Investment Performance of University Endowments

Abstract

University endowments provide long-term support for academic activities. Universities rely on the investment returns of endowments to continuously fund these activities. To pursue better investment performance, university endowments of all sizes have adopted the endowment model, which reduces holdings of public securities and increases allocation to alternative assets such as hedge funds, private equity, commodities, and real estate. This study documents the trend toward increasing allocation to alternative assets and evaluates the investment performance. Large university endowment funds have allocated a higher portion to alternatives and have higher rates of returns. Conversely, smaller university endowments have increased a lower percentage to alternatives and their performance trails that of larger peers, supporting prior studies showing that smaller endowments would achieve better performance by adopting the conventional 60/40 allocation in equity and fixed income strategy. We perform a regression analysis to examine the link between asset allocation and investment performance. The empirical results show that the impacts of equities and alternatives on performance are positive and significant. Furthermore, a comparative analysis indicates that investment returns exhibit high year-to-year volatility while the spending rates are stable and that the average rate of return is higher than the average spending rate.

1. Introduction and Literature Review

University endowments support many important academic activities, such as academic programs and research, endowed faculty positions, student financial aid, and the operations and maintenance of campus facilities. Investment returns are essential to provide stable and long-term funding for these activities. There are several approaches to investing university endowments. The conventional passive allocation model focuses on 60 percent stocks and 40 percent fixed income (60/40 allocation). This approach often utilizes low-cost index funds to reduce fees. In contrast, the Yale model, or the endowment model, moves away from the conventional 60/40 portfolio to include high allocations in illiquid alternative assets such as private equity, venture capital, hedge funds, and real assets. Large endowments tend to manage funds in-house, whereas smaller ones typically outsource to external firms.

To pursue better investment performance, many universities and colleges have modified their investment strategies to reduce public holdings of 60 percent stocks for growth and 40 percent fixed income for income and stability and to increase allocation to alternative assets. Allocation to these alternative assets offers a diversification benefit as alternative assets and public stocks and bonds are not highly correlated. In addition, the illiquidity of private alternative assets adds a premium to long-term portfolio returns.

Observers have cited the successes of mega-university endowments such as Yale, Harvard, and Princeton endowment funds as evidence of higher returns from increasing allocation to illiquid alternative assets. These large endowment funds adopted the endowment model early. They have a higher portion of portfolio invested in alternative assets and they have achieved higher rates of returns. Smaller endowments do not have the expertise and size to access the same top-tier alternative funds, yet they attempt to mimic the strategy, leading to lower performance. As excessive capital has flooded alternative markets in recent years, the premium on alternative investments has shrunk, causing the conventional passive portfolio to outperform endowments heavily weighted in alternatives. Blazek and McMillan (2025) confirm such observations, providing evidence of endowment model underperformance between 2023 and 2024.

Data from the National Association of College and University Business Officers (NACUBO) show that university endowment funds of all sizes have significantly increased asset allocation to alternative assets. The percentage for large university endowments with assets of more than $1 billion increased from 32.0 percent in 2002 to 62.6 percent in 2022. Smaller endowments, with assets between $101 million and $500 million, increased from 14.3 percent to 35.3 percent during the same period. University endowment funds with assets of $51 million to $100 million increased allocation to alternatives from 7.2 percent in 2002 to 20.4 percent in 2022. However, smaller university endowments, with lower alternative exposure, have not been able to replicate the strong investment performance of large university endowments. Other contributing factors to this underperformance include investment expertise, access to high-performing managers, and fee structures. The empirical results in Section 3 show that the majority of endowments would have achieved better investment performance if they had simply allocated 60 percent to the stock market index and 40 percent to the bond market index (the passive 60/40 strategy). Furthermore, the results show that the 60/40 allocation strategy delivers the highest Sharpe ratio.

