The U.S. public pension system spans more than 800 plans and $9 trillion in AUM, but for practical purposes it is defined by fewer than 100 institutions. The top 24 plans above $100B hold nearly 60% of all assets. Everything meaningful in private markets flows from that concentration.
Two decades of alternatives adoption have reshaped how these plans invest. Plans above $25B now average 37% in alternatives, up from roughly 10 to 15% in 2005, driven initially by the return assumption gap and sustained by the risk management and liability-matching properties that private markets provide. The consultant layer determines where capital flows for most of the universe. CIO turnover is constant, and the relationships that survive it are the ones built below the CIO level.
Private markets deployment is accelerating across a broader base of plans. The growth is no longer just the mega-allocators scaling existing programs. More plans are entering the asset class for the first time.
Private credit has arrived. It is now a core allocation at the most sophisticated plans, not a satellite position, and a large share of the universe has yet to adopt it at all.
The empirical case for alternatives is more nuanced than it is typically marketed. The data supports alternatives as a risk management tool. The long-run alpha case is harder to make.
The consultant is the real gatekeeper. For most plans, the general consultant determines which managers get evaluated. Knowing who advises a target plan is critical.
Dakota Marketplace tracks over 800 U.S. public pension plans with over $9 trillion of combined AUM. The shape of the universe is severely top-heavy.
Two-thirds of the plans in the universe sit below $1B in AUM. Many are municipal pension plans, small county systems, and special-purpose funds (drainage districts, port authorities). They are governed and managed differently from the larger plans, typically through a small in-house team or a consultant relationship that is increasingly being upgraded to a discretionary OCIO arrangement. They are also largely outside the institutional private markets ecosystem; most operate fully or predominantly in liquid public mandates.
The opposite end (the 74 plans above $25B in AUM) represents 87% of total system assets and almost all of the deployable capital for institutional alternatives. These plans run substantial in-house investment teams, engage specialty consultants alongside their general consultant, and increasingly run co-investment and SMA programs.
U.S. public pension plans fall into the below categories that shape how they invest and how they engage with managers:
Statewide Retirement Systems (e.g., CalPERS, NYSCRF, Florida SBA, Virginia Retirement System): the largest, most professionalized funds, typically $50B+ with substantial in-house staff.
Statewide Teacher Systems (e.g., CalSTRS, Texas TRS, NY State Teachers, Ohio STRS, TRS Illinois): distinct from general statewide systems. Teacher plans typically have their own boards and consultants.
Municipal and City Systems: run by city or county investment offices. The largest, NYC Comptroller's Office (which manages five city pension funds), operates at scale; most others are smaller and more consultant-reliant.
County Systems (e.g., LACERA, OCERS, Ventura County, Fresno County): often '37 Act' plans in California, usually $1B to $25B, consultant-led with small to mid-sized in-house teams.
Investment Authorities (e.g., Washington SIB, SWIB, NMSIC, Texas Treasury Safekeeping): centralized investment arms that manage assets on behalf of pension boards plus other state trust funds.
U.S. public pension plans have spent the better part of two decades building private market allocations. In 2005, the average large plan held roughly 10 to 15% of assets in alternatives broadly defined. By 2015, that figure had climbed to the mid-20s. Today, plans above $25B average approximately 37% in alternatives, and the largest and most sophisticated funds run private market programs spanning private equity, private credit, real assets, infrastructure, and venture capital, often with dedicated co-investment and separately managed account sleeves layered on top.
The growth has not been uniform. Private equity led the first wave, with most large plans establishing meaningful PE programs through the 2005 to 2015 period. Real assets and infrastructure followed as plans sought inflation sensitivity and liability-matching characteristics that equity-like alternatives could not provide. Private credit is the most recent major adoption cycle: a decade ago it was a niche allocation at a handful of the most sophisticated plans; today it represents the second-largest alternatives category by disclosed capital, sitting at 24% of all reported commitments collected by Dakota since the beginning of 2025.
The driver throughout has been the return assumption gap. Most public pension plans carry actuarial return assumptions in the 6.8 to 7.5% range. After the 2008 financial crisis pushed fixed income yields to historic lows, maintaining those assumptions with a traditional 60/40 portfolio became mathematically difficult. Alternatives offered the yield and return premium necessary to make the math work, and boards approved increasingly aggressive pacing targets throughout the 2010s. The post-2022 rate reset has partially rehabilitated public fixed income, but the institutional infrastructure built around private markets is now deeply embedded and continues to grow.
Note: totals reflect reported activity during each quarter, not a complete tally of dollars committed in that same window, since larger plans often report prior-quarter commitments in later board materials.
