Capital on the Ramp: Hangar Real Estate for Endowment Portfolios

Abstract

1. Introduction

University endowments face a persistent challenge: generating returns sufficient to support institutional spending, typically 4.5 to 5.5 percent of trailing average assets, while preserving intergenerational equity. The period since the global financial crisis has seen allocators respond through systematic diversification into alternatives and real assets. In fiscal year 2025 the average participating institution held roughly 45 percent of its portfolio in alternative strategies and 10 percent in real assets, with real asset weightings rising from about 5 percent among endowments under $50 million to about 11 percent among those over $5 billion (NACUBO and Commonfund 2026). Within the real assets category, most allocations remain concentrated in conventional property types that have experienced yield compression and structural demand disruption from remote work and changing retail patterns.

This paper introduces aviation hangar real estate as a candidate diversifying allocation. Hangars are purpose-built structures for the storage, maintenance, and operation of aircraft. They represent a substantial real estate category that has attracted institutional capital over the past decade while receiving almost no attention in the portfolio construction literature. The entry of publicly traded companies, private equity platforms, and institutional family offices has moved the sector from an opaque collection of individually owned properties toward an investable category with improving data availability and professional management.

The investment case rests on five arguments developed below. Supply is structurally constrained by federal funding priorities, airport authority capacity limits, and the near absence of speculative development. Demand is driven by corporate profitability, high-net-worth household formation, and executive time valuation rather than by retail foot traffic, remote work adoption, or e-commerce logistics. Income is contractual, frequently triple-net, and frequently escalated by index. Management requirements are simpler than for multi-tenant office, retail, or residential property. And an institutional ecosystem now exists, including a public company, private equity platforms, and the fixed-base operator consolidation that preceded them, that supplies the professional management and documentation fiduciaries require.

The contribution is threefold, and the third part is as important as the first two. First, the paper applies portfolio theory to aviation real estate and constructs the return, risk, and correlation estimates an allocation model requires. Second, it analyzes the risk characteristics of leasehold hangar ownership within a framework institutional risk managers can evaluate. Third, it is explicit about what those estimates are and are not. They are informed assumptions. They are not measurements, and no measurement is currently possible because the data does not exist. A paper that presents assumed inputs and reports the resulting frontier as a finding has not analyzed anything; it has restated its own assumptions in a more confident font. Section 5 is therefore constructed as a sensitivity demonstration, and Section 8 states plainly what would have to be measured for it to become something more.

Section 2 reviews the literature on endowment portfolio construction, the critical work on that model's recent performance, the estimation-error literature on mean-variance optimization, and the thin literature on aviation real estate. Section 3 establishes the institutional characteristics of the hangar market. Section 4 develops the parameter estimates. Section 5 presents the sensitivity analysis. Section 6 examines the specific risks. Section 7 discusses implementation. Section 8 sets out limitations. Section 9 concludes.

2. Literature Review and Theoretical Foundation

2.1 The Endowment Model and the Case Against It

The modern endowment paradigm was shaped by David Swensen at Yale, whose allocation to illiquid alternatives generated strong long-term returns that other institutions sought to replicate (Swensen 2000, 2009). The central claim, that a perpetual horizon and a tolerance for illiquidity are structural advantages that can be monetized through exposure to illiquidity premia, became close to axiomatic in institutional portfolio construction. Lerner, Schoar, and Wang (2008) documented that the largest endowments, which allocate most heavily to alternatives, outperformed smaller institutions concentrated in public equities and fixed income.

That consensus is no longer secure, and any paper proposing a further alternative allocation has an obligation to say so. Barber and Wang (2013) found that the outperformance of elite endowments was largely explained by exposure to public equity factors rather than by manager selection within alternatives. Ang, Ayala, and Goetzmann (2018), examining the stated investment beliefs of endowments against their realized results, found limited evidence that active allocation decisions added value. Ennis (2021) argued from performance data that the endowment model has failed on its own terms, producing returns below passive benchmarks of comparable risk once fees and factor exposures are accounted for. Dimmock, Wang, and Yang (2024) show formally that illiquidity constraints cause endowment holdings to deviate substantially from target allocations precisely during market stress, which undercuts the premise that a perpetual horizon neutralizes illiquidity risk.

The most recent data does not settle the argument in the model's favor. In the fiscal 2025 study, the smallest endowment cohort reported stronger three-year returns than the largest, a pattern inconsistent with the proposition that heavier alternatives exposure is reliably rewarded (NACUBO and Commonfund 2026). A hangar allocation is an alternatives allocation. It inherits this debate rather than escaping it, and the case made below should be read as conditional on the allocator's own view of it.