There is a rich body of literature on investments of university endowments. Arnold et al. (2024) use monthly data from 1997 to 2023 to construct mean variance-optimized portfolios of assets that university endowments have invested. They find that optimal allocations to these assets vary across different subperiods and that exposure to hedge funds is often present in optimized portfolios. They also show that international equity, real estate, and natural resources generally do not receive substantial allocation. Siegel (2021) states that the endowment model is merely active management. University endowments pursuing an active investment strategy must take into consideration the higher fees and other costs. The success of the endowment model is not guaranteed. Ennis (2021) discusses the failure of the endowment model. The performance drag is due to the higher annual fees and the failure to provide a diversification benefit. Ennis (2021) thus supports low-cost passive investments. Wallick et al. (2014) explain the operational advantages of large university endowments in implementing the endowment model. For smaller university endowments, it is challenging to replicate such an investment strategy. These university endowments are encouraged to invest in low-cost, diversified mutual funds.

Blazek and McMillan (2025) examine the challenges to the endowment model and discuss ways to adapt to the evolving market environment. They show that the performance of the endowment model has declined in recent years. During the 10-year period of 2015–2024, the average endowment did not outperform the 70/30 equity/bond passive strategy. During the latter two years of their sample period, i.e., 2023–2024, the 70/30 passive portfolio outperformed the average endowment by a cumulative 10.5 percent. They discuss the changes in the market environment and several steps to take advantage of the advancements in technology and analytics.

Binfare et al. (2023) examine how human capital impacts allocation to alternative assets and investment performance for university endowment funds. The three sources that contribute to higher investment returns are higher returns from alternative assets, selection of and access to high-performing managers, and investing in funds directly for lower fees. The proxies used for human capital in this study are expertise in alternatives, a chief investment officer, and the size of the investment staff. Fuss et al. (2023) examine the role of alumni ties in allocation to private equity by university endowments. Their empirical results show that university endowments are 70 percent more likely to invest in private equity funds managed by alumni. The strongest such ties are found for lower-ranked universities and for less experienced university endowments. Their results do not show any evidence of an association between alumni ties and performance.

Filosa (2024) discusses the role of university endowments. The study reveals that the majority of academic institutions increase endowment spending to fund growing operating expenses and that universities rely on endowment funding to balance the budget. The study shows that the average operating deficit before endowment spending is 15.3 percent. Endowment withdrawals provide a cushion and reduce the deficit to an average of 4.4 percent. Bulman (2022) studies how private universities use income from investment returns on their endowments. Growing endowments provide large and persistent increases in overall university spending and for instruction, student services, and administration as well. However, there is no evidence that spending increases expand enrollment or increase the fraction of entering freshmen eligible for financial aid. Universities and colleges with larger endowments do not increase access for low-income students. The study shows that wealthier institutions become more selective and thus the higher admissions yields do not result in a net increase in freshmen enrollment.

Aragon et al. (2021) examine whether socially responsible investing (SRI) in investment policies impacts the flow of donations to endowments. The empirical results show that the percentage of university endowments adopting SRI policies increases over time and that donations increase upon the adoption of SRI policies. Additional benefits of SRI policies include enhanced risk management practices, more student applications, and more funding for faculty research in sustainability-related projects. However, they show that SRI policies result in lower, albeit statistically insignificant, investment performance.

This paper is related to several areas of the literature on university endowments. First, we analyze the size group data obtained from NACUBO publications and provide empirical evidence that smaller university endowments underperform the conventional 60/40 stock/bond passive strategy, supporting prior studies challenging their pursuit of the endowment model. Furthermore, we show that the conventional 60/40 passive portfolio delivers the highest Sharpe ratio. Second, we show the annual rate of returns and spending rate that support the perpetual nature of university endowments. Finally, we utilize panel data regression to examine how university endowment asset allocation contributes to investment performance.

The remainder of the paper is organized as follows. In the next section, we describe the data coverage and data source. Section 3 examines asset allocation and investment returns for university endowments of different size groups. Section 4 discusses investment performance and spending. Section 5 investigates the link between asset allocation and investment returns. Finally, the last section provides the conclusions, discusses the limitations of this study, and outlines directions for future research.

