How Larry Summers lost Harvard $1.8 billion

By Felix Salmon
November 29, 2009
Larry Summers:

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Most people, if they’ve hired a legendary fund manager on a multi-million-dollar salary to look after investments and liquidity, would listen to the advice of that person. But most people aren’t Larry Summers:

It happened at least once a year, every year. In a roomful of a dozen Harvard University financial officials, Jack Meyer, the hugely successful head of Harvard’s endowment, and Lawrence Summers, then the school’s president, would face off in a heated debate. The topic: cash and how the university was managing – or mismanaging – its basic operating funds.

Through the first half of this decade, Meyer repeatedly warned Summers and other Harvard officials that the school was being too aggressive with billions of dollars in cash, according to people present for the discussions, investing almost all of it with the endowment’s risky mix of stocks, bonds, hedge funds, and private equity. Meyer’s successor, Mohamed El-Erian, would later sound the same warnings to Summers, and to Harvard financial staff and board members.

“Mohamed was having a heart attack,’’ said one former financial executive, who spoke on the condition of anonymity for fear of angering Harvard and Summers. He considered the cash investment a “doubling up’’ of the university’s investment risk.

But the warnings fell on deaf ears.

Summers, amazingly, wanted to invest 100% of the university’s cash in the endowment, and had to be talked down to investing a mere 80%. No wonder Meyer and El-Erian tried to talk him out of it: the Harvard endowment was never designed as a place to invest sums of cash which might be needed immediately. Instead, it’s designed to invest for the very long term, taking advantage of the higher returns on illiquid investments.

Summers was playing a high-risk carry-trade game with Harvard’s cash:

The aggressive investment of cash accounts is part of how the university has long run its “central bank,’’ an account that holds funds from its various schools and pays them a modest US Treasury rate of return. The “bank,’’ in turn, has invested the lion’s share of that money with the endowment, generating returns that are used to pay for shared needs, like graduate housing and financial aid.

No one had the stones to stand up to Summers when it came to this high-risk strategy of essentially borrowing at Treasury rates and investing the proceeds in an illiquid long-term endowment — certainly not James Rothenberg, Harvard’s part-time, unpaid, California-based treasurer.

After Summers left, sheer inertia took over, and nothing happened — maybe because El-Erian was soon on his way out as well. The result was that the university ended up losing 27% of its $6 billion in “cash”: a whopping $1.8 billion. There’s no indication, of course, of any kind of apology from Summers.

Update: Brad DeLong, in the comments, does some back-of-the-envelope math and reckons that Harvard came out ahead of the game, on net, even after accounting for that $1.8 billion loss. But that’s exactly the difference between a long-term endowment, on the one hand, and a “cash account”, on the other. If you have money in a cash account, you spend it. And money you’re spending should be liquid, not tied up in an endowment which can drop 27% in one year.

And Viyada York says that Harvard’s investment committee, rather than Summers, should be held responsible for the loss. It’s true that the managers of the endowment are responsible for its performance. But they’re investing for the long term. Summers should absolutely be held responsible for the decision — which was entirely his — to invest the Harvard cash account alongside the endowment, despite the fact that the cash account required much more liquidity than the endowment as a whole.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

So…isn’t whomever permitted ‘inertia’ to prevent the university from changing its endowment’s asset mix responsible for something that happened more than two years after Summers left the university?

Posted by Sterling | Report as abusive

I understand that Summers is more newsworthy than his successors, but what I get out of the Boston Globe article is that he paid attention to the endowment and made some risky but ultimately successful decisions, and his successors (Bok and Faust) paid no attention and reaped the whirlwind. Having read Summers’ articles of the time, I fully believe that he would have (and did!) noticed what was happening before it was too late. Faust has absolutely no financial expertise, of course. But Harvard’s ultimate managers, the so-called Corporation, should be taken to task for this — and for hiring Faust. Summers was long gone by the time the strategy he pushed went wrong.

Posted by Wordsmith | Report as abusive

So what margin did Harvard get on its invested cash? IIRC, endowment returns averaged 11% per year from 2001-2007, with short-term interest rates over that time span averaging 3%. So it would seem that 6 x (11% – 3%) -27% = 48% -27% = +21% puts Harvard ahead on the deal, no?

Posted by Brad DeLong | Report as abusive

Brad, this is actually a meta-post. Felix is simply giving us a practical demonstration of asymmetric loss aversion, as a prelude to a forthcoming discussion of cognitive biases in economics.

It’s a good thing Felix just writes about money & doesn’t manage it for anyone.

Posted by Joe P | Report as abusive

Joe P, I wouldn’t hire Felix to run money either. But I’d hire him a year before I hired Larry Summers, and a decade before I hired Alan Greenspan.

“Under Neil Rudenstine, Harvard’s president from 1991 to 2001, cash was heavily invested in the endowment and surged from $290 million to $2 billion. Under Summers, the figure more than doubled again…”I’m sure Larry Summers is an ass and everything, but given Harvard’s assumed ability to raise cash during these occasional losing periods, it really doesn’t seem so terrible to me to stick as much as they can in their long-term, higher-yield investments.This part strikes me as a bit more irresponsible: “Harvard also would pay $500 million to get out of the interest-rate swaps Summers had entered into, which imploded when rates fell instead of rising.”

Posted by TS | Report as abusive

The real travesty is that Summers did all this prior to resigning from the Harvard presidency in 2006, while Harvard lost its billions in 2008. So there was no president of Harvard for two years!Or wait, was there….?

Posted by Anonymous | Report as abusive

Felix … do some research and stop wasting web space with your platitudes. The Harvard endowment has an Investment Committee; they approved all asset allocations and investments. Summers DID NOT sign a single Investment Memorandum Agreement with any external investment managers. The fault of bad decisions and bad investments rests entirely with the overpaid CIOs and their overly aggressive (and wonderfully ‘accommodating’) Investment Committee. Blame,in order: Meyer, El-Erian, Kaplan, and Mendillo.