2.2 Estimation Error in Mean-Variance Optimization

A second body of work bears directly on the method used in Section 5. Michaud (1989) described mean-variance optimizers as estimation-error maximizers: because the optimizer allocates toward assets with favorable estimated inputs, any estimation error is systematically amplified rather than averaged away. Chopra and Ziemba (1993) quantified the asymmetry, finding that errors in estimated means are roughly an order of magnitude more costly to an investor than errors in estimated variances, and roughly twice as costly again relative to errors in covariances. DeMiguel, Garlappi, and Uppal (2009) found that across a range of datasets, sophisticated optimization models failed to consistently outperform naive equal weighting out of sample, because the estimation error in the inputs overwhelmed the theoretical gain from optimizing.

These findings are not a reason to avoid portfolio analysis. They are a reason to be explicit about which inputs carry the most estimation risk and to report results in a form that does not imply precision the inputs cannot support. For a new asset class with no return history, the expected return is the input with the largest estimation error and the largest consequence. That governs how Section 5 is constructed and how its results are reported.

2.3 Niche Real Estate and the Diversification Frontier

The literature on niche real estate allocation has expanded since 2015, driven by the institutionalization of property types previously considered too specialized. Self-storage, student housing, senior living, data centers, and life sciences laboratories have each moved from fragmented operational niches to institutional categories with dedicated vehicles and performance benchmarks. The common pattern is structural demand growth, supply constraint, and differentiated return drivers. Pagliari, Scherer, and Monopoli (2005) demonstrated that return and risk profiles vary substantially across property types, with direct implications for portfolio construction.

Aviation hangars share the niche characteristics but add a further dimension: the leasehold ownership structure. Unlike most commercial real estate, hangar ownership typically involves a ground lease from an airport authority. The literature on ground lease valuation is developed for urban commercial and residential property. Giglio, Maggiori, and Stroebel (2015) used very long leaseholds to estimate discount rates at long horizons. Caesar, Donner, and Kopsch (2019) measured the leasehold discount directly in the Swedish apartment market and found that price rises by approximately 0.22 percent for each additional year of remaining lease term, a finding that gives the term-structure effect an empirical magnitude rather than a theoretical sign. That literature has not been extended to aviation, where reversion without compensation and the bargaining position of the airport authority at expiration amplify term risk beyond what the residential estimates imply (Carter 2026a, 2026c).

One point of contrast matters. Hangars have not yet institutionalized in the way those other niches did. Data centers and self-storage entered institutional indices, acquired performance benchmarks, and developed transaction depth before allocation frameworks were written for them. For hangars that sequence has not occurred. The analysis below is therefore prospective in a way that the niche real estate literature generally is not.

2.4 Aviation Real Estate: An Academic Gap

The academic literature on aviation real estate valuation is thin. Lindsey (2008a, 2008b) published the only comprehensive peer-reviewed treatment of hangar valuation methodology, applying cost, sales comparison, and income capitalization to aircraft storage facilities. Since that work, the market has seen the entry of institutional capital, the emergence of a public vehicle, and a shift in demand toward larger-footprint hangars driven by fleet composition. No academic research has updated the valuation framework or examined portfolio implications. There is no hangar return index, no transaction-based performance series, and no property type classification for hangars in the principal institutional real estate indices, where the nearest available category is air cargo within the industrial sector. This paper begins to address the gap by treating hangars as portfolio assets rather than as isolated appraisal problems, while acknowledging that the data required to do so properly does not yet exist.

3. Institutional Characteristics of the Hangar Market

3.1 Market Structure and Supply

The United States aviation system includes roughly 19,800 landing facilities of all types, of which 3,285 are designated in the National Plan of Integrated Airport Systems and therefore eligible for federal funding. Hangar-related activity is a principal revenue source for general aviation airports, which gives airport sponsors a direct financial interest in the hangar inventory and in the lease terms governing it.

Structural undersupply is the feature most relevant to investors. Survey evidence from aircraft owner associations consistently reports multi-year waits for hangar space at desirable fields, with a minority of the existing inventory rated in good condition and a substantial share requiring major repair. State aviation system plans in several jurisdictions have identified hangar shortfalls measured in tens of percent of current inventory. Federal grant priority does not close the gap: hangar construction ranks near the bottom of eligible project types under the Airport Improvement Program priority system, so federal funding rarely finances the speculative capacity that would relieve the shortage. For investors, that combination produces barriers to entry, pricing power at existing facilities, and demand visibility, while also raising a question the optimism in this literature tends to skip: an undersupply that has persisted for decades under stable policy is evidence of a durable constraint on supply, and equally evidence of a market that has not cleared, which is not the same thing as a market that will reward new capital.