2. Data Source and Data Structure

Our study focuses on asset allocation, investment performance, and spending rates for university endowment funds. We obtain the data from publications by the National Association of College and University Business Officers (NACUBO). The annual data sample period is from 2002 to 2022, and the annual data are based on the academic year and not on the calendar year. The annual data include the rate of return, percentage allocation to equities, percentage allocation to fixed income, percentage allocation to alternative assets, and spending rate. The data sample period (2002–2022) is determined by the data availability of NACUBO publications. The publications are available from the NACUBO website.

For 2002 to 2017, NACUBO publications list six size groups of university endowments: (1) over $1 billion, (2) $501 million to $1 billion, (3) $101 million to $500 million, (4) $51 million to $100 million, (5) $25 million to $50 million, and (6) under $25 million. Starting year 2018, NACUBO publications separate asset size group “$101 million to $500 million” into two size groups consisting of “$101 million to $250 million” and “$251 million to $500 million”. We combine these two groups into one endowment size of “$101 million to $500 million” to maintain the same size grouping for the whole sample period of 2002–2022.

3. Asset Allocation and Investment Performance

Colleges and universities regularly evaluate their asset allocation mixes to pursue better investment performance. University endowments utilize various investment theories to establish their portfolio strategies. The modern portfolio theory is a framework for portfolio construction to maximize the expected returns for a chosen level of risk. The theory emphasizes diversification across assets. The endowment model leverages long-term investment horizons to heavily allocate toward alternatives to capture the illiquidity premium and to exploit inefficiencies in private markets. Agency theory and governance studies focus on how investment committees, board expertise, and governance structures influence the decision regarding active versus passive management. The spending policy integration approach guides asset allocation to meet specific spending rates while preserving the real value of the endowment. Furthermore, the background adjustment theory suggests conservative allocations for universities with high budget volatility. Several of these theories are related to this study in terms of risk-adjusted performance, the observed increasing allocations toward alternatives, and the comparative analysis of performance and spending.

Academic research publications and news media have reported that university endowments have adopted the endowment model, which reduces their asset allocation to conventional public market equities and bonds and increases allocation to alternative assets such as hedge funds, private equity, commodities, and real estate. Proponents claim that the endowment model benefits endowments in terms of portfolio risk reduction and better investment performance. Alternative assets and conventional public securities are not perfectly correlated, leading to a lower portfolio risk. These alternative assets are not as liquid as public securities and thus offer premiums to compensate for the illiquidity. University endowments invest for the long term and hence can reap both diversification and enhanced performance benefits. The Managed Funds Association (2022) states that a one-percent increase in allocation to hedge funds increases returns by 0.06 percent. In contrast, some have raised concerns that alternatives are not for every endowment fund. Wallick et al. (2014) identify three factors that contribute to the success of mega-university endowments. These factors are investment expertise, fee pricing power, and access. Smaller university endowments do not have access to top funds, and they do not have the levels of investment staff and expertise. Smaller university endowments also do not have the size to negotiate lower fees. The authors show that the average small and medium endowments would have achieved a better investment outcome if they had adopted the strategy of low-cost diversified mutual funds invested in conventional stocks and bonds. Ennis (2021) discusses the causes of the performance drag. Blazek and McMillan (2025) demonstrate that, during 2023–2024, the average endowment underperformed the 70/30 portfolio by a cumulative 10.5 percent.

Table 1. Asset allocation to alternative assets (percentage).

Size Group	Average 2002–2022	2002	2022
Over $1 billion	52.02	32.0	62.6
$501 million to $1 billion	38.50	22.0	42.9
$101 million to $500 million	28.7	14.3	35.3
$51 million to $100 million	18.89	7.2	20.4
$25 million to $50 million	14.19	6.3	16.1
Under $25 million	8.47	5.8	7.8

Note: The first column lists the six size groups, the second column lists the averages of the percentages allocated to alternatives for each size group during 2002 to 2022, the third column lists the percentage allocation to alternatives for each size group in 2002, and the last column lists the percentage allocation to alternatives in year 2022. Data Source: National Association of College and University Business Officers (NACUBO).

shows that, during the sample period of 2002 to 2022, university endowments of all size groups increased allocation to alternative assets and that the percentage of allocation increased with size. The size group “over $1 billion” increased allocation from 32.0 percent in 2002 to 62.6 percent by 2022. The next size group, $501 million to $1 billion, allocated 22.0 percent to alternative assets in 2002 and increased this allocation to 42.9 percent in 2022. This pattern is similar for all other size groups. The smallest size group, under $25 million, increased its allocation slightly by 2 percent. For the sample period, the average for the largest endowments (over $1 billion) is 52.02 percent. The average is lower for smaller endowments. The smallest (under $25 million) averages at 8.47 percent.

