U.S. public pension funds held an estimated $3.3 billion in cryptocurrency-linked assets by mid-2025, according to analysis of 13F filings across seventeen major funds. That figure has grown through 2026 as state legislatures, federal executive orders, and the expansion of spot Bitcoin ETFs conve...
"Our portfolio is balanced and is well above our 6.5% annual growth target since I assumed office as State Treasurer." — Brad Briner, North Carolina State Treasurer
U.S. public pension funds held an estimated $3.3 billion in cryptocurrency-linked assets by mid-2025, according to analysis of 13F filings across seventeen major funds. That figure has grown through 2026 as state legislatures, federal executive orders, and the expansion of spot Bitcoin ETFs converge to push retirement capital toward digital assets. At the same time, critics including Better Markets have called these allocations "a risky gamble with public retirement security," and at least one state — Wisconsin — reversed course entirely, liquidating a $321.5 million Bitcoin ETF position in Q1 2025.
The trend is uneven, often small in dollar terms relative to total AUM, and subject to sharp political pressure from both directions. But the structural plumbing — ETF wrappers, 401(k) rule changes, and state-level mandates — is being built. Whether pension beneficiaries benefit or absorb losses remains an open question that the data does not yet resolve.
As of the most recent 13F filings available (Q2 2026), the following U.S. public pension systems report cryptocurrency-linked holdings:
| Fund | Vehicle | Shares | Approx. Value | AUM | |------|---------|--------|---------------|-----| | CalPERS | Strategy (MSTR) | 410,596 | $165.9M | ~$500B | | Teacher Retirement System of Texas | MSTR + direct allocation | 80,844 MSTR + $400M allocation | $412M+ | ~$200B | | Michigan State Retirement System | ARKB + MSTR | 300,000 ARKB + 14,000 MSTR | ~$12.7M | ~$90B | | Florida Retirement System | Strategy (MSTR) | 240,026 | ~$97M | ~$250B | | Louisiana State Pension | Strategy (MSTR) | 21,300 | ~$2.1M | ~$16.3B | | North Carolina Retirement Systems | Various crypto-linked | Undisclosed | Down ~$30M on paper | ~$140B |
These figures are small relative to total assets under management. CalPERS' $165.9 million MSTR position represents approximately 0.03% of its portfolio. Michigan's $12.7 million across ARKB and MSTR rounds to 0.014%. The Teacher Retirement System of Texas is an outlier: its $400 million direct crypto allocation, the single largest pension fund crypto commitment disclosed to date, represents roughly 0.2% of its $200 billion in assets.
Pension funds access crypto exposure through three distinct mechanisms, each with different risk profiles and governance implications.
1. Equity Proxies (MSTR/Strategy)
The most common channel. Strategy Inc. (formerly MicroStrategy) held 687,410 BTC on its balance sheet as of its Q2 2026 earnings report, with Bitcoin holdings surging 11% quarter-over-quarter. Pension funds buying MSTR acquire indirect Bitcoin exposure bundled with Strategy's operating business, leverage profile, and management decisions. Pension MSTR concentration rose from 17.3% of crypto-linked holdings in Q1 2024 to 58.4% by Q1 2025, before settling at 43.2% by Q4 2025, according to academic analysis of 13F filings.
2. Spot Bitcoin ETFs
U.S. spot Bitcoin ETFs collectively hold approximately $76.4 billion in AUM and 1,212,749 BTC as of July 31, 2026. The category has accumulated $58.7 billion in cumulative net inflows since the January 2024 launch. BlackRock's IBIT dominates with roughly $54 billion — approximately 49% of total category AUM. Morgan Stanley's MSBT, launched April 8, 2026 at a 0.14% fee (undercutting IBIT by 11 basis points), drew $34 million on day one and $100 million in its first week. Michigan's tripling of its ARKB position to 300,000 shares ($11.4 million) in Q2 2026 illustrates how ETFs simplify pension board approvals relative to direct custody.
3. Direct Allocation
Rare but expanding. The Teacher Retirement System of Texas' $400 million allocation is the largest known example. Indiana's House Enrolled Act 1042, signed March 3, 2026, requires the $54.9 billion Indiana Public Retirement System (INPRS) to offer self-directed brokerage accounts with at least one crypto investment option by July 1, 2027. The law covers Hoosier START (the state's 457(b) and 401(a) plans) and specified teacher retirement accounts.
Three federal and state-level policy shifts are expanding the institutional corridor for pension crypto exposure.
Trump Executive Order (August 7, 2025). Directed the Department of Labor to reexamine regulations governing retirement plan investments, opening the $8.7 trillion 401(k) market to alternative assets including cryptocurrency. The DOL proposed a formal rule in early 2026. Senator Troy Downing (R-Mont.) introduced legislation to codify the order into statute.