3.2 The Institutional Capital Wave

Aviation real estate has moved over the past decade from a fragmented collection of individually owned properties toward an institutional category. Sky Harbour Group Corporation, listed on the New York Stock Exchange, is the sector's first publicly traded pure-play hangar company, operating a portfolio of campuses at metropolitan airports funded through public equity, tax-exempt private activity bonds, and a bank construction facility. The company is organized as an umbrella partnership C corporation rather than as a real estate investment trust, though management has discussed conversion as a longer-term option. Its access to public equity and to the tax-exempt bond market demonstrates that the sector has achieved a degree of institutional acceptance.

Private capital has entered alongside it. The acquisitions of the two largest fixed-base operator networks by major private equity firms represent multi-billion dollar commitments to airport infrastructure, and smaller platforms have assembled hangar portfolios directly. These entrants bring professional management, standardized documentation, and portfolio-scale operations. They also bring competition for the same constrained supply, which is the mechanism by which an identified inefficiency stops being one.

3.3 Demand Drivers and Correlation Structure

The economic drivers of hangar demand differ from those governing traditional commercial real estate. Business aviation demand tracks corporate profitability, high-net-worth household formation, and the value placed on executive time rather than retail foot traffic or logistics volumes. The composition of the fleet has shifted materially toward larger aircraft over the past fifteen years, which drives demand for newer and larger facilities precisely in the supply-constrained metropolitan markets institutional investors target.

This differentiated demand profile has implications for diversification. During the pandemic, when office and retail experienced severe dislocation, business aviation saw a demand surge as travelers sought private alternatives, and hangar occupancy held while urban office vacancy climbed. That episode is suggestive. It is also a single observation from an event with no precedent, and it should not be mistaken for a measured correlation. The correlation estimates used in Section 4 rest on this reasoning about demand drivers, not on a return series, because no return series exists.

3.4 Income Characteristics

Hangar leases exhibit income characteristics that align with endowment spending requirements. Institutional-grade hangar leases are commonly structured as triple-net arrangements, with tenants responsible for taxes, insurance, and maintenance. Tenant lease terms range from five to thirty years; ground leases from airport authorities commonly run twenty to fifty years and occasionally longer. Contractual escalation tied to a price index or to fixed annual increases of 2 to 4 percent provides built-in inflation protection.

The tenant credit profile is distinctive. Business aviation tenants include corporate flight departments, high-net-worth individuals, charter operators, and aircraft management companies, a cohort with above-average creditworthiness and low default rates. Waiting lists at established campuses allow later development phases to pre-lease before construction completes, a demand signal with few parallels in conventional commercial real estate.

3.5 What the Public Company Data Does and Does Not Show

Sky Harbour's filings are the only audited, continuously disclosed financial data in the sector, which makes them valuable and makes their limitations worth stating precisely.

Revenue grew from $7.6 million in fiscal 2023 to $14.8 million in fiscal 2024 to $27.5 million in fiscal 2025 (Sky Harbour Group Corporation 2026). The fiscal 2024 increase of approximately 95 percent is frequently quoted as evidence of organic growth in hangar demand. It is not that. The figure is inflated by an acquisition completed in December 2024 and by a change in the presentation of fuel revenue from net to gross. The Obligated Group revenue growth, which tracks the bond-financed hangar portfolio itself, was roughly 51 percent and then roughly 49 percent across the same periods, and is the more meaningful indicator of the underlying business.

The operating picture is also less settled than a revenue series alone suggests. The company reported positive net income of $18.8 million for fiscal 2025 against an operating loss of $28.0 million, with the difference driven by the fair value remeasurement of warrants rather than by operations. As of the second quarter of 2026 it reported occupancy of 83.7 percent, adjusted EBITDA of negative $0.9 million, and an accumulated deficit of $52.6 million. A company still building toward operating profitability is not evidence against the asset class, but it is not evidence of stabilized asset-class returns either, and it should not be used as though it were.

4. Parameter Estimation

4.1 Return Estimation

Estimating expected returns for hangar investment requires triangulating across sources, none of which is a return series. Three are available: implied capitalization rates from observable transactions, disclosed financial data from the one public company, and target return ranges reported by private investors active in the sector.

