The Purposes and Uses of Endowment

Paul Jansen, William Massy, Henry Riggs, and Timothy Warner

Take Aways

  • Prior to the 1950s, the rule for permissible spending from endowment was very simple: spend yield, all yield, and only yield.
  • Today, nearly three-quarters of all colleges and universities target their endowment spending at about 5 percent of a three-year rolling average of total endowment market value, reflecting the traditional conservative approach of trustees whose fiduciary responsibilities push them in the direction of endowment preservation.
  • Commonfund’s 2006 Benchmarks Study found that the 40 institutions with endowments greater than $1 billion earned an average three-year return of 11.6 percent. Yet the average payout for this group was just 4.3 percent for fiscal year 2005, a drop from previous years.
  • Perhaps the most compelling (although least inspiring) reason to reconsider endowment spending is the risk of legislative intervention or donor backlash.

The Forum for the Future of Higher Education’s Master Class for the past three years has been considering new ideas about spending from endowment. Spurred initially by the work of Forum scholar Perry Mehrling, professor of economics at Barnard College, the discussion has challenged some long-standing views about how much trustees should allocate for current operations and how much they should reinvest so as to preserve endowment for future generations. The Forum’s 2005 Master Class discussion flowed from work by Paul Jansen, director of McKinsey & Company’s nonprofit practice, and was led by William Massy, former chief financial officer at Stanford University, Henry Riggs, president emeritus of Keck Graduate Institute, and Timothy Warner, vice provost for budget and auxiliaries management at Stanford University. They addressed the fundamental question about whether traditional management practices that favor endowment preservation are equitable to current students, faculty, and others who benefit from higher education’s endeavors as well as to future generations, and they suggested several refinements to the processes by which spending rules are typically developed.

A Brief History of Spending Rules

Prior to the 1950s, the rule for permissible spending from endowment was simple: spend yield, all yield, and only yield. One could spend all dividends, interest, and rents received. And because capital appreciation is not yield, all such appreciation was automatically reinvested. Whether the appreciation was realized or unrealized, or positive or negative for that matter, made no difference.

The great bull market of the 1950s triggered the first round of out-of-the-box thinking about endowment payout. The new breed of investment managers tilted portfolios toward equities, and growth stocks in particular, which offered greater total return. It soon became apparent, however, that while a diversified portfolio of bonds and equities produced more endowment wealth, it might not be immediately beneficial for budgets supported by the endowment. Such a portfolio could in fact be detrimental, if only in the short run, because yield on growth stocks can easily be less than on a portfolio consisting mainly of more conservative holdings. Investment managers were pitted against budgeters—and not for the last time. Something had to give.

What gave was the idea of spending only yield. Why should budgets be starved while endowment market values were rising? Why spend only 1 or 2 percent of market value? Why, alternatively, should investment managers be pressed to make suboptimal asset allocations in order to generate yield for immediate spending? It took much argument, multiple legal opinions, and an influential report by the Ford Foundation, Managing Educational Endowment (1969), but eventually trustees were empowered to spend at least the realized portion of appreciation. The bonanza was at hand. But was that a really good thing?

Stock market declines soon demonstrated that the answer was no—that something important had been lost when the rules changed. That something was the discipline of reinvestment, a discipline that was so deeply embedded in the yield-only rule as to be almost invisible. The new rules allowed spending rates to approach total returns—Stanford’s climbed to more than 8 percent in the early 1970s, for instance—which left little money for reinvestment. People began searching for new principles to determine spending, ones that would provide the discipline needed to drive reinvestment without distorting asset allocation.

Enter the ideas of purchasing power maintenance and its close cousin, intergenerational equity. Because total return was too high a spending target and yield both too low and perverse in terms of its effect on asset allocation, the challenge was to find an appropriate balance that did not depend on yield. Purchasing power maintenance became the standard for “appropriate” in spending formulas. The standard was in wide use by the mid-1970s. The devil is in the details of any such formula, however, and 30 years of refinement have yet to establish consensus among either theoreticians or practitioners. And although they generally agree that spending rules should govern most of the time, a lively debate persists about when and what kind of exceptions should be allowed.

