The Skorina Report: Why You Can’t Clone Yale

The Skorina Report: Why You Can’t Clone Yale

By Charles Skorina
Can you copy great endowment investment performance?  The Yale endowment, for example?
Reverse-engineering a good performer is probably achievable.  The key attributes can often be identified and, to some extent, duplicated.
But can you replicate great performers?
That’s a much taller order.  There’s always something slightly mysterious about greatness.
We were reminded of that fact when we looked at a recent white paper by Drew Knowles.  He’s the founder and chief investment officer of Berkeley Square Capital Management, an investment advisor in Denver. (“Invest Like an Endowment”)
Mr. Knowles assembled data from NACUBO-Commonfund (NCSE) surveys and analyzed recent historical relationships between asset allocation, endowment size, and performance.  He did a good job and it’s worth reading.
As an investment advisor to (not necessarily wealthy) individuals he would like to demonstrate how they can use ETFs to replicate the investment performance of major endowments.  He’s particularly interested in showing the usefulness of so-called “liquid alternatives,” in which his firm seems to specialize.
He argues that this really can be done, at least to the extent of reproducing the average performance of big (over $1 billion AUM) endowments.
It would be even better if an amateur could match the performance of the very best endowments, such as Yale.  Unfortunately, his model says they can’t quite.  But he argues that his Yale Clone should still be an attractive portfolio for the average investor, and makes a good case for it.
Mr. Knowles suggests that an individual investor, by using certain low-fee index-tracking ETFs and rebalancing every year per the latest NCSE data, can do (very nearly) as well as the average big endowment with its professional investment staff and well-compensated external asset managers.  He only actually demonstrates this for 2003-2015, since not all his preferred tracking ETFs existed in earlier periods; but so far, so good.
(This is not an altogether original idea.  Our friend Mebane Fabermade a similar argument in his book The Ivy Endowment a few years ago.  But more and better tracking ETFs have become available since then.)
Mr. Knowles then attempts to reverse-engineer Yale, which hugely outperformed the average big endowment 1997-2014.  It earned an annualized 13.2 in that period versus 9.9 percent for its sisters.
He writes: “While the Yale Clone mimics the actual Yale portfolio asset allocation to the best of its index-based ability, and matches the Yale endowment standard deviation, the Yale Clone lags its target by three-percent a year.
It is highly evident that the Yale allocators add alpha (300 basis points a year!) by picking active managers that handily beat comparable indices.  Yale’s endowment consistently allocates capital to first quartile managers… Such winning allocations compound in Yale’s favor to the benefit of the school.”
I recently had a chance to talk to some people close to Yale and tried to hone in on how that “picking active managers that handily bear comparable indices” thing is actually accomplished.
As many of our readers might suspect, it’s not really all that mysterious; but it’s hard to do in practice if you’re not Yale, and very difficult to replicate.
An elite school can often detect unusually talented people when they are still students, track them into their professional lives and become early investors.  We should underline “can,” because not even all the elite schools have demonstrated the ability to do this as consistently and effectively as the Yale Investment Office under Dr. Swensen and his team.
That applies to startup external fund managers.  But it also applies to Yale’s internal staff, which judiciously draws from the same best-and-brightest cohort.
There’s a story abroad that I hear is true.  Years ago, a senior Yale officer working after hours happened to look into a wastebasket full of resumes of people applying to work at the YIO.
He spotted a young Yalie with an unusual background: sixteen years as a world-class professional ballet dancer, who was now raising a family and currently earning an econ degree.  A check with his professors indicated that he was an outstanding student (who would soon graduate summa cum laude).  So, Robert F. Wallace was brought in as an intern, then subsequently as a full-time associate for four years.
That’s the same Robert F. Wallace who was recently hired as chief investment officer at Stanford Management Company.
In our story about Mr. Wallace a few months ago we opined that he himself was clearly trying to replicate Yale’s performance and offered some evidence for that assessment.  That won’t be easy to do, but he probably has a better shot at it than Mr. Knowles with his Yale Clone portfolio.
In our conversation with Erik Lundberg, chief investment officer at the University of Michigan endowment a while back, he indicated that he is also well aware of this tactic.  He said he keeps very careful track of his graduates and networks with them to access investment opportunities.
Michigan is a terrific school with a well-run endowment, but it’s still probably not as favorably situated culturally or geographically as Yale vis-à-vis the very top tier of fund-managers.
For the rest of you, there’s always Mr. Knowles’ Yale Clone.  It seems to be doing better than most big endowments, albeit with a higher standard deviation than either the average big endowment or Yale itself.
Here’s the performance of Mr. Knowles’ Yale Clone versus Yale, a 60/40 portfolio, and the NCSE over-$1 billion endowments from 2003-2015:
Comparative Performance: 13 years 2003-2015
Yale Clone
$1 bn
Annualized Rtn %
Std Dev %
Berkeley Square Yale Clone Portfolio: Categories and
Tracking ETFs.  With Yale actual allocations 2014
Allocation Catagory
Index used
Yale Actual Percent FY2014
Domestic Equity
S&P 500 Index
(SPTR Index)
Foreign Equity
(MXWDU Index)
Fixed Income
Barclays Aggregate Index
Absolute Return
HFRI Fund Weighted Composite (HFRIFWI Index)
Real Assets/Real Estate
Dow Jones REIT Index
(REIT Index)
Natural Resources
S&P No. American Natural Res. Index (SPGINRTR Index)
Private Equity
Cambridge Assoc. US Private Equity Index (Note 1)
Barclays 1-3-month Treasury Bill Index (LD12TRUU)