The first of these is weaker than it appears. There is no published dataset of hangar transaction capitalization rates. The major institutional real estate indices do not classify hangars as a property type. What exists is a set of individually reported transactions, disclosed inconsistently, with lease structures that are not comparable across deals. Practitioner observation places stabilized hangar capitalization rates above core real estate by a risk premium of roughly 200 to 400 basis points, and that observation is the basis for the estimates used here. It is not a measurement, and the reader should treat it as the judgment of a practitioner rather than as a statistic.

Combining an initial yield in that range with contractual escalation of 2 to 4 percent and the potential for income growth driven by supply constraint produces an unlevered total return estimate in the range of 10 to 14 percent. For context, the principal institutional real estate indices reported a trailing one-year total return of approximately 5 percent for the property index in mid-2026, and the open-end diversified core fund index reported a three-year annualized return that remained negative on both a gross and a net basis (NCREIF 2026). A hangar return assumption of 10 to 14 percent is therefore an assumption that this sector will deliver a wide margin over core real estate through a period in which core real estate has not delivered. That is the assumption doing the most work in Section 5, and it is the one least supported by evidence.

4.2 Risk Estimation

Volatility estimation is complicated by the private nature of most transactions and by the absence of an appraisal-based index comparable to those available for traditional property types. The estimate used here is a standard deviation of 10 to 15 percent, reflecting offsetting factors. Contractual escalation smooths income volatility, supply constraint limits vacancy risk, and the tenant base reduces credit risk. Against that, leasehold ownership creates finite-horizon risk, airport authority actions can change operating conditions, and a thin transaction market creates valuation uncertainty.

The appraisal smoothing problem documented in the private real estate literature (Geltner 1993) is likely more acute here than in conventional property, given how few comparable sales exist. Reported volatility will understate true economic risk, particularly during stress when transaction activity declines and appraisals lag. An allocator using a 10 to 15 percent standard deviation should understand that the figure describes how a smoothed appraisal series would behave, not how the underlying asset value behaves.

4.3 Correlation Estimates

Correlation estimates rest on the demand-driver reasoning in Section 3.3 rather than on observed co-movement. Dependence on corporate profitability and high-net-worth wealth formation suggests moderate positive correlation with domestic equities, in the region of 0.25 to 0.40. Real asset characteristics and contractual income suggest low correlation with core fixed income, in the region of 0.05 to 0.15. Differentiated demand drivers suggest low to moderate correlation with core real estate, in the region of 0.20 to 0.35.

An earlier version of this analysis supported the equity correlation estimate by reference to the trading behavior of the sector's public company. That support has been withdrawn, and the reason is worth stating because the same inference is common in emerging sectors. The company completed its public listing in January 2022, giving roughly eighteen quarters of history. Average daily volume is on the order of one hundred thousand shares, volume across the period has varied by a factor of more than forty, the share price has ranged from roughly $2.50 to roughly $43, and the free float is a fraction of market capitalization. A single thinly traded small-capitalization security over eighteen quarters cannot identify the correlation structure of an asset class, and a correlation estimated from it would carry a standard error wide enough to encompass most of the plausible range. The correlations used here are assumptions derived from economic reasoning, and Appendix A labels them as such.

4.4 Comparative Asset Class Characteristics

Table 1 summarizes the assumed investment characteristics used as parameter inputs in Section 5.

Table 1. Assumed asset class characteristics for endowment portfolio analysis.

All figures are forward-looking assumptions, not historical measurements. Hangar estimates derive from practitioner observation of transactions, public company disclosure, and reported private return targets, as described in Sections 4.1 through 4.3. Traditional asset class assumptions are conventional long-horizon planning figures and should be replaced with the allocator's own capital market assumptions before use. The hangar row is the least well supported row in the table.

5. Portfolio Sensitivity Analysis

This section examines what a hangar allocation does to a representative endowment portfolio under the assumptions above. It is not an optimization, and its output is not a result. Every input is assumed, and Michaud (1989) and Chopra and Ziemba (1993) establish that an optimizer applied to assumed means will amplify the error in them. What follows is therefore a demonstration of directional sensitivity: under which assumptions does the allocation help, by roughly how much, and at what point does it stop helping.

5.1 The Representative Portfolio

The base portfolio is 35 percent domestic equity, 15 percent international equity, 15 percent fixed income, 15 percent private equity, 10 percent hedge funds, and 10 percent core real estate. Under the Table 1 assumptions and the Appendix A correlations, this portfolio produces an expected return of roughly 8 percent with a standard deviation of roughly 10 percent, which is in the region reported by mid-size endowments.