To address the exceptions question, note what is not included under the spending rule rubric: The rule does nothing to inform trustees about the trade-off between financial investments and those in, say, bricks and mortar or human capital. Institutions regularly transfer expendable funds to quasi-endowment, which includes funds such as accumulated operating surpluses and reinvested yield that are added to the true endowment by trustee action. Withdrawals from endowment to fund construction or make new program investments are less common, but they are certainly permissible providing applicable donor covenants are met.

Institutions that add to quasi-endowment do so because they feel the benefits of a larger endowment outweigh those of current spending. They are enhancing the endowment’s purchasing power, not maintaining it.

Institutions that add to quasi-endowment do so because they feel the benefits of a larger endowment outweigh those of current spending. They are enhancing the endowment’s purchasing power, not maintaining it. A larger endowment insulates the institution from the student-tuition and sponsored-research markets. A larger endowment also confers prestige on the institution and, as some critics have charged, psychic benefits on trustees.

Conversely, there may be times when financial investments should be shifted to other uses. Disasters such as Hurricane Katrina present an obvious case—what are financial assets for if not to serve on a rainy day? Or suppose that an institution can move to the next level of quality, but the opportunity requires an up-front investment in faculty? In that case, spending now will produce significant and persistent future value, so why shouldn’t the endowment be tapped? At the level of principle, the only problem is to separate the exceptional problem or opportunity from the myriad of day-to-day spending demands that, if acceded to, will fritter away the endowment.

One job of trustees is to lean against the demands for immediate spending that characterize any dynamic university. Doing that job well requires a disciplined spending rule. But trustees also should be alert to opportunities for effecting extraordinary improvements in their institution’s academic prowess or market position. When the case for such an opportunity is truly made, they are justified in setting aside the spending rule for a limited time and purpose. Finally, trustees should insist on effective financial engineering with respect to the endowment. For example, it may be better to borrow at tax-exempt rates than to liquidate endowment investments to fund building construction. The point is not that extraordinary spending or borrowing for construction are good or bad per se, but that maintenance of purchasing power is just one consideration in strategic planning for the endowment.

Today, nearly three-quarters of all colleges and universities target their endowment spending at about 5 percent of a three-year rolling average of total endowment market value, reflecting the traditional conservative approach of trustees whose fiduciary responsibilities push them in the direction of endowment preservation.

Intergenerational Inequity?

Clearly, the notion of intergenerational equity dominates the mindset of trustees today, as evidenced by their endowment management practices. Verne Sedlacek, president and chief executive officer of Commonfund, has defined intergenerational equity as “the state in which the nominal market value [of the endowment] is equal to or greater than the inflation-adjusted market value from one generation to the next.” Perry Mehrling’s alternative approach to endowment management would refine that definition by saying that the endowment’s inflation-adjusted value should simply be equal from one generation to the next, rather than growing larger over time. Mehrling argues that the common practice of adding investment returns in excess of spending back into the endowment works only if the spending rate is so low that excess returns are always ensured. He believes this is because the endowment corpus is ratcheted up when returns exceed spending but not ratcheted back down when returns fall short of spending. If that were true, every new level of achieved endowment accumulation would become the new perpetual goal. The demands of intergenerational equity might even seem to require that any shortfall be made up. While many students of endowment spending rules don’t agree that the corpus of endowment is ratcheted upward, Mehrling’s general point is worthy of consideration.

A policy of holding current spending below the expected rate of return shifts all the risk involved in future asset returns onto present shoulders, and none of it onto future generations. Mehrling wonders what is equitable about that approach, and he encourages consideration of increased current endowment spending under certain conditions. (For a complete discussion of Mehrling’s proposed Alpha-Beta approach to endowment spending and responses to it, see http://www.educause.edu/ir/library/pdf/ffp0516s.pdf.)