From 2025 through Q1 2026, Dakota tracked approximately $309 billion in private market commitments across 2,880 commitments. Quarter-over-quarter, the trend is one of consistent expansion:
QUARTERLY CAPITAL COMMITTED BY ASSET CLASS ($B)
The headline commitment figures are only part of the picture. How plans deploy, manage, and govern private market allocations in practice varies considerably and shapes what it means, operationally, to be an active private markets investor at this scale.
Large plans manage private markets through multi-year pacing programs, committing capital annually to stay on target allocations despite the lag between commitment and deployment. Because capital is drawn down over two to four years and distributions begin years later, plans must commit well ahead of their target exposure to avoid being perpetually underweight. The J-curve effect, where early-period fees and costs produce negative returns before the portfolio matures, means a plan's private markets book is always a mix of vintages at different stages of the value creation cycle. Managing the vintage mix is as important as manager selection for many CIOs.
The illiquidity of private market commitments creates a liquidity management challenge. Plans with significant alternatives programs must ensure they maintain enough liquid assets to cover benefit payments, rebalancing, and opportunistic deployment without being forced sellers. This is why even the most alternatives-heavy plans rarely push beyond 50 to 55% in illiquid assets, and why unfunded commitments are tracked as carefully as NAV exposure. The post-2022 denominator effect, where public market drawdowns inflated alternatives' share of total portfolios, forced several plans to pause new commitments or slow pacing, a phenomenon visible in moderated 2022 to 2023 vintage counts among some of the most aggressive programs.
The largest plans have materially expanded co-investment programs over the past five to seven years. Co-investments, which are direct investments alongside a fund manager in a specific deal (typically at reduced or zero fees) offer three advantages: fee savings at scale, the ability to deploy larger amounts of capital per relationship than a fund commitment alone allows, and direct deal experience that builds internal staff capability over time. CalPERS, CalSTRS, NYSCRF, and Texas TRS all run active co-invest programs.
At the very top of the market, the largest plans have moved beyond fund commitments and co-investments into fully customized separate account structures with their largest manager relationships. These SMAs give the plan direct control over portfolio construction within the strategy, typically in exchange for a meaningful capital commitment ($500M+) and a long-term relationship. Blackstone, KKR, Ares, and a handful of other mega-managers have built significant SMA businesses on the back of large pension relationships. For the plan, the SMA offers fee efficiency, transparency, and the ability to exclude certain investment types; for the manager, it provides stable, long-duration capital.
Sophisticated plans increasingly use the secondary market on both sides of the ledger. On the sell side, LP-led secondary transactions allow plans to exit fund stakes before the end of a fund's life, which reduces manager count, trims underperforming relationships, and can free up capacity for new commitments without waiting for natural distributions. The denominator effect accelerated this: several large plans that found themselves overallocated to alternatives in 2022 to 2023 used secondary sales to reduce exposure rather than simply pausing new commitments. On the buy side, a growing number of the largest plans, including CalPERS and NYSCRF, have built dedicated secondary allocations or co-invest in GP-led continuation vehicles, using the secondary market as an entry mechanism that offers compressed J-curves and faster capital deployment than primary commitments. The secondary market has matured from a distressed-seller's last resort into a standard portfolio construction tool, and for the most sophisticated programs it now factors directly into pacing decisions.
Public pension investment offices vary enormously in scale and sophistication. Some of the largest groups run large in-house teams of 100 to even over 200 investment professionals organized by asset class. Mid-sized plans typically run 10 to 50 investment staff. Plans below $5B often have only a handful of investment officers and rely heavily on their general consultant.
Most U.S. public plans share a common governance template, with significant variation in execution:
Board of Trustees. Typically 7 to 15 members in mixed composition: elected member representatives, gubernatorial or legislative appointees, ex-officio public officials (state treasurer, comptroller). The board sets strategic asset allocation, approves the investment policy statement, hires and oversees the executive director and CIO, and ratifies large manager hires. Board composition meaningfully shapes investment behavior: boards weighted toward investment professionals tend to delegate more authority to staff; boards weighted toward political appointees and member representatives tend to retain direct control over manager selection, and tend to produce more CIO turnover.
Sole-Trustee Model (Rare). The New York State Common Retirement Fund is the most prominent example: the State Comptroller is the sole fiduciary, with no board of trustees. This concentrates authority and shortens decision cycles but also concentrates political risk in a single elected official. Texas Treasury Safekeeping Trust operates under a similar structure.
Investment Committee. A subcommittee of the board, often supported by independent fiduciaries or external advisors, that focuses specifically on investment policy, manager selection, and performance review. Typically meets monthly or bi-monthly.