Charles Skorina works with leaders of Endowments, Foundations, and Institutional Asset Managers to recruit Board Members, Executives Officers, Chief Investment Officers and Fund Managers. Mr. Skorina also publishes THE SKORINA LETTER, a widely-read professional publication providing news, research and analysis on institutional asset managers and tax-exempt funds. Prior to founding CASCo, Mr. Skorina worked for JP MorganChase in New York City and Chicago and for Ernst & Young in Washington, D.C. Mr. Skorina graduated from Culver Academies, attended Michigan State University and The Middlebury Institute of International Studies at Monterey where he graduated with a BA, and earned a MBA in Finance from the University of Chicago. He served in the US Army as a Russian Linguist stationed in Japan.

Be Sociable, Share!

One Comment

  1. Brad Case, PhD, CFA, CAIA
    April 20, 2016 at 5:16 pm

    Very interesting article, Charles, as was the paper published by Drew Knowles. I have one major concern about the prospect of cloning Yale’s performance, and I would very much value your thoughts. My concern is this: I simply don’t believe Yale’s numbers.
    Yale’s latest endowment report ( includes three statements about the returns on their Venture Capital portfolio, each of them striking:
    (1) “Since inception in 1976, Yale’s venture capital portfolio generated a 33.8% annual return” (page 5).
    (2) “Over the past twenty years, the venture capital program has earned an outstanding 92.7% per annum” (page 16).
    (3) “The venture capital portfolio earned an annualized return of 18.0% for the ten years ending June 30, 2015” (page 28).
    The market value of the endowment ($25,572.1 million) and the VC allocation (16.3%) reported on the inside front cover peg the size of the VC portfolio at approximately $4,168,214,733. If that were the outcome of returns averaging 33.8% per year since 1976, then (assuming no additional capital invested, and also no withdrawals) that suggests they started in 1976 with an initial investment of $48,771. That’s possible–but do find it believable?
    (Yeah, yeah, maybe they started with a much greater initial investment and simply withdrew a stupendous amount each year. Possible–but do you believe it?)
    Similarly, if the VC portfolio size were the outcome of returns averaging 97.2% per year over the past 20 years (!), then (under the same assumptions) that suggests the portfolio was worth $8,363 in 1995. Again, possible–but do you believe it?
    Moreover, if they started with $48,771 in 1976 but had just $8,363 in 1995, then that suggests they lost 8.86% per year over the first 19 years of VC investing before embarking on their amazing 20-year run of returns averaging 97.2% per year. Do you believe it?
    Finally, if the VC portfolio size were the outcome of returns averaging 18.0% per year over the past 10 years, then that suggests the portfolio was worth $796,404,904 in 2005. That sounds much more reasonable–until you realize that implies that the return grew from $8,362 in 1995 to $796,404,904 just ten years later, for an average annual return of 215%. Now that would be an investment performance to trumpet–if it were true.
    There is, of course, one remaining explanation: what they mean when they use words like “generated” and “earned” is not what normal people mean when they use that sort of term. Specifically, of course, they presumably measure VC “returns” using IRR, which Ludovico Phalippou noted in 2009 “is probably the worst performance metric one may use in an investment context” because it “exaggerates the performance of the best funds, can be readily inflated and provides perverse incentives to fund managers.”
    It’s very interesting that Yale’s endowment report uses the general term “return” (or “investment return”) something like 80 times, but never once uses the more clearly defined terms “IRR” and “internal rate of return.” In fact, the report uses a clearly defined term exactly one time, in comparing Yale’s 10-year “dollar-weighted” return (that is, IRR) to active and passive benchmarks in various assets classes. In other words, even in an investment context where it would have made perfect sense to use time-weighted returns, Yale chose instead to use “probably the worst performance metric one may use in an investment context.”
    As I say, I would be very interested in your thoughts.

Leave A Reply

← 2016 Allocator Trends Report: Consensus and Dissension Greek Debt: Back in the Headlines →