The collective assets under management (AUM) for the 17 largest U.S. university endowments have surged from approximately $249 billion in fiscal year (FY) 2016 to a remarkable $426 billion by FY 2025. This substantial growth, representing an increase of about 70-73% depending on rounding, is a testament to both robust investment performance and the consistent influx of new gifts. However, beneath this aggregate success lies a more nuanced narrative, characterized by three distinct allocation philosophies, differing risk profiles, and a structural efficiency gap that becomes evident when these endowments are benchmarked against their global institutional peers. This analysis, drawing from the 2025 NACUBO-Commonfund Study of Endowments, individual fund annual reports, and sophisticated multi-factor statistical risk modeling, delves into what the performance and strategies of these top endowments reveal about allocation, risk management, and overall investment efficiency.
A Decade of Growth: Strong Returns, Divergent Long-Term Leaders
Over the past decade, the aggregate AUM of the top 17 U.S. university endowments has expanded significantly. From roughly $247 billion in FY 2016, these institutions have grown their portfolios to approximately $426 billion by FY 2025, achieving an impressive compound annual growth rate. This expansion is a direct result of strong investment returns and, crucially, sustained philanthropic contributions.
Examining individual performance, the University of Michigan stood out in FY 2025, reporting a net return of 15.5%. Close behind were the Massachusetts Institute of Technology (MIT) at 14.8%, Washington University in St. Louis (WashU) at 14.7%, and Stanford University at 14.3%. Looking at longer time horizons, Michigan, Stanford, the University of Notre Dame, WashU, and MIT have all demonstrated annualized returns exceeding 12% over the past five years. The ten-year horizon sees MIT, Michigan, Stanford, Notre Dame, and Duke University leading the pack. Notably, Harvard University, despite holding the largest AUM at $57 billion, registered a ten-year annualized return of 8.2%, placing it at the lower end among the top five by asset size. This divergence highlights a critical observation: the institutions achieving the best one-year returns are not always the most consistent long-term compounders. Strategic allocation and risk management play a pivotal role in sustained wealth creation.
Table 1: AUM Growth and Net Returns: Top 17 Endowments (FY2016-FY2025)
(Image of Table 1 would be inserted here, displaying AUM and net returns for each of the top 17 endowments over the specified period)
Note: Source: 2025 NACUBO-Commonfund Study of Endowments; individual endowment annual reports. Returns are net of investment management fees and exclude distributions and operating expenses. AUM growth reflects investment returns and net new gifts. Some figures estimated where not publicly disclosed.
Three Philosophies: Unpacking Distinct Allocation Strategies
The investment landscape of these elite endowments is not monolithic. A closer examination of three institutions—Harvard, Yale, and Michigan—reveals the spectrum of strategic philosophies employed. These institutions, representing the largest endowment, the progenitor of the diversified alternatives model, and a top recent performer, respectively, underscore that there is no single blueprint for long-term outperformance.
Harvard University: The Concentration Model
Harvard’s endowment exemplifies a "concentration model," with a significant 41% allocation to private equity and 31% to hedge funds. Notably, it maintains no direct allocation to venture capital. This strategy embraces substantial illiquidity and manager concentration, aiming to capture the higher return premia often associated with private markets. While Harvard has shown strong recent performance, its ten-year return of 8.2% reflects the challenges and costs associated with earlier strategic transitions, even in the face of robust recent one-year gains of 11.9%. This approach prioritizes the potential for outsized returns from specific asset classes, accepting the inherent risks of illiquidity and concentrated manager risk.
Yale University: The Diversified Alternatives Model
Yale University’s endowment, under the pioneering leadership of David Swensen, has long been synonymous with the "diversified alternatives model." Its current allocation—24% in venture capital, 20% in private equity, and 22% in hedge funds—illustrates a broad diversification across alternative asset classes. This model, which has profoundly influenced institutional investing, emphasizes diversification to achieve consistent compounding returns over extended periods. Yale’s 11.1% one-year return and a solid 9.4% ten-year annualized return are indicative of the enduring success of disciplined diversification strategies in navigating market cycles.
University of Michigan: The Growth-Concentrated Model
The University of Michigan’s endowment, the top performer in FY 2025 with a 15.5% net return, showcases a "growth-concentrated model." This strategy features a substantial 33% allocation to venture capital and 11% to private equity, complemented by 13% each in hedge funds and real assets. The impressive 13.7% five-year and 10.4% ten-year annualized returns suggest that this strategic posture is not a fleeting outcome of market exuberance, such as the recent AI boom, but rather a durable approach built over multiple market cycles. This allocation prioritizes high-growth potential sectors, particularly in early-stage companies, while maintaining a diversified base.