Table 2. Investment performance summary statistics 2002–2022 (percentage).

	S&P 500 Index	Bloomberg U.S. Bond Aggregate Index	Average	Over $1 Billion	$501 Million to $1 Billion	$101 Million to $500 Million	$51 Million to $100 Million	$25 Million to $50 Million	Under $25 Million	60/40
Mean	8.82	4.22	6.50	8.35	7.66	6.50	5.95	5.89	5.89	6.98
Standard deviation	16.13	4.90	10.94	11.95	11.81	11.23	10.96	10.59	10.69	9.57
Minimum	−26.2	−10.3	−18.7	−20.5	−19.8	−19.7	−18.6	−18.5	−16.8	−13.3
Maximum	40.8	10.4	30.6	37.3	33.9	30.2	26.6	27.3	23.9	24.4
Sharpe ratio	0.47	0.62	0.49	0.60	0.55	0.47	0.43	0.44	0.44	0.61

Notes: (1) The fourth column, “Average”, indicates the average of all university endowments of different size groups. (2) The last column, “60/40”, indicates the rate of return for allocation of 60 percent in the S&P 500 Index and 40 percent in the Bloomberg U.S. Bond Aggregate Index. Data sources: NACUBO and Bloomberg.

provides summary statistics of returns for the S&P 500 Index, Bloomberg U.S. Bond Aggregate Index, investment performance for each of the six size groups, and the rate of returns for 60/40 allocation (60 percent allocation in S&P 500 index and 40 percent in Bloomberg U.S. Aggregate Bond index). The average annual rate of returns for the S&P 500 index is 8.82 percent, and the average for the Bloomberg U.S. Aggregate Bond Index is 4.22 percent. The average annual rate of returns for all university endowments is 6.50 percent. As Table 2 shows, the average rate of returns increases with the size group of the endowment. The strategy of 60/40 allocation in equity and fixed-income indexes generates an average of 6.98 percent per year during 2002–2022. The standard deviation of returns for 60/40 allocation is lower than that of any size group of endowments. Table 2 also lists the minimum and maximum rates of returns for each size group. Finally, Table 2 shows the risk-adjusted performance, i.e., the Sharpe ratio, for each. The Sharpe ratio for the average of all endowments is 0.49 and that for the 60/40 passive allocation strategy is 0.61. The Sharpe ratios for the largest two size groups are 0.60 and 0.55, respectively. The Sharpe ratios for other, smaller size groups are all less than 0.50.

There are several interesting observations. The average investment rate of returns for all university endowments underperforms the 60/40 passive index investing strategy. The average rate of returns for all university endowments is 6.50 percent, and that of the 60/40 passive strategy is 6.98 percent. Furthermore, the standard deviation of returns for all university endowments is higher than in the 60/40 strategy. The largest two size groups, “over $1 billion” and “$501 million to $1 billion”, have higher rates of returns than in the 60/40 passive strategy. However, 60/40 allocation has a lower standard deviation. All other groups have lower rates of returns and higher standard deviations.

Observations from Table 1 and Table 2 indicate that larger endowments allocate a higher percentage to alternative assets and that larger endowments also have a higher rate of investment returns. Thus, some observers have promoted the endowment model to increase allocation to alternatives to generate better performance. However, the data from Table 1 and Table 2 also show that the results do not support such a claim. The largest two size groups have higher rates of returns (8.35 percent and 7.66 percent) than in the 60/40 strategy (6.98 percent), but the standard deviations of returns for these two groups (11.95 percent and 11.81 percent) are higher than the standard deviation of the 60/40 strategy (9.57 percent) as well. The other four smaller size groups have increased allocations to alternatives during the sample period but underperform compared to the 60/40 passive strategy with lower average rates of returns and higher standard deviations of returns. Furthermore, the Sharpe ratio for the 60/40 stock/bond allocation is higher than that of any size group.