The alternative portfolio reallocates 3 percentage points from core real estate and 2 from domestic equity into a 5 percent hangar allocation. The rationale is that hangars are assumed to offer a higher return than core real estate with differentiated risk factors, and that a modest reduction in domestic equity lowers sensitivity to public market drawdowns.

5.2 Directional Result

Under the central assumptions, the alternative portfolio produces an expected return roughly 30 basis points higher and a standard deviation roughly 30 basis points lower, improving the Sharpe ratio by something on the order of 0.05 at an assumed 2.5 percent risk-free rate. Current income yield rises by roughly 30 basis points.

Table 2. Directional effect of a 5 percent hangar allocation.

Computed from the assumptions in Table 1 and Appendix A. Figures are rounded deliberately. They are arithmetic consequences of assumed inputs and carry no more precision than those inputs, which is to say very little. Base portfolio: 35 percent domestic equity, 15 percent international equity, 15 percent fixed income, 15 percent private equity, 10 percent hedge funds, 10 percent core real estate. Alternative portfolio reallocates 3 points from core real estate and 2 points from domestic equity.

5.3 Where the Result Breaks

The more useful question is what has to be true for the allocation to help, and what makes it stop helping.

The result is driven primarily by the assumed hangar return. At an assumed return of 9 percent with a 15 percent standard deviation and a 0.40 equity correlation, the Sharpe improvement is marginal, on the order of a few basis points, and is within the range that any reasonable revision to the other assumptions would erase. At an assumed 13 percent return with a 10 percent standard deviation and a 0.25 equity correlation, the improvement is an order of magnitude larger. The entire case therefore turns on an expected return that no one has measured, which is exactly the input Chopra and Ziemba (1993) identify as most costly to get wrong.

The result is also sensitive to the correlation assumption, though less so. Because the allocation is small, a correlation error of 0.15 moves portfolio standard deviation by a few basis points rather than by a percentage point. This is a general property of small allocations and is the strongest argument for sizing an initial position conservatively: at 2 to 5 percent, the cost of being wrong about the correlation is bounded, while at 10 percent it is not.

Varying the allocation from 1 to 10 percent, the marginal improvement in portfolio efficiency is greatest between 2 and 4 percent and diminishes above 5 percent as exposure to hangar-specific risks, including lease term, airport authority behavior, and illiquidity, begins to offset the diversification benefit. That pattern supports a target allocation in the 2 to 5 percent range, consistent with initial allocations to other niche categories, and it holds under every set of assumptions tested because it is driven by the concentration of idiosyncratic risk rather than by the return assumption.

5.4 Income Distribution Implications

Beyond the risk and return arithmetic, a hangar allocation addresses a practical problem: the gap between portfolio income yield and spending policy. The base portfolio generates roughly 2.8 percent in current income against an effective spending rate of approximately 4.9 percent across participating institutions in fiscal 2025 (NACUBO and Commonfund 2026), which requires the balance to come from appreciation. Raising current income narrows that gap and reduces dependence on selling appreciated assets to fund distributions, which matters most during periods of depressed values when forced selling is costliest.

The index-linked escalation in institutional hangar leases further aligns the income stream with spending. Many spending policies incorporate inflation adjustment either explicitly or through the smoothing formula, and an income stream that escalates with inflation matches that obligation more naturally than fixed income does. This argument does not depend on the return assumption and is the most robust element of the case made in this paper.

6. Risk Factors Specific to Hangar Investment

Several risks in aviation real estate have no counterpart in conventional property. They are structural features of the asset class rather than contingencies, and they require specific mitigation.

6.1 Ground Lease and Reversion Risk

The distinctive risk is the leasehold structure. Hangars are typically constructed on airport authority land under ground leases granting occupancy for a defined period, commonly twenty to fifty years. At expiration, improvements revert to the authority without payment to the tenant. The investment therefore has a finite horizon and a terminal value that may be zero.

The effect of remaining term on value is not linear. As the remaining term shortens, lender constraints bind before the lease itself expires, because debt must be retired within the lease term with a margin, and conventional loan amortization schedules then cease to fit. The result is a range of remaining term over which financing availability deteriorates faster than the remaining income stream shortens, which reduces marketability and compresses the buyer pool. Caesar, Donner, and Kopsch (2019) measured the underlying effect in a residential leasehold market at approximately 0.22 percent of price per additional year of remaining term; the aviation case should be expected to be steeper because reversion is without compensation and the improvement has no use off the airport. Allocators can mitigate this by targeting properties with substantial remaining term, negotiating renewal provisions, and monitoring the term composition of the portfolio to avoid concentration in near-expiration assets (Carter 2026a).

