By Stephen Stonberg, CEO of Tabit Insurance

Most insurers exclude “risky” Bitcoin from their balance sheets, yet comfortably hold nearly a third of their assets in opaque private credit. Bitcoin is among the most frequently and transparently priced assets available; private credit is among the least. It may be time for the Insurance Industry to consider “actual” vs “perceived” reputation risk.

The insurance industry depends on a core premise: when a future obligation becomes due, capital will be there to meet it. That is why insurers have typically shunned Bitcoin, deeming it too volatile and speculative for conservative balance sheets. Yet many have opened those same balance sheets to the $1.8 trillion private credit market, where risks are often opaque, infrequently priced, and difficult to verify. As a publicly traded asset, BTC volatility is transparent. Private credit is just that – private – and is not subject to market pricing. The actual valuation (what a buyer would pay in a forced liquidation) in a time of stress may be very different than the book value reported on the insurance balance sheet. Summer 2026 has some private credit funds gating (like Summer 2008), and Michael Burry recently called the 2026 version of the AI private credit / insurance bubble “fugazi.” U.S. prosecutors have now opened an inquiry into affiliated private credit holdings at two U.S. life insurers, which are racing to divest the assets to avoid potential credit rating downgrades.

Private credit has become deeply embedded in the insurance industry. A recent Moody’s report warned about this rise, finding that U.S. life insurers held $807 billion in some of the least liquid fixed-income instruments last year. As banks pulled back from lending after the global financial crisis, insurers, particularly life insurers and annuity providers, became major buyers of private credit, attracted by the potential for higher returns without the volatility of public markets. Yet, these are opaque, privately negotiated loans that are often valued by models rather than markets, and in some cases marked by the same managers who originated them. The absence of volatility is not the same thing as the absence of risk. A Columbia Business School study estimated that U.S. life insurers may need to hold as much as $4.5 billion more in regulatory capital each year to properly account for the risks in their private credit portfolios.

A significant share of the private credit now sitting on insurance balance sheets is financing the artificial intelligence infrastructure buildout. Morgan Stanley estimates that private credit will supply more than half the capital to close the $1.5 trillion external financing gap for data center construction through 2028. The scale of individual transactions is unprecedented: Meta completed a $30 billion off-balance-sheet private credit data center deal; Blue Owl and JPMorgan invested $13 billion through a special purpose vehicle (SPV) into a facility built for OpenAI; Oracle has constructed data centers through SPVs backed by tens of billions in debt. None of this debt is visible in the financial statements that policyholders rely on.

The particular concern with this exposure is that much of the AI sector's reported earnings growth relies on circular financial flows rather than independent customer demand. In the first quarter of 2026, nearly half of Alphabet's record $62.6 billion profit came from a paper markup on its Anthropic stake, and Amazon reported $16.8 billion of its $30.3 billion net income from the same source while its free cash flow fell 95%. OpenAI's cloud spending runs at more than double its reported revenue, with the gap covered by recycled investment rather than paying customers. If AI monetization disappoints, the borrowers behind these loans face refinancing conditions that may not support current marks, and those marks are already embedded in insurer balance sheets.

The early warning signs are real. Fitch reported in May 2026 that the U.S. private credit default rate hit a record 6.0%, payment-in-kind loans, where borrowers pay interest with more debt rather than cash, have risen from 5% to 11% of the market since 2022, and redemption requests at several major semi-liquid funds have significantly exceeded their stated quarterly limits. A $162 billion maturity wall arrived in 2026 just as capital appetite for rolling over weaker credits is declining. An insurer holding these assets against long-duration policyholder obligations cannot rely on the appearance of stability alone. What matters is whether the marks are real.

Bitcoin, by contrast, is volatile, and in its raw form unsuitable for many traditional insurance portfolios. However, when structured with conservative haircuts and appropriate overcollateralization, even significant drawdowns do not impair the collateral backing a reinsurance contract. A 50% haircut, for example, means that Bitcoin would need to lose more than half its value before the collateral falls below the obligation it secures. The contract remains denominated in dollars, claims are paid in dollars, and the overcollateralization absorbs the price movement. Beyond structure, Bitcoin's risk is visible. It trades continuously and by nature it is transparent. Its price is not determined quarterly by a manager. Its supply is fixed by protocol, and its ownership and transfer history can be verified on a public ledger. Bitcoin may be risky, but it certainly isn't opaque.