Smaller endowments do not replicate the success of their large counterparts, despite increasing their allocation to alternative assets and reducing exposure to public securities. This is because smaller university endowments often lack the resources to hire specialized investment staff. Smaller endowments do not have access to top-tier managers or the size to negotiate lower fees. It is thus helpful to re-evaluate the market environment and revise the investment strategies. The empirical data suggest that active strategies (and increasing exposure to alternative assets) are not always better than the simple 60 percent equity and 40 percent fixed-income passive, low-cost strategy.

4. Investment Performance and Spending

University endowments have become an important source of funding for designated academic activities and related expenses. Universities and colleges utilize capital campaigns to raise money and investment strategies to increase their investment performance. In pursuit of higher returns, they have shifted endowment investments from fixed income to equities and then toward alternative assets.

Universities and colleges have spending policies that govern how they spend the withdrawn money from the endowment. Most spending policies use a formula designed to maintain stable withdrawals and to also grow the endowment over time. The formula could be a percentage of the average endowment values from the previous several quarters or several years. Alternatively, some universities use a smoothing rule that incorporates a portion of the previous-year spending plus a portion of the long-term target spending rate. Yale University is one such example. Yale’s withdrawal is set at 80 percent of the previous year’s spending plus 20 percent of the long-term spending rate of 5.25 percent.

Table 3. Annual spending rates (percentage).

	Over $1 Billion	$501 Million to $1 Billion	$101 Million to $500 Million	$51 Million to $100 Million	$25 Million to $50 Million	Under $25 Million
2002	4.8	5.0	5.1	5.3	4.9	5.2
2003	5.3	5.2	5.2	5.3	4.9	4.8
2004	5.2	5.0	5.0	4.9	4.7	4.5
2005	4.8	4.7	4.7	4.8	4.7	4.7
2006	4.5	4.5	4.6	4.7	4.7	4.7
2007	4.4	4.4	4.6	4.8	4.8	4.6
2008	4.2	4.5	4.2	4.6	4.3	4.1
2009	4.6	4.9	4.4	4.7	4.3	3.9
2010	5.6	5.7	4.9	4.6	4.1	3.5
2011	5.2	5.2	5.0	4.5	4.0	3.7
2012	4.7	4.7	4.3	4.3	3.8	3.7
2013	4.8	4.6	4.4	4.4	4.3	4.1
2014	4.6	4.3	4.3	4.4	4.2	4.6
2015	4.3	4.1	4.1	4.4	4.0	4.5
2016	4.4	4.3	4.3	4.4	4.1	3.8
2017	4.8	4.6	4.5	4.5	4.2	4.0
2018	4.6	4.2	4.4	4.6	4.1	4.1
2019	4.6	4.4	4.5	4.6	4.5	4.1
2020	4.5	4.4	4.6	5.0	4.6	4.1
2021	4.8	4.4	4.7	4.8	5.6	4.3
2022	3.9	3.6	4.1	4.2	5.3	3.8
Average	4.70	4.60	4.57	4.66	4.48	4.23
Standard deviation	0.39	0.45	0.32	0.29	0.46	0.44

Note: All annual spending rates for the period 2002–2016 are as reported by NACUBO. The annual spending rate for years 2017–2022 for endowment size “$101 million to $500 million” is the average of those for “$101 million to $250 million” and “$251 million to $500 million”. Data source: NACUBO.

lists the annual spending rates from 2002 to 2022 for endowments of the six size groups. The spending rate is defined as the percentage of the amount of withdrawal from the endowment relative to the value of the endowment at the beginning of the period. During the sample period of 2002–2022, the spending rate declines slightly for all endowments. On average, larger university endowments have a higher spending rate. For example, the average spending rate for the largest endowments (over $1 billion) is 4.70 percent—higher than that for smaller endowments. The spending rates in other size groups of endowments are 4.60 percent, 4.57 percent, 4.66 percent, 4.48 percent, and 4.23 percent, respectively. The observations in Table 3 also indicate that there is no pattern in the variation (standard deviation) in spending rates for different size groups over the sample period.