DOL Guidance Rescission (May 2025). The Department of Labor withdrew its 2022 cautionary guidance (Compliance Assistance Release No. 2022-01) that had warned 401(k) fiduciaries against offering crypto options. The rescission removed the primary regulatory chill on plan-level crypto inclusion.
State Mandates. At least 21 states have introduced or passed legislation enabling public pension crypto exposure as of mid-2026, according to research by Reason Foundation. Indiana mandates it. Texas Governor Greg Abbott signed a Strategic Bitcoin Reserve law in June 2025; the state executed its first IBIT purchase ($5 million) in November. North Carolina's House passed a bill to expand crypto investments in the state pension plan.
The State of Wisconsin Investment Board (SWIB) provides the clearest case study in pension crypto volatility risk. SWIB was the first U.S. state pension fund to hold a spot Bitcoin ETF, disclosing $163 million in IBIT in May 2024. By Q4 2024, the position had grown to over 6 million IBIT shares worth approximately $321.5 million.
In Q1 2025, SWIB liquidated the entire position — all 6,060,351 IBIT shares — and exited completely. The fund retained 127,528 MSTR shares worth approximately $36.8 million, a fraction of its prior crypto footprint.
The reasons for the exit were not publicly disclosed. SWIB's 13F filing is a regulatory document, not a strategy memo. But the sequence — rapid scaling from $163 million to $321 million, followed by complete liquidation — illustrates the governance challenges pension boards face with volatile assets. SWIB is not a hedge fund. Its beneficiaries are state employees expecting defined-benefit retirement payments.
Better Markets published a report in August 2025 titled "State Pension Fund Investment in Cryptocurrency: A Risky Gamble with Public Retirement Security." The core arguments:
North Carolina's experience provides a partial illustration. The state's pension plan reported approximately $30 million in paper losses on its crypto-linked positions as of early 2026, according to WRAL. The loss represents less than 0.02% of the $140 billion fund, and Treasurer Brad Briner noted the portfolio remains above its 6.5% annual growth target. But the political optics of any crypto loss in a pension context attract disproportionate scrutiny.
Proponents point to structural adoption data rather than price forecasts.
ETF scale. Spot Bitcoin ETFs reached $200 billion in AUM within 28 months of launch — described by ETF analyst Eric Balchunas as "the fastest ETF accumulation in history." The product wrapper eliminates custody, private-key, and direct regulatory exposure for institutional allocators.
Corporate treasury growth. Some 190+ public companies hold a combined 1.2 million BTC (approximately 6% of Bitcoin's 21 million fixed supply), according to data compiled by Dakota Research. This creates a liquid equity market for indirect exposure.
Wirehouse entry. Morgan Stanley's MSBT launch marked the first spot Bitcoin ETF issued directly by a major U.S. commercial bank. Merrill Lynch is reportedly preparing a comparable product. Wirehouses provide distribution infrastructure to wealth management clients whose accounts often include pension rollovers and IRAs.
Fee compression. MSBT's 0.14% expense ratio set a new floor for Bitcoin ETF fees, suggesting the product is being positioned for volume-based, long-term institutional capital — not speculative retail flows.
Several critical questions remain unanswered by current disclosures.
Performance attribution. No public pension fund has published a standalone performance report isolating crypto returns from overall portfolio returns. The actual contribution of crypto-linked holdings to funded status is unknown.
Rebalancing discipline. 13F filings show point-in-time snapshots. Whether funds are systematically rebalancing crypto positions or allowing them to drift is not visible from regulatory filings alone.
Beneficiary impact. No actuarial analysis has been published showing the effect of crypto allocation on funded ratios or contribution rates for any U.S. public pension system.
Counterparty and custody risk. Pension funds using MSTR as a crypto proxy inherit Strategy's capital structure, including approximately $8.2 billion in convertible debt. A Bitcoin downturn that triggers margin calls or share dilution transmits directly to pension portfolios.
The entry of U.S. public pension capital into crypto markets is measurable but modest. A $3.3 billion aggregate allocation spread across funds managing trillions does not constitute a structural shift in asset allocation. It constitutes a test.
The infrastructure being built — ETF wrappers, regulatory clearances, state mandates, wirehouse products — suggests the test is meant to scale. Indiana's mandate requiring crypto options in pension accounts by July 2027 creates a legal obligation for exposure. The DOL rule change opens defined-contribution plans to crypto for the first time.
Whether this infrastructure channels meaningful capital into digital assets or remains a rounding error depends on two variables the data cannot predict: Bitcoin price performance over retirement-relevant time horizons, and the political tolerance for crypto losses in public pension portfolios. Wisconsin's exit demonstrates that the second variable has a low threshold.