This article is a continuation of MPI Quantitative Research series focusing on 2021 fiscal year endowment performance:
Harvard 2021 | Bowdoin 2021 | UPenn 2021
Duke University’s endowment grew to $12.7 B in FY 2021, on the strength of a 56% return. Even in a year where endowments are putting up superb numbers, that’s a staggering result.
The latest annual report is also a reminder of how good a single allocation bet can be for your bottom line. It’s well-known that Duke was an early investor in Coinbase, which went on to hold its IPO in April of this year. There was no question that this would pay out for the university – the only question was by how much. A March 2021 Coindesk report perfectly captured the ongoing interest leading into the IPO, quoting a source that predicted “They probably 100x’d their money, after dilution.”
A 100x return? Crypto chasers scan message boards and social media looking for a moonshot like that. But as Coindesk points out, Duke didn’t have to go chasing anything. It was Coinbase co-founder – and Duke alumnus – Fred Ehrsam who approached them with the offer. Every endowment should be so lucky.
Whatever the actual size of their investment (more on that later), there’s no question that Duke’s endowment can call 2021 a win. But as the endowment return season comes to a close, CIOs will have to turn their eyes to FY 2022, and that raises a natural question: will endowments try and top themselves next year or should everyone be preparing for returns that are a lot more grounded?
As Bitcoin Goes, So Goes the Coinbase
To understand the true drivers of Duke’s returns, we use the same approach we’ve used in our past endowment analyses. We start with the endowment’s annual returns, as well as various indices representing specific asset classes, and then use MPI’s proprietary Dynamic Style Analysis (DSA) model to come up with a replication portfolio that explains the true drivers of performance.
This approach isn’t limited to endowments; we’ve used it to dive into fixed income, target-date funds, and complex hedge funds too. It always comes down to finding a good set of indices, and that meant we had to solve a very particular problem: how to approximate an equity (Coinbase) with a very limited return history?
As it turns out, Coinbase (COIN) stock price movements are very closely related to Bitcoin’s, as we show in the 30-day rolling performance chart below.
To be precise, for the fiscal year 2021 ending June 30th, we use Bitcoin prices through April and Coinbase stock prices afterwards. When combined this way, we estimate the Coinbase pro-forma return for FY2021 at 430%.
With Bitcoin as our proxy for Coinbase, we’re able to create a replication portfolio that helps us understand the behavior of the endowment’s investments. Right off the bat, our model begins picking up the estimated Coinbase exposure correctly in the past few years, separating it out from other venture capital and private equity investments.
This same replication portfolio also does a fantastic job of mimicking the actual returns of the endowment, shown below:
In the chart below we show DSA-estimated exposures of the endowment to various asset classes – both public and private – at the beginning of the 2021 fiscal year. Our system estimated Coinbase exposure at a whopping 2.32%, which translates into $200M for a $8.5B endowment. Given our earlier estimate of Coinbase’s return for the FY2021 at 430% – this places the value of the Coinbase investment slightly over $1B as of June 2021 and, most likely, 30% more today. Not bad!
It’s also worth mentioning that the combined estimated exposure to private equity (18.25%) and venture capital (20.45%) is approximately 41% if we include our Coinbase 2.32% estimate. That’s very close to the 42% 2020 private equity allocation for the separate $4B Duke Endowment, managed by the same management company as the Duke University endowment. Also, historical allocation trends of private asset allocations disclosed for The Duke Endowment very closely resemble the DSA-estimated historical exposures for the university endowment based on their annual returns.
Confident in our analytical approach, we next look to understand each asset class’s estimated contribution to returns for both FY 2021 and over the past several years.
Our model estimated Coinbase’s contribution to the endowment’s FY2021 return at 10%. In most recent years that would have doubled Duke’s returns, but as we mentioned in our Harvard piece, this wasn’t an ordinary year. In other words, even without Coinbase, it would have been a great year for Duke’s endowment. But adding in Coinbase…what an illustration of how a single investment can be an outsized driver of performance!
While describing Duke’s investment in Coinbase, Coindesk includes an anonymous source speculating that their shares “are now worth a small fortune, on the order of nine figures.” Looks right to us, and we’ll say this much: if endowments were committed to VC and PE before FY 2021, this year’s results will do nothing to shake that conviction.
What Have You Done for Me Lately?
This brings us back to our original question (and many others, come to think of it). As the FY 2021 endowment season draws to a close, how do CIOs plan to top this year’s windfall? Have risk tolerances been pushed to their limits? Should endowments prepare for a return to more conventional performance, or is there an appetite to increase investment in illiquid or highly volatile asset classes in the hopes of striking gold?
After legendary Duke basketball coach Mike Krzyzewski led the team to its first national championship in 1991, and then managed to do it again the next year, he famously said, “Once you win a National Championship, how do you do that again? A lot of it had to do with the fact that I didn’t give myself an opportunity to enjoy the first one.”
Of course, in basketball there is only one winner at the end of the season. Endowments face a less dire situation; it’s hard to find a loser in this year’s landscape. When the curtains are lifted next year, it will be interesting to see what everyone had planned for an encore.
Factors used in DSA Analysis:
- Cash & Equiv – 3M TBills
- Broad US Equity – Russell 3,000 Index
- Developed Equity – MSCI EAFE USD Index
- Emerging Equity – MSCI EM USD Index
- Natural Resources – Preqin Natural Resources Index (estimate 2021)
- Real Estate – Cambridge Associates Real Estate Index
- Private Equity – Cambridge Associates Private Equity Index
- Venture Capital – Cambridge Associates Venture Capital Index
- Hedge Funds – HFRI Fund Weighted Composite Index
- Bitcoin+COIN – CMBI Bitcoin TR USD through April 2021, COIN thereafter
DISCLAIMER: MPI conducts performance-based analyses and, beyond any public information, does not claim to know or insinuate what the actual strategy, positions or holdings of the funds discussed are, nor are we commenting on the quality or merits of the strategies. This analysis is purely returns-based and does not reflect actual holdings. Deviations between our analysis and the actual holdings and/or management decisions made by funds are expected and inherent in any quantitative analysis. MPI makes no warranties or guarantees as to the accuracy of this statistical analysis, nor does it take any responsibility for investment decisions made by any parties based on this analysis.