I incorporated, and promised that he would, in deed, offer a response. I also invited him to submit a “guest blog” post, if he’d like.
Carl is well respected in our industry, and deservedly so. However, one of the things many of us like about our segment of the industry is the controversy that surrounds many topics, including the applicability of money-weighting. I hope you find this post of interest, and look forward to hearing from Carl.