Posted by Viyada York | Report as abusive

Felix, it’s not like Harvard is ever going to run out of cash. They can always raise a few hundred million in short-term loans if push comes to shove. Summers was right; Harvard made money on its cash and is much better for it.What’s wrong with making money?

Posted by Myles SG | Report as abusive

And I mean, Harvard of anybody doesn’t need like 5 billion in cash. it was completely excessive. Investing it to make more money was the right way to go.

Posted by Myles SG | Report as abusive

Don’t think I or DeLong have all the numbers or the interest in finding them, but the 27% loss is on a higher amount than the gains.I made 20% a year for 7 years, but then lost 100% in one year puts me at ZERO dolars, not a 40% gain. Likely DeLong’s point holds, but not as clear – efficient market would say close to equal?Actually saw reports that Harvard was OVER 100% invested, with private equity commitments, hedge funds and had “margin calls” in addition to needing cash to pay the bills.

now I remember, it was Forbes that said that Harvard was 105% invested  /harvard-endowment-failed-business_harv ard.html”It would have been nice to have cash on hand to meet margin calls, but Harvard had next to none. That was because these supremely self-confident money managers were more than fully invested. As of June 30, they had, thanks to the fancy derivatives, a 105% long position in risky assets. The effect is akin to putting every last dollar of your portfolio to work and then borrowing another 5% to buy more stocks”

If they could rely on short term borrowing to compensate for cash shortfalls they would likely come out ahead. If they couldn’t they were taking on too much risk.

Posted by Lord | Report as abusive

Summers can also take comfort in the envelope theorem where even if he is off from the optimal asset allocation a bit near the maximum, the “flat” loss near the maximum is not that much.I think both I and Delong ignored compounding in the back-of-the envelope calculations.What strikes me as odd if true in this story, man-bites-dog, is given the different incentives of a college administrator vs. fund manager, why was Summers seeking more risk than his fund manager was seeking?

Re: Steve Baba.Touche…Percentages don’t add, but log percentages do.In logs, a 27% loss is a -0.315, while an 8% relative gain is a +0.077. So the actual log change is 6 x 0.077 – .315 = 0.147 in logs, or a percentage change of +15.8% rather than +21%

Posted by Brad DeLong | Report as abusive

Vanity Fair also had a long article suggesting that Harvard was over 100% invested: ics/features/2009/08/harvard200908?curre ntPage=1“Is Harvard desperate now? One clue is this: last December, the university sold $2.5 billion worth of bonds, increasing its total debt to just over $6 billion. Servicing that debt alone will cost Harvard an average of $517 million a year through 2038, according to Standard & Poor’s.”"To be clear, even if you’d tried hard, you could not have picked a worse time to sell bonds than December 2008; that was the precise moment when credit markets seized up. But Harvard, it seems, had no choice. Unwilling to sell its assets at fire-sale prices, it needed immediate cash to cover, among other things, what my sources say was approximately a $1 billion unrealized loss from interest-rate swaps. That’s a staggering figure: $1 billion, roughly a third of the university’s entire operating budget for last year.”

Seems like someone else informed Summers.From Dec. Vanity Fair: features/2009/12/summers-200912?currentP age=1“There were also charges of betrayal from Iris Mack, a former derivatives specialist at the Harvard Management Company (responsible for investing Harvard’s endowment) and the second black woman to receive a doctorate in applied mathematics at Harvard. Mack claims that soon after she started working at Harvard Management, in early 2002—after a stint at Enron—she became uncomfortable with the lack of understanding she thought her colleagues had with the risky derivatives they were investing in. (She was proved correct in the past fiscal year, when the endowment dropped 27.3 percent.) On May 12, 2002, she wrote an e-mail to Summers, alerting him to her concerns: “As a proud Harvard alum I am deeply troubled and surprised by what I have been exposed to thus far at HMC, and the potential consequences for my alma mater’s endowment. In addition, I strongly believe that if my fellow alum[s] knew how the endowment is being managed and the caliber of some of the portfolio managers, they probably would not give another dime to our endowment.”

Re: DeLongThe more I think about it, which is not much, it’s unfair to compare cash returns to endowment returns because1) cash funds have to be liquid while endowment funds don’t have to be liquid – giving the endowment a slight advantage in returns from illiquidy (Felix’s objection reworded?) and2) The endowment funds should have taken all the best opportunities they identified (Alpha) first (The “best” private equity limited-subscription firms) leaving less attractive fair-bet or minimal-alpha investments for the additional cash investment (decreasing returns to scale).Over all possible outcomes, investing cash should have left Harvard with just the Beta minus costs of having to borrow in down years, assuming Harvard did not try to do more like outguess the market with interest rate swaps?What makes Summers/Harvard look bad, in my opinion, is he looks and spent like one of these homeowners who thought the good times would never end – even if he still made money on his house.

10 Days later the NY Times reports:
“BOSTON — Harvard announced Thursday that it would indefinitely suspend construction on a high-tech science complex in the Allston neighborhood of Boston because of money problems.”

Regardless of the optimal policy and slight differences between the optimal and good policies, I don’t think anyone can claim that it was a “good” policy covering all economic conditions?

Posted by SteveBaba | Report as abusive

I would like to point out my passion for your kind-heartedness supporting men who really want guidance on this one matter. Your very own commitment to getting the message around had become extraordinarily beneficial and have in most cases helped women like me to get to their ambitions. Your own invaluable instruction signifies much to me and substantially more to my fellow workers. Many thanks; from all of us.