This distinction is crucial. The industry has traditionally viewed visible volatility as the primary source of danger, while underestimating the danger of assets that appear calm on the surface. This is especially relevant for reinsurance, where capital must be available when stress arrives. Insurers are facing rising demand for capacity across natural catastrophe, geopolitical, cyber, digital asset, and other emerging risks. That does not mean abandoning conservative capital standards, but rather evaluating new forms of collateral on their actual properties. On this basis, Bitcoin deserves to be looked at more seriously.

Used properly, Bitcoin-backed capital does not mean writing insurance in Bitcoin, paying claims in Bitcoin, or asking cedants to take crypto risk. Insurance and reinsurance can remain denominated in dollars, governed by conventional contracts, and subject to regulatory capital rules. The innovation is on the asset side: whether transparent, liquid, externally priced digital collateral can support real-world insurance obligations when structured with appropriate safeguards.

The insurance industry relies on the quality, liquidity, and transparency of the assets sitting behind the balance sheet. Now is the time for Bitcoin to be evaluated seriously as a potential source of transparent, externally priced insurance capital. It's very visible price volatility that makes it a perceived risk may prove its strength vs opaque / illiquid securities whose pricing is not determined by the market (until a liquidation is needed).

References: 

1. Antoine Gara, Eric Platt, and Oliver Barnes, "Mark Walter's insurers race to divest affiliated private credit assets," Financial Times, August 13, 2026. https://giftarticle.ft.com/giftarticle/actions/redeem/cf040ef2-3ac2-4ae7-a3c8-4754ec346294 

2. US Life Insurers Held $807 Billion of Highly Illiquid Credit," Bloomberg, June 8, 2026. https://www.bloomberg.com/news/articles/2026-06-08/us-life-insurers-held-807-billion-of-highly-illiquid-credit 

3. Christopher Marx, "Study Says Life Insurers May Need $4.5B More Capital for Private Credit Holdings," Institutional Investor, July 2026. https://www.institutionalinvestor.com/article/study-says-life-insurers-may-need-45b-more-capital-private-credit-holdings 

4. Moody's Ratings, "Private Credit Transforms Life Insurance Industry," June 2026. Reported in: Josh Recamara, "US life insurers' private credit push is creating liquidity and concentration risks, Moody's warns," Insurance Business, June 8, 2026. https://www.insurancebusinessmag.com/us/news/life-insurance/us-life-insurers-private-credit-push-is-creating-liquidity-and-concentration-risks-moodys-warns-578168.aspx 

5. Morgan Stanley Research, "Bridging a $1.5trn Data Center Financing Gap," July 2025. Discussed in: Vishy Tirupattur and Vishwas Patkar, "Credit Markets' Role in AI Financing Gap," Thoughts on the Market, Morgan Stanley, August 6, 2025. https://www.morganstanley.com/insights/podcasts/thoughts-on-the-market/credit-markets-ai-financing-gap-vishy-tirupattur-vishwas-patkar 

6. Alphabet Inc., Form 8-K, First Quarter 2026 Results, filed April 29, 2026, U.S. Securities and Exchange Commission. https://www.sec.gov/Archives/edgar/data/0001652044/000165204426000043/googexhibit991q12026.htm

7. Amazon.com Inc., "Amazon.com Announces First Quarter Results," press release, April 29, 2026. https://ir.aboutamazon.com/news-release/news-release-details/2026/Amazon-com-Announces-First-Quarter-Results/default.aspx 

8. Fitch Ratings, U.S. Private Credit Default Rate, trailing twelve months ended April 30, 2026. Reported in: Nabila Ahmed, "US Private Credit Defaults Hit New Record of 6% in April: Fitch," Bloomberg, May 18, 2026. https://news.bloombergtax.com/financial-accounting/us-private-credit-defaults-hit-new-record-of-6-in-april-fitch