Table 2: Three Endowment Case Studies: Harvard, Yale, and Michigan (FY2025 Standardized Allocation Categories)
(Image of Table 2 would be inserted here, detailing asset allocations for Harvard, Yale, and Michigan)
Note: Source: FY2025 institutional reports and APT/UMass standardized allocation mapping. Categories are presented in comparable asset-class buckets and may differ from each institution’s exact public-report labels. Rows may not sum to 100% due to rounding and category mapping.
Diversification’s Nuances: Real Assets and Hedge Funds in Focus
While the broad embrace of alternative investments is a hallmark of these endowments, the degree to which different alternatives contribute to risk reduction varies significantly. Sophisticated risk modeling reveals that not all alternatives offer the same diversification benefits.
Table 3: Selected APT Correlation Estimates Across Asset Classes (APT Model)
(Image of Table 3 would be inserted here, showing correlation estimates between various asset classes)
Note: Source: APT model, UMass Amherst Endowment Research Project (2026). Real Assets is an approximate composite based on several real asset sub-categories, so pairwise values should be interpreted as selected model estimates rather than a mathematically complete symmetric correlation matrix. Full model correlation output is available in the underlying workbook.
The analysis of ex-ante correlation estimates from the APT model provides crucial insights. Private equity exhibits a correlation of 0.71 with U.S. equities, and venture capital a correlation of 0.52, both retaining meaningful equity beta. Similarly, real estate and private energy assets carry comparable exposures. In stark contrast, hedge funds, as modeled by the Global HFR Index, demonstrate a much lower correlation with U.S. equities (0.25) and non-U.S. equities (0.12). Commodities, a component often considered within real assets, show a negative correlation of -0.05 with hedge funds.
This distinction is critical: even with allocations exceeding 60% in alternatives, approximately 86% of the ex-ante volatility across large endowments still originates from the equity factor. This implies that endowments have not eliminated equity risk; rather, they have shifted its manifestation towards manager selection, market timing, and the inherent illiquidity of their investments.
Risk Profile: Higher Returns, Yet Lower Risk-Adjusted Efficiency
When evaluating performance, raw returns alone can be misleading. A comprehensive assessment requires an examination of risk-adjusted returns. The largest 17 U.S. endowments, categorized as the "Over $5B" cohort, have indeed outperformed U.S. and Canadian pension peers over one, five, and ten-year horizons, with average net returns of 12%, 11%, and 9% respectively. However, this outperformance comes with higher ex-ante risk, estimated at approximately 11%, and a greater potential for losses, with an approximate 20% drawdown in a 1-in-20-year event.
Consequently, the risk-adjusted performance metrics, such as the Sharpe Ratio and Information Ratio, reveal a gap. The Over $5B cohort exhibits a Sharpe Ratio of 0.56 and an Information Ratio of 0.76. This stands in contrast to the Canadian "Maple 8" benchmark, which achieves a Sharpe Ratio of 0.90 and an Information Ratio of 1.43.
Table 4: Ex-Ante Risk, Net Returns, and Risk-Adjusted Performance by Endowment Plan (APT Model Output)
(Image of Table 4 would be inserted here, detailing risk metrics and performance ratios)
Note: Source: APT model output, UMass Amherst Endowment Research Project (2026). Net returns are from 2025 NACUBO-Commonfund Study and individual annual reports. Ex post volatility is 10-year realized. Sharpe Ratio uses the workbook’s 3.1% risk-free rate. Information Ratio is reported directly from the APT workbook. Ex-ante volatility and systematic/specific risk from the TLA_ST tab.
The Sharpe ratio range across the 17 endowments spans from 0.38 for WashU to 0.74 for the University of Texas System. The Canadian Maple 8, a benchmark of leading Canadian pension plans, consistently outperforms with a Sharpe ratio of 0.90. WashU’s high ex-post volatility over ten years (19.5%) explains its lower Sharpe ratio of 0.38, despite a competitive 9.7% ten-year return. Conversely, Cornell University and Columbia University, despite lower headline returns, demonstrate the lowest ex-ante volatility within the individual endowment group (9.1% and 9.6% respectively) and achieve above-average Sharpe ratios relative to peers with higher raw returns. The significant difference in Information Ratios—1.43 for the Maple 8 versus 0.76 for the Over $5B U.S. endowment cohort—serves as a clear indicator of this persistent efficiency gap.
Stress Tests: Unpacking the Dominance of Equity Risk
To further understand the risk profiles, stress tests were conducted to assess potential losses under various market scenarios. The findings reiterate the pervasive influence of equity risk. Despite substantial diversification into alternatives, 86% of ex-ante volatility across large endowments is still attributable to equity factor risk, leading to significant losses in adverse scenarios.