Investment performance fluctuated, as evidenced by the rates of returns listed in

Table 4. Average annual spending rates and investment returns (percentage).

Year	Spending Rate	Investment Returns
2002	5.1	−6.2
2003	5.1	3.2
2004	4.9	15.3
2005	4.7	9.3
2006	4.7	10.8
2007	4.6	17.2
2008	4.3	−3.0
2009	4.4	−18.7
2010	4.5	11.9
2011	4.6	19.2
2012	4.2	−0.3
2013	4.4	11.7
2014	4.4	15.5
2015	4.2	2.4
2016	4.3	−1.9
2017	4.4	12.2
2018	4.4	8.2
2019	4.5	5.3
2020	4.6	1.8
2021	4.8	30.6
2022	4.0	−8.0
Average	4.55	6.50
Standard deviation	0.32	10.94

Data source: NACUBO.

. For example, university endowments post a return of −6.2 percent in 2002 and −18.7 percent in 2009. The average rate of returns for all university endowments exceeds 15% for 2004, 2011, 2014, and 2021. The average rate of investment returns for years 2002–2022 is 6.50 percent. The average spending rate for the sample period is 4.55 percent. During the period, the spending rate decreases from 5.1 percent in 2002 to 4.0 percent in 2022.

5. Contribution of Asset Type to Investment Performance

We next turn to the examination of how each asset class (equities, fixed income, and alternatives) contributes to the investment performance of university endowments during the sample period. There are six size groups, as listed previously. For each of the size groups of university endowments, the data consist of the annual rate of return, percentage allocation to equities, percentage allocation to fixed income, and percentage allocation to alternative assets for the sample period of 2002–2022. The data set is panel data, as it consists of time series (2002–2022) and cross-sections (six size groups).

Public equity and fixed-income securities generally have closing prices every trading day. These prices are used to determine performance. Alternative assets—at least some of them—are not traded every day. Furthermore, it usually takes time to source targets and to invest the allocated capital. After the investment, additional time is required to enhance the value and then to exit. As such, investments in alternative assets might not generate immediate returns. Furthermore, these alternatives are not publicly traded. They are valued only periodically, and their valuations are generally reported with a one-quarter or two-quarter delay. Thus, we assume that there is a one-year lag in the contribution of alternatives to investment performance.

A panel data regression is performed to analyze the impact of asset allocation on investment returns. The panel data regression moves beyond descriptive statistics to explicitly test the statistical significance of asset allocation on investment performance. While the descriptive statistics provide a snapshot, the panel data regression analysis controls for covariates such as the year and size of the endowment. Equation (1) lists the panel data regression model:

Y_it = a_i + bEQ_it + cFI_it + dALT_i(t−1) + u_it

(1)

where

Y is the annual rate of return of university endowments;

a is a constant term;

EQ is the percentage allocation in equities;

FI is the percentage allocation in fixed income;

ALT is the percentage in alternative assets;

u is the error term.

In Equation (1), i is the index for the size group of the university endowment and t is the time index. The coefficients (b, c, and d) measure the impacts of asset allocation in equities, fixed-income, and alternative assets on investment returns, respectively.

Table 5. Panel data regression estimates.

	Coefficient Estimate	t-Value
Allocation in equities	2.343	8.803 ***
Allocation in fixed income	0.375	1.036
Allocation in alternatives (1-year lag)	1.630	6.962 ***
R-squared, within	0.4124

Note: *** denotes significance level at 1 percent. Data source: NACUBO.

lists the results of the panel data regression. The estimated coefficient for allocation in equities is 2.343 with a t-value of 8.803, significant at 1 percent. The estimated coefficient for allocation to fixed income is 0.374 with a t-value of 1.036, not significant at 10 percent. The estimated coefficient for allocation in alternatives with a one-year lag is 1.630 with a t-value of 6.962, significant at 1 percent. Note that the estimated coefficient indicates the incremental marginal impact for a small percentage change in allocation.