6.2 Bargaining Position of the Airport Authority

The airport authority occupies two distinct positions relative to a hangar investor, and the distinction matters for how the risk is managed.

Before the lease is signed, the authority is a monopoly supplier. It controls the only land on the airfield, and federal policy constrains the through-the-fence arrangements that might otherwise supply an alternative. The investor's protection at this stage is the ability to walk away and develop at a different airport, which is real but weakens as the number of viable airports in a market falls.

After the improvement is built, the position reverses. The hangar cannot be moved and has no use off the airfield, so at reversion or at reset the authority is the only party that can acquire or continue to use it. At that point the authority is a monopsonist with respect to the improvement, and the investor has no alternative buyer. This is the fundamental transformation Williamson (1985) described: a market that was competitive before a relationship-specific investment becomes bilateral afterward, and the quasi-rents generated by the improvement become appropriable by the party controlling the complementary asset (Klein, Crawford, and Alchian 1978; Carter 2026b).

The practical consequences include exposure to above-market ground rent at reset, limited leverage on terms, and the risk that authority decisions on capital improvements, access, and services will not align with investor interests. Mitigation runs through diversification across airports and regions, through negotiating reset mechanics and renewal rights at the outset while bargaining position is strongest, and through the political economy that constrains authorities from extracting the maximum available rent, since doing so deters the private investment the airport system relies on. That last constraint is real but is not contractual, and an investor should not price it as though it were.

6.3 Illiquidity and Thin Market Risk

Hangar transactions are infrequent relative to conventional commercial real estate. The specialized improvements, the leasehold structure, and the limited buyer pool compound the illiquidity. Allocators should expect holding periods of seven to twelve years and should size positions so that liquidity needs can be met without a forced disposition.

The thin market also creates valuation uncertainty, as noted in Section 4.2. Appraisal-based valuations may lag market conditions, creating apparent outperformance during rising markets and delayed loss recognition during downturns (Geltner 1993). Dimmock, Wang, and Yang (2024) show that illiquid holdings drift furthest from target weights precisely when rebalancing would be most valuable, which is the mechanism by which an illiquidity premium can become an illiquidity cost.

6.4 Regulatory and Environmental Risks

Aviation real estate sits within a regulatory environment that includes federal oversight, state aeronautics regulation, environmental compliance, and local land use control. Contamination from legacy firefighting foam is a material environmental exposure at many airfields and should be diligenced specifically rather than as part of a general Phase I scope. Federal grant assurances attach to airport property for a period following each grant and constrain the sponsor's flexibility in ways that reach tenant operations and lease terms. Changes in any of these can affect operating costs, permitted uses, and value.

7. Implementation Considerations

7.1 Vehicle Selection

Endowments can access hangar real estate through several routes. Public equity investment offers daily liquidity and disclosure but introduces public market beta that may undermine the diversification the allocation is meant to provide, and at present the public route means a single small-capitalization company, which is a concentrated position in one management team rather than exposure to an asset class. Private fund vehicles offer more direct property exposure with reduced mark-to-market volatility but carry higher fees, longer lock-ups, and governance complexity. Direct investment and joint ventures provide control and fee efficiency but require operating expertise and concentrate risk.

For an initial allocation, a blended approach is reasonable: a core position through a diversified private vehicle or joint venture, with any public exposure sized as a liquidity sleeve and understood as company risk rather than sector risk.

7.2 Due Diligence Framework

Diligence for hangar investment extends beyond standard commercial real estate analysis. The aviation-specific items are the remaining ground lease term and its renewal provisions, the authority's documented behavior at previous resets and expirations, the airport's operational characteristics including runway length, approach capability, and based aircraft mix, the competitive facility supply within the draw area and the development pipeline, tenant credit and lease structure, regulatory compliance history, and environmental condition with specific attention to firefighting foam contamination. The operational characteristics of the airport itself deserve the same weight as the building, for reasons developed at length in the companion analysis of airport-level value drivers (Carter 2026b).

7.3 Performance Monitoring and Benchmarking

The absence of a performance benchmark is the practical obstacle to institutional adoption, and it should be treated as such rather than worked around. No index covers this property type. Until one exists, a dual-benchmark approach is workable: compare performance against a core-plus real estate benchmark, such as a diversified core fund index plus a spread, to evaluate risk-adjusted return relative to the broader real estate allocation, and separately track observable sector metrics including the public company's reported occupancy and rent, disclosed transaction activity, and survey data on waiting lists and rental trends.