Paul Jansen agrees with Mehrling’s conclusion that endowment spending driven by maintaining so-called intergenerational equity is too conservative. He notes as well that under that approach the current generation bears all the risks of market volatility, and he suggests that equity needs to mean something closer to “equitable given the best information at the time of the [investment] decision,” which indeed is the only thing a trustee can influence. However, even defining equity as an equal probability of greater or lesser future value falls short of achieving true intergenerational equity because current methodologies for developing spending rules ignore the impact of future gifts on endowment value over time—even though, of course, the likelihood of a stream of future gifts is high.

Thus Jansen proposes another definition of intergenerational equity, one based on a “payout rate for which there is an equal chance of greater or lesser inflation-adjusted value, with an acceptable level of uncertainty, taking into account expectations for future gifts.” Bill Massy agrees, as he notes that future generations benefit from additional total returns, so why shouldn’t they share in the cost of uncertainty? Holding them harmless would unbalance the allocation of costs and benefits.

Tim Warner and Hank Riggs note the massive run-ups in endowment values of the top few dozen U.S. colleges and universities and question their ongoing pursuit of even greater endowments. Stanford, for example, is gearing up a campaign with a target of more than $4 billion, notwithstanding its current endowment of about $15 billion. Commonfund’s 2006 Benchmarks Study found that the 40 institutions with endowments greater than $1 billion earned an average three-year return of 11.6 percent and have an average return assumption of 8.8 percent for the next three years. Yet the average payout for this group was just 4.3 percent for fiscal year 2005, a drop from previous years—and lower than spending rates among all other size categories except the smallest, with endowments of less than $10 million.

The Purposes of Endowment and Higher Education’s Mission

Riggs decries today’s prevalence of ambitious facilities construction and student amenities such as sumptuous food services, opulent student dormitories and lounges, and exercise facilities that rival exclusive private clubs that proliferate on wealthy campuses—while facilities at poorer institutions accumulate deferred maintenance. At the wealthy institutions, student financial aid plans are being further sweetened as the family income threshold for aid rises, while poorer colleges succumb to deep tuition discounting to maintain enrollment or increase student quality.

Could colleges and universities spend their endowments in ways that promise broader societal benefits? Riggs notes that indeed a small fraction of endowments will be spent that way—for example, to pursue medical research, advances in science and engineering, or improved K–12 education. He believes that institutions with vast endowments could do far more and suggests increasing enrollments—on their primary campus or on branch campuses, either in the United States or in underserved parts of the world. Further, he asks whether quality education programs could be delivered electronically to a larger fraction of the public, noting MIT’s moves in that direction. In Riggs’s view, those are the only possible ways such institutions can simultaneously and honestly pursue both access and excellence. Finally, Riggs also suggests the development of innovative programs to entice more U.S. students to pursue science, engineering, and mathematics degrees.

Jansen believes that increasing graduation rates is critical. More students drop off the path to a four-year degree in college than to a diploma in high school. Although quality of preparation, life circumstances, maturity, and other factors are no doubt involved, almost 30 percent of every class of eighth graders will drop out at the college level. Clearly, society benefits if more qualified students graduate. Recent moves at some institutions to convert loans to grants mean hundreds of students may gain access, but the effect will be incremental and will not significantly alter the productivity decline. Students need new support systems, time and attention, and a culture that makes their success a shared objective. That takes money, but net productivity improvement is clearly achievable.

In a globalizing economy with new sources of competition, universities need to look for ways to break the paradigm of squeezing a bit more out of the same organization or supporting largely disconnected research initiatives that do not add up to meaningful impact or reputation building.

Technology could also help reduce administrative costs and expand student bodies without increasing faculty size, thus improving productivity. However, real barriers to productivity improvement remain, such as a history of sterile technology investments and organizational resistance to change. But new models and mindsets are essential. Endowments permit investments that can reposition an institution in a way that fancier student centers or new dorms simply cannot. In a globalizing economy with new sources of competition, universities need to look for ways to break the paradigm of squeezing a bit more out of the same organization or supporting largely disconnected research initiatives that do not add up to meaningful impact or reputation building. Echoing Riggs, Jansen asks why there isn’t a Harvard in India, or a multidisciplinary initiative to reform the teaching of math and science in U.S. schools?