In a scenario mirroring the COVID-19 pandemic’s initial shock (March-April 2020), most endowments experienced losses ranging from 21% to 27%. This period was characterized by a rapid, global liquidity crunch, where even diversified portfolios struggled to fully insulate against the broad sell-off in equities.
However, during a hypothetical repeat of the early 2022 inflationary environment, endowments generally fared better than their pension fund counterparts. The Over $5B cohort experienced an approximate 8% loss, whereas benchmark portfolios lost between 22% and 25%. This resilience in an inflationary scenario is attributed to the partial inflation hedging provided by real assets and certain hedge fund strategies.
Table 5: Potential Losses and Factor Risk Attribution by Endowment Plan (APT Model Output)
(Image of Table 5 would be inserted here, detailing stress test losses and factor contributions)
Note: Source: APT model output, UMass Amherst Endowment Research Project (2026). VaR and Average Loss are 1-in-20-year annual loss estimates. Max Drawdown Horizon is 20 days. Factor attribution percentages are selected model factors and may not sum to 100% because Currencies and Other factor exposures are omitted for space. Stress test losses are scenario-based estimates: Financial Crisis (June-December 2008), COVID (March-April 2020), Stagflation (January-October 2022).
Several specific findings are noteworthy:
- Notre Dame: Exhibited the highest equity factor attribution (96%) and a significant stagflation loss of -18.1%, underscoring its public-equity-heavy posture.
- Duke University: Benefited from a 11.7% commodities factor share, the highest in its peer group, contributing to a smaller stagflation loss of -4.3%. This highlights the inflation-hedging capabilities of real assets and natural resources exposure.
- Canadian Maple 8: Their 7.0% duration factor attribution, compared to near zero for most endowments, reflects a higher allocation to fixed income (partially achieved through leverage). This strategy contributed to their lower ex-ante volatility but also a stagflation loss of -10.0%, which was worse than most endowments despite a lower overall volatility profile.
- UC System and Notre Dame: Showed the largest stagflation losses (-15.7% and -18.1% respectively), directly linked to their high equity factor concentrations.
- U.S. Public Pension Benchmark: Experienced the largest stagflation loss (-18.8%), consistent with mandatory fixed income allocations that face mark-to-market losses during sharp rate increases.
- COVID Scenario: The dispersion of losses was less pronounced during the COVID scenario, with most portfolios losing 21-27%, indicating that even alternative investments could not fully offset the impact of a rapid, widespread equity beta shock.
Key Insights for Investment Professionals
The analysis of these top U.S. university endowments yields several critical takeaways for investment professionals:
-
Raw Returns Require Volatility Adjustment: Comparing raw returns without considering risk is insufficient. Institutions like WashU (9.7% ten-year return with 14.1% ex-ante volatility) and Cornell (8.6% ten-year return with 9.1% ex-ante volatility) demonstrate that Cornell’s more disciplined approach yields a more efficient outcome for the risk taken, as evidenced by its superior Sharpe Ratio. Table 4 clearly illustrates this risk-adjusted performance across all 17 institutions.
-
Factor Attribution Overrides Allocation Labels: Understanding how different factors drive portfolio performance is more crucial than simply looking at broad asset allocation categories. Institutions with similar alternative allocations can have vastly different equity factor exposures, which are the primary drivers of losses during financial crises and market shocks. Specific exposures to real assets, commodities (particularly energy), natural resources, and infrastructure are key differentiators in stagflationary environments.
-
Structural Differences in Global Benchmarks: The success of benchmarks like the Maple 8 highlights a fundamental difference in approach. While U.S. endowments primarily use real assets and hedge funds for diversification away from equities, Canadian pensions often employ leverage—through debt, repo markets, and derivatives—to diversify exposures, notably in fixed income. Both strategies can enhance risk-adjusted returns, but they do so through distinct mechanisms with differing risk profiles. The Maple 8’s superior Sharpe and Information Ratios suggest that U.S. endowments could potentially improve their efficiency by exploring balance sheet leverage and direct investment infrastructure, though current governance structures may present significant hurdles to such adoption.
In conclusion, the journey of the largest U.S. university endowments over the past decade showcases remarkable wealth creation driven by strategic investment. Yet, it also illuminates persistent challenges in achieving optimal risk-adjusted efficiency when compared to global peers. The nuanced interplay of allocation philosophy, risk management, and structural differences in investment approaches continues to shape the landscape of institutional investing, offering valuable lessons for all stakeholders.