An investment committee should be told directly that the second of these is not a benchmark. It is a set of indicators. A committee that cannot benchmark an allocation cannot evaluate the manager running it, and that is a governance cost that should be weighed against the diversification benefit rather than assumed away.

8. Limitations

Five limitations bear on how this analysis should be used.

Every hangar parameter is an assumption. There is no hangar return series, no appraisal index, and no transaction-based performance history. The expected return, standard deviation, and correlations in Table 1 and Appendix A are informed judgments, and the expected return in particular is the input the conclusion depends on most and the evidence supports least.

The optimization framework amplifies input error. Michaud (1989), Chopra and Ziemba (1993), and DeMiguel, Garlappi, and Uppal (2009) establish that mean-variance methods applied to estimated inputs systematically overweight the assets whose inputs happen to be estimated favorably. Section 5 is reported in rounded, directional terms for that reason, and no figure in it should be quoted to a decimal place.

The correlation matrix cannot be validated. Appendix A reports assumed correlations. The public company's trading history is too short, too thin, and too volatile to identify them, and no private return series exists against which to test them. The matrix is internally consistent and economically reasoned. It is not estimated.

The endowment model itself is contested. Barber and Wang (2013), Ang, Ayala, and Goetzmann (2018), Ennis (2021), and Dimmock, Wang, and Yang (2024) each raise substantive challenges to the proposition that heavier alternatives exposure is reliably rewarded, and the most recent cohort data does not resolve them. An allocator who does not accept that premise should not find this paper's conclusion persuasive, and that is a reasonable position rather than an error.

The sector has not been tested through a downturn. Institutional capital entered aviation real estate during a period of strong business aviation demand. The counter-cyclical behavior described in Section 3.3 rests on a single episode. The asset class has not been observed through a corporate profit contraction of the kind that would test the demand driver the entire thesis rests on.

9. Conclusion

Aviation hangar real estate has characteristics that make it a credible candidate for an endowment allocation. Supply is constrained by policy and by the economics of airport development. Demand is driven by variables that differ from those governing office and retail. Income is contractual, index-linked, and paid by creditworthy tenants. Management is simpler than in multi-tenant property. An institutional ecosystem now exists where a decade ago it did not.

The sensitivity analysis here indicates that under the assumptions set out, a 2 to 5 percent allocation improves the risk and return profile of a representative endowment portfolio and raises current income. That improvement is directional, not measured, and it depends most heavily on a return assumption for which no evidence exists. The finding that survives every set of assumptions tested is a narrower one: at 2 to 5 percent, the cost of being wrong about these parameters is bounded, and the current income benefit does not depend on the return assumption at all. That is a smaller claim than the one this literature usually makes, and it is the one the evidence supports.

The risks are real and structural. Reversion without compensation, the reversal of bargaining position once the improvement is built, illiquidity, and a regulatory environment the investor does not control are features of the asset class rather than contingencies within it. They are manageable through diversification, lease term selection, and terms negotiated at the outset, and they are not eliminable.

What would change this analysis from argument into evidence is specific and achievable. A transaction-based index covering hangar sales, even a small one, would permit return and volatility estimation. A panel of ground lease rates from multi-airport authorities, where the operator and policy environment are constant, would permit estimation of the lease-term and airport-characteristic effects that are currently asserted. And a classification for aviation facilities within the established institutional property indices would let allocators benchmark the position they are being asked to take. None of these requires new theory. All three require data collection that the sector has not yet undertaken, and until it does, an allocation to this asset class is a judgment about an argument rather than a decision supported by measurement. Committees should be told which one they are making.

References

Ang, A., A. Ayala, and W. N. Goetzmann. 2018. "Investment Beliefs of Endowments." European Financial Management 24 (1): 3-33. https://doi.org/10.1111/eufm.12149

Barber, B. M., and G. Wang. 2013. "Do (Some) University Endowments Earn Alpha?" Financial Analysts Journal 69 (5): 26-44. https://doi.org/10.2469/faj.v69.n5.4

Caesar, C., H. Donner, and F. Kopsch. 2019. "The Impact of Leasehold Status on Apartment Price." Journal of Housing Economics 46: 101629. https://doi.org/10.1016/j.jhe.2019.04.001

Carter, C. W. 2026a. "Ground-Lease Reversion Risk: Why a Hangar Can Be Worth Less Than It Cost to Build." Aeronautical Valuation Research Series, Valuation Takes Flight LLC. https://valuationtakesflight.com/research/ground-lease-reversion-risk.html