Executive Director / CEO and CIO. The ED runs the system as an organization (benefits administration, member services, IT, HR, legal); the CIO runs the investment office. Reporting lines vary in ways that meaningfully affect CIO authority. CalPERS' CIO reports to CEO Marcie Frost. NYSCRF's CIO reports to the Comptroller directly. Florida SBA splits authority between an Executive Director and a CIO. At plans with stronger CEO authority over the CIO, manager relationships built through investment staff tend to survive CIO transitions more cleanly.
The Investment Committee is where strategic asset allocation gets set and where the largest manager commitments get approved. The agenda is typically built jointly by the CIO's office and the general consultant. Three categories of business consume most meetings:
Performance reporting against benchmarks and peer groups, usually presented by the consultant. This is where individual asset class performance gets put into peer-group context.
Policy decisions: IPS amendments, asset allocation rebalancing, manager structure reviews, and asset-liability study results (typically refreshed every 3 to 5 years).
Investment recommendations: specific manager hires and structural decisions, almost always brought jointly by staff and the consultant.
Public meeting agendas are an underused information asset. NYSCRF publishes monthly investment and transaction reports; CalSTRS publishes quarterly Portfolio Risk Reports; CalPERS publishes IC agenda items and board webcasts. This is the cleanest available window into how the largest investment offices actually deliberate. All of this information is available on pension plan profiles in Dakota Marketplace.
CIO authority sits on a spectrum. At one end, plans like CalPERS, CalSTRS, and NYSCRF delegate substantial day-to-day commitment authority to staff under board-approved policies. The CIO is empowered to commit hundreds of millions per fund without IC pre-approval, subject to public disclosure post-commitment. At the other end, smaller and more politically constrained plans require board ratification for most manager hires.
Three trends are reshaping CIO authority (2020 to 2025):
Total Portfolio Approach (TPA): Collapses asset-class silos, gives CIO greater discretion. CalPERS announced full implementation for July 2026. CalSTRS is transitioning under CIO Scott Chan. Texas TRS CIO Jase Auby has publicly questioned TPA, a meaningful philosophical divide.
Longer-tenured CIO compensation packages: CalPERS' CIO compensation reaches $2.4M max with five-year performance vesting, designed to extend tenure beyond the historical 18-month-to-4-year norm.
Expanded co-investment and SMA programs: The largest plans now give CIOs and senior staff direct deal-by-deal selection authority alongside fund commitment authority.
Q1 (Jan–Mar): Asset Allocation Review & Pacing Study. Annual review with consultant. Capital markets assumptions refreshed; 12-month commitment plan set by asset class.
Q2 (Apr–Jun): Heavy Commitment Quarter. Peak deployment, particularly for June 30 FY-end plans (~75% of universe). Manager pipelines reviewed; commitments approved before year-end.
Q3 (Jul–Sep): Performance Reporting & Forward Planning. FY-end performance reporting for June 30 plans; next year's pacing refreshed. CalPERS, CalSTRS, NYSCRF release preliminary returns.
Q4 (Oct–Dec): Calendar Year-End Commitments. Commitment activity for Dec 31 year-end plans. Annual investment plan adopted for the following year.
The consultant is, for most plans, the gatekeeper to manager selection. They screen, qualify, and recommend, and at smaller plans they functionally drive the decision. The vast majority of pension plans use a consultant, and plans also hire specialist consultants for specific asset classes.
Meketa has developed a strong position at the top of the market. 12 of the 45 plans above $50B use Meketa as general consultant, giving the firm outsized influence over where institutional capital flows.
NEPC is the most evenly distributed, top-five across all three tiers. That breadth makes NEPC's platform unusual: few firms maintain genuine relevance from the mega-allocators down through the long tail.
Mariner Institutional's 95 plans under $5B represents a different business model. Marquette and Dahab operate similarly, with Dahab particularly active among smaller public safety and municipal plans.
Consultant coverage strategy should be tier-specific. At plans above $50B, the general consultant is necessary but not sufficient; large plans layer in specialty consultants (Cambridge, StepStone, Aksia, Albourne). At plans below $5B, the general consultant typically drives manager selection.
Dakota collected current allocation data from a sample of 128 plans spanning the universe. The sample provides a broader picture of allocation behavior than aggregate dollar-weighted surveys.
Key observations:
Public equity dispersion is enormous, from 20% (a private-markets-heavy fund like Oregon PERF) to a plan allocating over 90%.
Real assets is the most universally adopted alternative, with only 2% of plans holding zero exposure.
Private credit remains a partial-adoption asset class. 37% of plans hold no private credit.
