Limited partners (LPs) are the backbone of private equity (PE) markets, providing the capital that general partners (GPs) deploy across investments. Yet the term “LP” encompasses a heterogenous set of institutions—public and private pension funds, endowments, foundations, insurance companies, sovereign wealth funds, and family offices, among others—whose surface-level similarity as capital allocators masks differences in how and why they invest.
Each LP operates under a distinct set of constraints that shapes their approach to portfolio construction. Public pension funds face statutory investment guidelines and public accountability obligations, and must manage assets against long-dated, relatively predictable liabilities like the future retirement payments of their beneficiaries. Their historically risk-averse nature has anchored them to fixed income assets, though many have shifted significantly toward alternatives in search of higher returns as funding gaps have widened.[1] Private sector pension funds face similar liability structures but often enjoy greater flexibility in investment mandates.
Insurance companies are among the most constrained institutional investors. Risk-based capital regulations (such as Solvency II in Europe or RBC frameworks in the US)[2] assign capital charges to different asset classes, making illiquid or equity-like investments more expensive from a regulatory capital perspective. This creates a structural bias toward investment-grade fixed income that no amount of return-seeking can easily overcome.[3]
Endowments and foundations, on the other hand, are relatively less constrained in their investment approaches. With perpetual time horizons and a less rigid liability structure, they can tolerate illiquidity more than pension funds or insurers. This has positioned them as leaders among institutional investors who build alternatives-heavy portfolios. Their primary constraint is the spending rule (typically a 4-5% annual distribution)[4] which demands sufficient long-run real returns rather than short-term income generation.
Sovereign wealth fund investment theses vary significantly from entity to entity, but most target the general mandate of producing long-term financial returns while supporting domestic economic growth strategies.[5] Family offices, meanwhile, are the least constrained of all institutional allocators. Absent external regulation or beneficiary obligations, their portfolios reflect the risk preferences, tax considerations, and legacy objectives of the underlying family. This idiosyncratic nature of family offices has produced the widest dispersion of allocation strategies of any LP type.
As shown in Figure 1, allocation patterns vary sharply across LP types in ways that reflect the previously discussed structural constraints, regulatory obligations, and risk tolerances.[6] Endowment plans, family offices, and foundations, which are the least constrained institutional allocators, exhibit the highest median allocations to both alternatives and PE. They also tend to pair these with high equity allocations. Rather than substituting risk across asset classes, these LPs appear to run uniformly risky portfolios, suggesting that risk is set at the institutional level and distributes consistently across the entire portfolio.
Insurance companies sit at the opposite extreme, allocating very little to alternatives or private equity and concentrating instead in fixed income. This is a direct consequence of the capital charges that risk-based regulatory frameworks impose on illiquid and equity-like holdings. Pension funds present a more nuanced picture. Despite carrying a high median equity allocation, they lag endowments and foundations in alternatives and PE exposure. This tension could be explained by the liquidity requirements, liability-matching obligations, and governance constraints that limit their allocation decisions.
To further understand the relationship between LP investor type and allocations to PE, we run a simple ordinary least-squares regression (OLS)[7] of the effects of firm type and size of assets under management (AUM) of the investor on the LP’s actual PE allocation (%). We use pension funds, a relatively risk-averse investor,[8] as the baseline, with the results shown in Figure 2 below.[9]
A pension fund with an average AUM (estimated at $18,276.1M) has a predicted PE allocation of 8.64%. Family offices show the largest difference, allocating 24.5 percentage points more to PE than a similarly sized pension fund. This is highly significant and consistent with their unconstrained mandates and return-seeking orientation. Endowment plans and foundations are also associated with higher PE allocations, at 9.7 and 8.97 percentage points above the baseline pension fund respectively, consistent with their longer investment horizons and higher tolerance for illiquidity. Insurance companies are the only investor type with lower PE allocations than pension funds, at -3.93 percentage points, likely reflecting (at least in part) the regulatory capital constraints discussed above. Sovereign wealth funds show a positive but statistically insignificant coefficient (2.38 percentage points), which may reflect the comparatively smaller sample of sovereign wealth funds in the data.
Log AUM is small and statistically insignificant (-0.07 percentage points), meaning that once LP type is controlled for, size has no meaningful independent association with PE allocation. This suggests that institutional mandate, rather than scale, is the stronger correlate of PE allocation behavior.