The Potential for Backlash

Perhaps the most compelling (although least inspiring) reason to reconsider endowment spending is the risk of legislative intervention or donor backlash. Voters and politicians eye with suspicion—and perhaps with taxation in mind—excessive endowments of, say, $10 billion at major research universities or $1 billion at small liberal arts colleges. Do those institutions really need enormous stockpiles of resources when they have no announced intention of serving a broader segment of society or more students? At a time when the top five managers at Harvard Management Company together are paid $100 million in a single year and the additional revenue that federal, state, and local governments would generate from just a 1 percent tax on Harvard’s endowment—approximately $220 million—is widely understood, we in the academy will have an increasingly difficult time arguing that we are not greedy. Greedy institutions and individuals are politically vulnerable.

Indeed, Harvard’s success may well be the case that prods legislators to act. From 1997 through 2002 (the date of the last available IRS Form 990 from Harvard), this nonprofit produced a cumulative profit of more than $10 billion. Its current endowment of about $27 billion, driven by Harvard’s astute investing, spins off more than $2 billion per year, tax free. While the Bill and Melinda Gates Foundation also has an endowment above $20 billion, it operates under legislative restrictions and gives more than $1.3 billion each year to causes such as immunizing poor children, searching for an HIV vaccine, reforming urban schools, and helping the poorest students attend college. Harvard’s “cause,” too easily characterized as educating the elite in increasingly luxurious facilities, might not fare so well in the hands of politicians looking to make a mark and win an election.

The nonprofit sector’s cumulative trillion-dollar investment account can yield almost $100 billion in annual returns. At a 20 percent capital gains tax rate, that’s $20 billion to fund the many items on politicians’ wish lists. Throw in the budget strains from Katrina or Iraq—along with a few more scandals involving administrator compensation or expense levels—and the unsuccessful effort a decade or so ago for a “tax on capital” aimed at endowments could well gain more support. More recently, in 2003 a House bill sought to raise foundation distributions by disallowing expenses from payout calculations. It too was unsuccessful, but the threat of legislation governing higher education institutions becomes greater as their endowment levels reach astounding heights and they continue to pay out at a rate below the 5 percent minimum legally mandated for endowed foundations.

A second source of risk lies with donors, who are becoming more demanding about the use of their gifts. Charity-rating agencies are on the rise and increasingly ask questions such as “How much working capital does the charity hold?” A prudent level is a plus, but too much can be a negative. Large endowments are bound to make donors wonder, “If you can’t spend what you have now, why do you need my gift?” While many donors are attracted to the immortality aspects of an in-perpetuity naming gift, an equal number are frustrated by low endowment payout rates and want their gifts to make a difference now. The rising share of philanthropy captured by higher education suggests that donors haven’t yet become disillusioned with giving to wealthy institutions, but will we be able to see the signs of a potential backlash before it is too late?

Recommendations

There is no question that spending rules should be established based on very long time horizons and rigorous financial analyses. But the conservative bias toward endowment expansion should be revisited. There is much that trustees can do now to help chart a more strategic course that focuses on spending endowment funds for the greatest benefit in an increasingly competitive global environment. Here are a few recommendations:

Payout rates should take future gifts into account. Expectations for future gifts should appear explicitly in payout formulas because their omission results in considerable real asset growth and, therefore, generational inequity. In particular, bequests and various deferred trusts already in place should be recognized. The current generation benefits from none of these, yet current budgets support fundraising and alumni relations expenditures designed to ensure future gift flows. Simply put, the current payout formulas, which operate on the assumption that the stream of new endowment gifts will suddenly dry up, should be revised.

The promise of “in perpetuity” for endowment gifts should be reconsidered. Fundraisers and trustees seem to view in-perpetuity gifts as the gold standard. Given that operating budgets are always tight, might it make sense to encourage donors to make their gifts expendable—not in a single year, but perhaps over a 10-year period—rather than to “true” endowment? If we believe that the opportunity being presented to the donor is of high priority, then why not front-load the benefit to the early years of the project? Is it honest to tell donors that the opportunity they are considering will remain a high priority in perpetuity? If so, then our institutions lack dynamism in the worst way. Given the current pace of change in our society and higher education’s research and teaching interests, both society in general and institutions in particular would be better served if endowments had a shorter term than in perpetuity.