Carter, C. W. 2026b. "Cleared for Value: How Airport Characteristics Drive Hangar Value." Aeronautical Valuation Research Series, Valuation Takes Flight LLC. https://valuationtakesflight.com/research/airport-characteristics-and-hangar-value.html

Carter, C. W. 2026c. "Grounded Assets and Appropriable Rents: Hold-Up Risk in Hangar Ground Leases." Aeronautical Valuation Research Series, Valuation Takes Flight LLC. https://valuationtakesflight.com/research/grounded-assets-and-appropriable-rents.html

Chopra, V. K., and W. T. Ziemba. 1993. "The Effect of Errors in Means, Variances, and Covariances on Optimal Portfolio Choice." The Journal of Portfolio Management 19 (2): 6-11.

DeMiguel, V., L. Garlappi, and R. Uppal. 2009. "Optimal Versus Naive Diversification: How Inefficient Is the 1/N Portfolio Strategy?" The Review of Financial Studies 22 (5): 1915-1953.

Dimmock, S. G., N. Wang, and J. Yang. 2024. "The Endowment Model and Modern Portfolio Theory." Management Science 70 (3): 1554-1579.

Ennis, R. M. 2021. "Failure of the Endowment Model." The Journal of Portfolio Management 47 (5): 128.

Geltner, D. 1993. "Temporal Aggregation in Real Estate Return Indices." Journal of the American Real Estate and Urban Economics Association 21 (2): 141-166.

Giglio, S., M. Maggiori, and J. Stroebel. 2015. "Very Long-Run Discount Rates." Quarterly Journal of Economics 130 (1): 1-53.

Klein, B., R. G. Crawford, and A. A. Alchian. 1978. "Vertical Integration, Appropriable Rents, and the Competitive Contracting Process." Journal of Law and Economics 21 (2): 297-326.

Lerner, J., A. Schoar, and J. Wang. 2008. "Secrets of the Academy: The Drivers of University Endowment Success." Journal of Economic Perspectives 22 (3): 207-222.

Lindsey, T. J. 2008a. "An Introduction to the Valuation of Aircraft Hangars." The Appraisal Journal 76 (1): 26-38.

Lindsey, T. J. 2008b. "An Introduction to the Valuation of Aircraft Hangars, Part 2." The Appraisal Journal 76 (2): 132-154.

Michaud, R. O. 1989. "The Markowitz Optimization Enigma: Is Optimized Optimal?" Financial Analysts Journal 45 (1): 31-42.

National Association of College and University Business Officers and Commonfund Institute. 2026. 2025 NACUBO-Commonfund Study of Endowments. Washington, DC: NACUBO. https://www.nacubo.org/Research/2025/Public-NCSE-Tables

National Council of Real Estate Investment Fiduciaries. 2026. NCREIF Property Index and NFI-ODCE quarterly performance reports, second quarter 2026. Chicago: NCREIF.

Pagliari, J. L., K. A. Scherer, and R. T. Monopoli. 2005. "Public Versus Private Real Estate Equities: A More Refined, Long-Term Comparison." Real Estate Economics 33 (1): 147-187.

Sky Harbour Group Corporation. 2026. Annual Report on Form 10-K for the fiscal year ended December 31, 2025. U.S. Securities and Exchange Commission EDGAR.

Swensen, D. F. 2000. Pioneering Portfolio Management: An Unconventional Approach to Institutional Investment. New York: Free Press.

Swensen, D. F. 2009. Pioneering Portfolio Management: An Unconventional Approach to Institutional Investment. 2nd ed. New York: Free Press.

Williamson, O. E. 1985. The Economic Institutions of Capitalism. New York: Free Press.

Appendix A: Assumed Correlation Matrix

The following correlation matrix underlies the sensitivity analysis in Section 5. The traditional asset class entries are conventional long-horizon planning assumptions. The hangar row reflects the demand-driver reasoning in Section 3.3 and nothing else. It is not estimated from returns, because no hangar return series exists, and it is not supported by the trading behavior of the sector's public company, for the reasons given in Section 4.3.

Table A1. Assumed correlation matrix.

Assumed values, not estimates. Users substituting their own capital market assumptions should replace the entire matrix rather than the hangar row alone, since the row was reasoned against these particular values.

This paper is research and general information for professionals evaluating aviation real estate. It is not appraisal, legal, investment, or tax advice, and it does not create an engagement. Nothing here is a recommendation to buy or sell any security.

Related reading: are hangars a good investment, an ROI framework for the same asset class.