The relationship between LP risk tolerance and PE allocation has attracted growing academic attention. Giommetti and Sorensen (2024) analyzes the gap between actual and theoretically optimal PE allocations across LPs of different risk profiles and finds a clear asymmetry.[10] While conservative LPs allocate close to their optimum, committing on average 35.6% of their portfolio to PE, aggressive LPs tend to underallocate, committing on average 53.4% of their portfolio to PE against an optimum of 100%.[11] The implication is that the LPs best positioned to bear illiquidity risk are leaving the most return on the table, constrained by mandates that optimization models do not capture.
Institutional mandate, which determines investor risk appetite, is a primary driver of LP portfolio construction. This takeaway translates into meaningful implications for fund managers, particularly concerning fundraising strategies. Nevertheless, mandate doesn’t tell the full story, as investor size introduces a second binding constraint. An allocation ask that is entirely reasonable for a large endowment may be structurally impossible for a small foundation. However, these boundaries are not fixed, as sustained periods of low yields in fixed income have already pushed historically conservative LPs towards private markets, and as the risk-return profile of PE continues to evolve, the LP universe may broaden further still.
[1] Seb Murray, “Why More Public Pensions Are Taking a Chance on Alternative Investments,” Stanford Business, July 22, 2025, https://www.gsb.stanford.edu/insights/why-more-public-pensions-are-taking-chance-alternative-investments.
[2] “Solvency II,” European Insurance and Occupational Pensions Authority, accessed on May 26, 2026, https://www.eiopa.europa.eu/browse/regulation-and-policy/solvency-ii_en.; “Risk-Based Capital,” NAIC, last updated October 2, 2025, https://content.naic.org/insurance-topics/risk-based-capital.
[3] Stijn Van Nieuwerburgh and Mark Perez Clanton, “The Intersection of Private Equity, Insurance Companies, and Commercial Real Estate Debt,” Columbia Business School, June 3, 2025, https://business.columbia.edu/milstein-center-research-lab/milstein-center/intersection-private-equity-insurance-companies-and.
[4] “Your Endowment Questions, Answered,” NACUBO, February 12, 2026, https://www.nacubo.org/Topics/Endowment-Management/Your-Endowment-Questions-Answered.
[5] “The rise of sovereign wealth funds in private markets,” CFA Institute, May 20, 2026, https://www.cfainstitute.org/insights/articles/sovereign-wealth-funds-investment-private-markets.
[6] Data sourced from Preqin Pro on May 20, 2026. Includes global LPs and was filtered to include only the top six LP types by PE allocation. Data restricted to only those LPs that have updated their allocations within the past two years. Each box represents the interquartile range of reported allocations, with the upper and lower bounds indicating the 75th and 25th percentiles respectively, and the horizontal line denoting the median.
[7] OLS regression estimates the relationship between a dependent variable and one or more independent variables, producing coefficients that represent the average change in the dependent variable associated with a one-unit change in each independent variable, holding all others constant. Here, each coefficient captures the percentage point difference in PE allocation relative to the pension fund baseline, after accounting for AUM.
[8] Pension funds represent the largest sample of PE-allocating LPs in the dataset, making them a natural baseline for comparison.
[9] Data sourced from Preqin Pro on May 20, 2026. Includes global LPs and was filtered to include only the top six LP types by private equity allocation. Data restricted to only those LPs that have updated their allocations within the past two years.
[10] Nicola Giommetti and Morten Sorensen, “Optimal Allocation to Private Equity,” Tuck School of Business Working Paper No. 3761243, May 3, 2024, https://dx.doi.org/10.2139/ssrn.3761243.
[11] Giommetti and Sorensen (2024) defines an aggressive LP as one with low relative risk aversion (γ = 1), which seeks maximum exposure to risky assets and consequently faces significant liquidity constraints that prevent it from achieving its first-best allocation, resulting in a substantially underallocated PE position. A conservative LP, by contrast, has high relative risk aversion (γ = 3) and naturally holds a larger share of safe assets, which means it is rarely liquidity-constrained and can effectively replicate its first-best allocation when those investments are illiquid, treating PE as simply another traded stock that adjusts with a lag. To understand allocation decisions, the authors use a discrete-time, infinite-horizon portfolio model with constant relative risk aversion preferences and a “linear fund dynamics” assumption borrowed from Takahashi and Alexander (2002) that simplifies PE cash flows into linear functions of uncalled commitments and NAV.
