Kai Nip

After a decade of unsatisfying endowment returns, the Harvard Management Company is reconfiguring its methods to align more closely with those employed by its counterpart at Yale.

In the ten years leading up to 2016, Harvard’s endowment saw an annualized return of 5.7 percent — a figure well below the university’s 8 percent expectation and the 8.9 percent return on the US 60–40 mix of stock and bonds, a traditional low-risk portfolio often used as a market indicator.

In a September letter to the Harvard community, HMC President Nirmal Narvekar, who was chief executive officer of Columbia University’s endowment before joining Harvard in December 2016, vowed to tackle the organization’s “deep structural problems.” He promised to outsource most investments to external managers and to take up a generalist model where staff are compensated based on the entire team’s performance. Both these new approaches have long been key features of what is commonly known as the Yale model. The model was developed by David Swensen and Dean Takahashi, who have led Yale’s investment office for the past three decades.

“Yale has a process for locating and identifying what they feel are good investments, but Harvard’s essentially starting from scratch again with Narvekar,” investment executive recruiter Charles Skorina said.

While Harvard has historically had a larger endowment than Yale has, the gap is quickly closing. On June 30, 2007 Harvard boasted a $34.9 billion endowment while Yale held a $22.5 billion endowment. Ten years later — following two major economic recessions — Harvard’s endowment had increased by just 6 percent, now totalling $37.1 billion, while Yale’s had increased by a full 21 percent, coming in $27.2 billion.

Swensen and Takahashi have been working on Yale’s endowment investment strategies for thirty years, but the leadership of HMC has been in constant flux since former president Jack Meyer left his job in 2005, after 15 years of service. Between 2005 and 2016, three different people sat at the helm of HMC.

Roger Ibbotson, a hedge fund manager and Yale School of Management professor, said changing leaders and, with them, strategies may cause “distortions.”

“When you have a chief investment officer who is steering the ship and implementing a number of different investment ideas, themes and exposures he or she believes in, you want to make sure their tenure is long enough to see those investments through,” said Michael Chase, a partner at consulting firm Fiduciary Investment Advisors who is familiar with Narvekar’s work.

Even though he has never worked at the Yale Investment Office, Narvekar is said to share the University’s investment philosophy.

“Generally the model that Yale thinks it’s best to use is to use outside money managers,” Chase said. “And Narv has kind of grown up in that same school of thought. He was doing the same thing when he was at Columbia.”

Investment experts interviewed by the News detailed a variety of reasons why using only external money managers is a good investment strategy.

For one thing, external money management gives endowment officers more flexibility, Ibbotson said.

“It’s harder to fire [an] old team whereas you can easily fire [a] manager team,” he explained. “It’s harder to be unbiased when you judge them when they are your next-door neighbors.”

According to Chase, outsourcing money management better aligns with the interests of money managers, who prefer greater access to clients and the opportunity to grow their portfolios rather than taking directions from the chief investment officer. He added that the benefits of external money management outweigh the cost of any extra service fees.

Using external money managers also better allows for the adoption of a generalist model, which aligns compensation for in-house staff with team-wide results. Such a model generates more camaraderie among staff and promotes discussions across asset classes, according to Chase.

“It’s bad morale when you have such star professors who are paid much less and who also see sometimes mediocre performance [of the endowment],” Skorina explained.

Because luck plays an important role in investment results, Ibbotson said, staff should not be compensated based on the performance of just their assigned part of the portfolio. He added that models based on individual performance can lead to paying some staff huge compensations even if the endowment sees poor overall returns.

According to Skorina, it may take three to five years for Narvekar to reset Harvard’s endowment portfolio.

“You don’t just wave your hand and change investments. You have to look at the contract that you signed for each investment,” Skorina said. “When you have the money out, you have to have something new to invest it in, and that takes work and efforts.”

Kevin Quirk, a principal with Deloitte Consulting, said Harvard’s restructuring reflects a broader trend in the endowment industry to rethink investing practices.

According to Quirk, between the 1980s and early 2000s — when Harvard initially rose to prominence for its endowment investment — the market environment was far more “friendly” than it currently is.

“What has happened since the financial crisis is an existential self-assessment of many endowment structures and teams,” Quirk said.

He added that the current capital market — with its rising interest rates and hard-to-come-by valuations — presents significant challenges for endowment investors.

Still, Chase said he is hopeful that Narvekar will reinvigorate Harvard’s endowment.

“Narv has a very good track record,” Chase said. “[He] has the capability that is needed to realign what their objectives are and make sure their portfolios get them where they want to be.”

Jingyi Cui | jingyi.cui@yale.edu

  • Nancy Morris

    Very disappointing article.

    It is entirly misleading to describe the Yale model as: “Generally the model that Yale thinks it’s best to use is to use outside money managers.” Yes, Yale uses outside managers and a relatively lean staff of “generalists.” But those are a relatively small parts of what makes Swensen’s team work as well as it does. Finding and maintaining relationships with the RIGHT managers who invest in the RIGHT MIX of assets is a bigger part: Separating superfical diversity (things that “look different”) from true, valuable diversification is not easy. Not easy at all.

    There is also what one might call a “character” issue involved. Any endowment filled with illiquid assets – as any really diversified endowment must be – relies on a process of ANNUAL APPRAISAL of those illiquid assets, and appraisals are easily distorted. The Yale model requires a firm commitment to employing only highly conservative appraisals in determining performance. Anything else eventually brings nothing but heartbreak. Concealed losses may salve egos and budgets in the short run, but they are soon disastrous. There is much evidence that the Harvard endowment has had very weak character in this sense … including its spectacular $1 Billion one-time write-down of illiquid natural resources assets just disclosed. Those assets almost certainly did not decline $1 Billion in ONE YEAR. That write down almost certainly represents belated recognition of long concealed losses. And there is evidence of a lot more of that kind of thing to come, including Narvekar’s recent admission that the Harvard endowment continues to have “deep structural problems.” Yes, huge unrecognized losses ARE “deep structural problems.” VERY deep. Refusals by the best outside managers to allow the Harvard endowment to participate in the best deals (something Narvekar also admits to be a problem for him) also counts as a “deep structural problem” in another sense.

    Then there is the trick of maintaining sufficient liquidity to protect the university from unexpected shocks without sacrificing the returns only achieveable with a largely illiquid, diversified endowment. During the recent financial crisis Harvard was pushed to the brink of insolvency by its failure to deal adequately with that issue. Not so, Swensen.

    Historically there has also been a perverse insistence by the Harvard president to directly meddle in endowment decisions, which is completely nuts. One might also consider intrusive and destructive Harvard alumni activists, including the notorious Class of ‘69, to be informal but permanent parts of the “Harvard Model.” Financing a university whose community consists largely of arrogant monomaniacs faces unusual obstacles.

    There is more. Much more. Why doesn’t the YDN try actually talking honestly to Swensen & Co. about such things, or at least watching his excellent videos explaining what they do?

    • SVV

      Because Yale enjoys snubbing Swensen.

      • Nancy Morris

        Snub Swensen? Yale is naming a building after him and in the past has forced him to accept big pay hikes. Yale LOVES Swensen!