The regression results listed in Table 5 are based on a fixed-effects model. We perform the Hausman test to determine if fixed effects or random effects should be used in the panel data regression. The chi-squared statistic is 69.589 with a p-value of 0.033. Thus, the fixed-effects model is used in our analysis.

The reported results are based on a size group fixed-effects model. Size group fixed effects control for unobserved time-invariant factors across the six size groups. A one-way rather than two-way fixed-effects approach is chosen because of sample size constraints. Adding time fixed effects introduces an - additional twenty time dummies, resulting in insignificant estimates.

Asset allocation in equities, fixed income, and alternatives might be endogenous to performance. To address potential endogeneity, we utilize a dynamic panel model. The results are similar to those reported in Table 5. The estimated coefficient for allocation to fixed income is 0.205, insignificant at the 10 percent level. The estimated coefficient for equities is 2.406, significant at 1 percent. The estimated coefficient for alternatives with a one-year lag is 1.591, significant at 1 percent. For the dynamic panel regression, the Wald test chi-squared value is 2569.04. This indicates that the model is significant and asset allocation decisions have a strong influence on investment performance.

Table 6. Dynamic panel regression estimates.

	Coefficient Estimate	z-Value
Allocation in equities	2.406	21.67 ***
Allocation in fixed income	0.205	0.749
Allocation in alternatives (1-year lag)	1.591	10.64 ***
Wald test	Chi-square = 2569.04

Note: *** denotes significance level at 1 percent. Data source: NACUBO.

lists the results of the dynamic panel regression.

6. Conclusions

Universities and colleges withdraw billions of dollars from their endowments to fund student financial aid, academic programs and research, endowed faculty positions, campus operations and maintenance, and other related expenses. A primary objective of university endowment management is to generate adequate returns to support these activities. To pursue better investment performance, university endowments have shifted asset allocation from fixed income to equities and to alternatives. The comparative analysis in this study shows that large university endowments have achieved a higher rate of return for increased allocation to alternative assets. Smaller university endowment funds, employing a similar strategy but with lower alternative exposure, achieve limited success.

The analyses in Section 3 show that the 60/40 allocation strategy has an average rate of return that is better than that of any size group at $500 million or smaller. Furthermore, the standard deviation of returns of the 60/40 strategy is lower than that of any endowment size group. In the risk-adjusted performance analysis, the Sharpe ratio of 60/40 allocation is higher than that of any size group.

To examine how each type of asset contributes to investment performance, a panel data regression has been performed. The results indicate that the coefficients are all positive, and the coefficient estimate for equities is significant at the 1 percent level; for fixed income, it is not significant at 10 percent; and for alternatives with a one-year lag, it is significant at the 1 percent level. To address potential endogeneity, we utilize a dynamic panel model. The results are similar to those from the panel data regression.

This study has several limitations. Due to a lack of endowment-level data, the analysis relies on size group data. The unavailability of data prevents an analysis of management quality, fees, access, or macroeconomic variables. Additionally, this study does not explore the specific objectives and constraints of each university endowment. These limitations will be addressed as better data become available.

After 2023, NACUBO publications include detailed allocations to each type of alternative asset. An area for future research is to examine how each type of alternative asset contributes to university endowment investment performance. With data breaking down allocations into specific types of alternatives, future research could investigate whether specific alternative asset types contribute disproportionately to performance. Another interesting topic is to investigate whether and how gift giving from stakeholders is correlated with the investment performance of the endowment. One approach is to examine how investment performance impacts stakeholder gifts for different endowment sizes, different institution types (for example, private versus public), and asset allocation. Furthermore, many universities offer programs with student-managed investment funds. These programs provide real-world asset management experience for students. It would also be interesting to compare the differences between student-managed funds and professionally managed university endowments. There are several distinct differences, including size, invested assets, turnovers, risk tolerance, and objectives. This might shed light on how these contrasting characteristics impact their performance.