Proposed institutional investments should be subjected to discounted cash flow (DCF) analyses. The usefulness of analyzing proposed investments using well-accepted DCF techniques is widely agreed upon, yet they are seldom used in analyzing higher education’s investments. In nonprofits with reasonable endowments, an alternative use of funds—the endowment—is always available, and analysts can estimate rather well what the long-term return on endowment investments will be. Colleges and universities tend to construct buildings that will have substantially longer lives than run-of-the-mill commercial or industrial buildings. That may well make sense, given their lower cost of capital, but a more rigorous financial analysis would almost certainly better inform such decisions. If a lower-investment option to accomplish a certain objective is selected, it can be reasonably assumed that the difference can be invested in the endowment with a quite predictable return. That return is the rate at which the cash flows of the project should be discounted.

Why aren’t DCF analyses more frequently used? Probably because in private higher education essentially all investment funds come in the form of gifts, except for the minor amounts that are borrowed—at subsidized, tax-free rates. Are we assuming that these gift funds have an essentially zero cost of capital? And is that an appropriate way to treat resources that donors and taxpayers entrust to us? Well-endowed institutions—indeed all institutions, although the rest are less likely to overspend because they can’t—have an obligation to donors and to society to seek higher-return investment opportunities than, for example, those flowing from lavish facilities and student amenities.

Know where the money is. For institutions whose actual payout rates have been below their target rates, it would be a safe bet that most don’t know how much has been reinvested. Knowing that number and tracking it is important. With that data, trustees can consider from a more informed basis the establishment of funds to make excesses available for onetime projects or to address extraordinary conditions. Yale’s deferred maintenance challenge (a long-term problem); Harvard’s Allston campus initiative (an opportunity); and expenditures to recover from a hurricane (an emergency) are examples of extraordinary conditions. Further, the buildup of accumulated restricted funds should be addressed. In some institutions, unspent endowment income has accumulated in highly restricted funds. Trustees need to be creative and aggressive in trying to use such funds. In some cases that may mean going to court; more likely, however, it simply means being proactive in reviewing the accounts and their accumulated balances.

Conclusion

From a capital markets perspective, it is important to emphasize the need for a sustainable endowment spending level based on a rigorous study of expected returns and volatility across asset classes over long periods. In evaluating the financial health of endowed institutions, bond investors and rating agencies look for spending policies that sustain the real value of endowment assets over time. That said, these capital market participants also recognize the role of an institution’s overall strategy in the development of the spending policy—not to mention the need to make adjustments to the policy over time as conditions warrant.

The case for rethinking payout policy is compelling. The core arguments in favor of doing so revolve around a more holistic view of intergenerational equity, the need to manage the growing risks associated with ever-increasing asset accumulation, and the strategic need to thoughtfully consider investment opportunities. Simply hoarding endowment for the future is an increasingly suspect strategy. Surely the higher education arena is replete with attractive and exciting intellectual and capital investment opportunities that will benefit students, faculty, and society—now and in the future. Undoubtedly, all will benefit when the nation’s wealthy institutions’ aspirations are as big as their endowments.


Paul Jansen is the director of McKinsey & Company’s global nonprofit practice. Jansen can be reached at paul_jansen@mckinsey.com.

William Massy is president of the Jackson Hole Higher Education Group. Previously he was the chief financial officer at Stanford University. His most recent book is Honoring the Trust: Quality and Cost Containment in Higher Education (2003). Massy can be reached at wmassy@jhheg.com.

Henry Riggs is the retired founding president of Keck Graduate Institute of Applied Life Sciences. His most recent book is Financial and Cost Analysis for Engineering and Technology Management (2004). Riggs can be reached at henryriggs@comcast.net.

Timothy Warner is vice provost for budget and auxiliaries management at Stanford University. Warner can be reached at trw@stanford.edu.