You’ll never be a Yale superman: How Yale’s influential endownment fund and its CIO are reining in their ambitions

You’ll never be a Yale superman

Dan McCrum

| Oct 22 12:02 | 4 comments | Share

David Swensen is an investing superman. His pioneering use of alternative investments after he took over the Yale endowment in 1985 prompted a thousand imitators and created an industry. Indeed, the Yale model became the endowment model, so widely was it embraced. Then, with alternative investments legitimised as suitable for big investors, pension funds began to follow his lead, albeit tentatively. Yet you can never be David. Or rather, you can never match up to the myth that Mr Swensen’s performance has become. Consider the recent decision by the $21bn Yale endowment to cut its allocation to private equity for the first time in eight years, to 31 per cent from 35 per cent. That shift may simply reflect a lack of opportunities at present, but note the long-term return target which Yale expects its private equity investments to deliver.In aggregate, the private equity portfolio is expected to generate real returns of 10.5 percent with risk [volatility] of 26.8 percent.

Then compare it to the long run performance of Yale’s private equity portfolio touted in its annual update last year:

…the University is frequently cited as a role model by other investors. Since inception in 1973, private equity investments have generated a 30.0 per cent annualized return to the University.

Even this Yale doesn’t hope to match old Yale. But there is, to put it politely, some myth-making going on here as well.

Compounding investment returns at that rate for 39 years would be genuinely superhuman. Here is what it would look like graphically, if Yale had just allocated $1m to private equity in 1973 and grown it at 30 per cent a year:

 

That $1m would be worth $28bn, or half as much again as the current total endowment.

The problem, as outlined in an excellent paper by Ludovic Phalippou of the Said business school at Oxford, is that Yale has not, in fact, been superhuman.

The 30 per cent return cannot be compared to simple returns from other asset classes because it is an internal rate of return, or more specifically, the “since-inception pooled mean return” — a methodology used as part of the Global Investment Performance Standards set by the CFA Institute.

The weirdness of this number can be seen in the way it has barely changed over the last decade, standing at 30.7 per cent in 2003. The amounts invested since then are far larger than the previous 30 years, and returns have been nowhere near 30 per cent. (In 2010 the ten year anualised rate of return from private equity, a slightly different IRR calculation, was reported at 6 per cent).

To quote Dr Phalippou:

This is a third oddity with these return numbers; they simply do not add up. The reason they do not add up is because IRRs do not add up. IRR is not a rate of return.

Problems with IRRs are well known. It’s why cash multiples — total capital returned relative to that invested — tend to be used alongside reports of private equity performance as a sanity check. But few recognise quite how dramatic the overstatement can be, or look beyond the headline of Yale’s tremendous performance.

There is also something else going on here. Returns to venture capital are included alongside the buy-out funds in private equity, and Yale made several small investments that paid off handsomely in the 1990s, including in netscape with Kleiner Perkins, as well as Amazon, Yahoo.

Such venture returns likely had incredible IRRs, but represented small allocations relative to the sums committed to buy-outs.

Building a hypothetical model to replicate the returns, Dr Philappou is able to reproduce a from-inception IRR of 30 per cent, but as he concludes:

It is evident that the 30 per cent that is so often cited is unlikely to be anywhere close to the true rate of return. That true rate of return could be anywhere from single digit to maybe 20 per cent. Hence making Yale Endowment a model based on what is in their annual report is rather premature.

Now we expect that a response to this will be: of course a 30 per cent a year return is unrealistic, but there is still value to extensive diversification in alternatives. Yale has performed reasonably well since 2000.

What this line of argument, and the many reports advocating the endowment model tend to gloss over, is the central fact of Yale’s performance: it is only marginally about asset allocation. Yale has actually been unusually and exceptionally good at picking investment managers.

From Yale’s 2012 report (click chart to enlarge):

Yale compares itself to the median result of endowments tracked by Cambridge Associates, the investment consultancy, and a benchmark of median asset class returns based on its asset allocation.

For the twenty years ending June 30, 2012, nearly 80 percent of Yale’s outperformance relative to the average Cambridge Associates endowment was attributable to the value added by Yale’s active managers, while only 20 percent was the result of Yale’s asset allocation.

Unfortunately for mere mortals, while large institutions have actually shown some ability to pick stocks, the same study found no ability to pick alternative managers.

So you’ll never be David Swensen the superman, because he doesn’t exist. And if you want to adopt the so-called Yale model, you’d better hope that you are as skilled, or lucky, at picking managers as they are.