Private equity is broadly adopted, but the median allocation is still just under 10%.
Alternatives exposure scales with plan size but the relationship is not linear. The biggest step-up (+7.5pp) occurs between under-$1B and $1 to $10B, the boundary where plans begin to have institutional-quality private markets programs. Above $25B it plateaus at ~37%.
Dakota assembled a performance sample with reported alternatives allocation alongside 3, 5, and 10-year returns through December 31, 2025. The findings call into question one of the more common framings in industry discussion: that public pensions need alternatives to hit their long-run return targets.
The 3-year correlation is sharply negative. Over 2023 to 2025, plans with higher alternatives allocations underperformed, driven by strong public equity returns. The 5-year correlation flips: including 2022's drawdown, plans with more alternatives benefited from the cushion private markets provided. The 10-year correlation is not significant.
The 10-year column is essentially flat: 8.4% in Q1, 8.8% in Q5, a spread of just 40 basis points across quintiles. Economically meaningful but smaller than the long-run alpha case implied by most institutional alternatives marketing.
The return-target framing assumes total fund return maximization is the primary objective. For most public pension plans, it is not. Public pensions optimize for meeting future benefit payments with high probability, maintaining funded status across market cycles, and reducing contribution volatility. Those goals lead portfolio construction in a direction that does not always point toward the highest-returning asset classes.
The 40 basis points separating Q1 from Q5 over 10 years likely understates the intentionality behind the gap. Plans with 50% alternatives allocations are running a different portfolio with a different objective function, one where return is one input rather than the primary output.
The empirical case for alternatives is more accurately framed around risk reshaping and liability management than long-run alpha generation. The 5-year correlation supports a drawdown protection story. The 10-year correlation does not support a total fund outperformance story. The mechanism by which alternatives deliver outperformance may exist at the individual asset class level without aggregating to outperformance at the total fund level.
Leadership turnover is a persistent feature of the U.S. public pension system and one of the most consequential dynamics shaping institutional relationships. Dakota Marketplace tracks CIO job changes, including new appointments, retirements, interim roles, active transitions, and departures to the private sector. Below are some notable moves so far this year.
The most visible pattern we have noticed over the past several years is senior pension investment leaders leaving for private market firms. The reason is straightforward: even CalPERS' $2.4M maximum CIO package is well below what private firms pay for comparable talent. That pay gap isn't closing, so CIO turnover should be treated as a permanent feature of the system rather than something that will resolve. Just as important is movement between public plans. When a senior person leaves one plan, they often show up at another within a year or two.
The smoothest transitions follow a simple pattern: promote from within and hire someone to backfill the deputy role. Maryland SRPS did this well (Palmer retired, Burd stepped in as interim, Sandoval was hired as permanent CIO, Burd stayed on as deputy). The practical takeaway is that relationships built around a single CIO will reset every few years. Relationships that run deeper, to the deputy CIO, asset class heads, and senior portfolio managers, are the ones that survive transitions.
Private markets deployment will keep growing, but the next phase looks different. The marginal growth is coming from mid-market plans in the $1 to $10B range building out first-generation alternatives programs, not from the mega-allocators scaling existing ones. As those programs mature, often through OCIO structures or upgraded consultant relationships, the pool of active private market participants will widen considerably.
Private credit has the clearest structural runway: the share of plans holding zero exposure remains large, and the plans already in the asset class are scaling. Real estate's sharp recovery in Q1 2026 is the single data point most worth monitoring. Infrastructure's consistent quarter-over-quarter growth looks durable regardless; the policy and liability-matching tailwinds are not going away.
CIO turnover will remain constant; the compensation gap is not narrowing. The plans that manage transitions well do so through deep teams and deliberate succession. For anyone building durable relationships in this channel: coverage needs to run deeper than the CIO, and it needs to start before the transition, not after.
Finally, the Total Portfolio Approach debate will shape how the largest plans structure manager engagement over the next two years. CalPERS' full TPA implementation is projected for July 2026, and its execution will either accelerate adoption or validate the skeptics. TPA collapses the asset-class-specific decision structures that most managers have built their coverage models around, and a system-wide shift would require a meaningful rethink of how fund managers engage with their largest allocators.
This report was built entirely on data from Dakota Marketplace — the most comprehensive private markets database designed for the institutional investment community. Every allocation figure, consultant relationship, CIO transition, and performance data point was researched and verified by Dakota's 60-plus person data team. For further insights, visit dakota.com.
Dakota Marketplace is a database of LPs, GPs, Private Companies and Public Companies used by fundraising teams and deal sourcing teams to both raise money as well source investment opportunities. Over 1,300 global investment firms use Dakota daily.