The Accounting Is Not the Event
The losses already exist. What follows is the process by which they become visible.
D.T. FranklyPublished:
This analysis describes structural patterns and their likely trajectories. It is not investment advice, financial guidance, or market timing recommendation. No content should be interpreted as suggesting specific investment actions. Consult qualified financial professionals before making investment decisions.
What Is Already Gone
Start with arithmetic, not projection.
US commercial banks hold approximately $2.3–2.5 trillion in securities classified as Hold-to-Maturity — primarily Treasury bonds and agency mortgage-backed securities issued between 2020 and 2022 at near-zero interest rates. Those securities now trade at 55–75 cents on the dollar in a 5%+ rate environment. Unrealized losses on the HTM portion alone are estimated at $280–300 billion by the Federal Reserve’s H.8 data and the FDIC Quarterly Banking Profile. Combined with available-for-sale securities, system-wide unrealized losses run $450–500 billion.
These are not projected losses. They are present-tense arithmetic: current yield minus coupon rate, applied to face value, expressed as price. The losses exist on every balance sheet in the US banking system. They are not reported because Hold-to-Maturity accounting rules — a regulatory classification requiring only “intent and ability” to hold to maturity — permit a bank to carry these securities at cost. The market value is irrelevant to the balance sheet until a sale is forced. The regulatory framework was designed to prevent the appearance of loss, not the loss itself.
Private equity and private credit carry the same structure at larger scale. Approximately $1.1–1.3 trillion in life insurance general account assets are held in private credit vehicles — Apollo/Athene, Blackstone/F&G, and KKR/Global Atlantic are the largest concentrations. These are marked using Level 3 inputs: management-determined valuations, not market-observable prices. Private credit default rates reached 9.2% in 2025, the highest on record, per Fitch Ratings. Payment-in-kind arrangements — where borrowers defer cash interest in favor of additional debt obligations — more than doubled from 5% to 11% of the market over 2022 through end of 2025, per Lincoln International, which values approximately one-third of US private credit portfolios. PIK doubling is a standard indicator: the cash flows are no longer arriving, but the reported valuations have not moved to reflect their absence. Blue Owl Capital Corporation II (OBDC II) permanently halted quarterly redemptions in February 2026 and is proceeding with orderly liquidation. Multiple business development companies are enforcing redemption gates against request volumes running four to five times their quarterly caps. The gap between what these portfolios report and what they would clear at in a competitive market is the already-vaporized amount. It is not a forecast.
The basis trade adds a third dimension. Hedge funds hold an estimated $1.0–1.5 trillion in notional Treasury positions, levered 20-to-50-to-one, financed through the overnight repo market. At 50:1 leverage, the equity capital supporting these positions is approximately $20–30 billion against $1.0–1.5 trillion in notional exposure. A 2% adverse move in the underlying position eliminates the capital base. The Bank Term Funding Program — the backstop established after SVB for precisely this kind of stress — expired in March 2024 and has not been replaced. The value at risk is not the notional; it is the downstream market disruption when forced unwinding hits Treasury markets with insufficient active absorption capacity. That mechanism is covered in the transmission section below.
Across these three nodes, the losses already embedded in the system run into the hundreds of billions before the PE phantom valuation gap is fully assessed. None of this requires a forecast. It requires reading currently available data.
The real economy confirms the stress is running ahead of market recognition. The ISM Manufacturing Index registered 47.9 for the tenth consecutive month as of December 2025, with 85% of manufacturing GDP contracting — near Great Financial Crisis levels — while official GDP data shows approximately 4% growth, a divergence driven by concentrated AI capital expenditure whose productivity returns remain unvalidated. Hardship withdrawals reached a record 6% of 401(k) participants in 2025, triple the pre-pandemic rate of approximately 2%, per Vanguard’s How America Saves 2026 report, despite 4.4% unemployment, meaning behavioral deterioration in the savings population is already 12–18 months ahead of official employment statistics. Investment-grade credit spreads stand at 77 basis points as of May 12, 2026 — near 25-year lows — meaning credit markets are priced for near-perfection at the precise moment the documented stresses are accelerating. These are not early warning indicators. They are current-state readings confirming the vaporization is already in progress while market pricing has not yet moved to reflect it.
The Template: 1933
Understanding how this happens requires a template. The clearest one in American history is Executive Order 6102, signed April 5, 1933.
The United States in early 1933 was in existential fiscal and monetary crisis. The banking system was effectively insolvent: thousands of institutions had already failed, taking depositors’ savings with them. The remaining banks were experiencing runs. The dollar was under gold standard pressure the Treasury could not maintain. Federal deficits had no credible path to closure. The system required a repricing, and that repricing had to happen at the expense of domestic asset holders, because no external party was available to absorb the cost.
The mechanism was direct. EO 6102 required all US citizens, partnerships, and associations to surrender their gold coin, gold bullion, and gold certificates to the Federal Reserve by May 1, 1933, at the statutory price of $20.67 per troy ounce. Exemptions were narrow. Penalties for non-compliance were severe: up to $10,000 fine — approximately $240,000 in current terms — or ten years imprisonment.
Nine months later, the Gold Reserve Act of January 1934 set the new official gold price at $35 per troy ounce. The value of the compulsorily acquired gold increased 69% overnight. The federal government captured an immediate book-value windfall of $2.8 billion, used to capitalize the Exchange Stabilization Fund and stabilize the dollar’s new external value. The citizens who surrendered gold at $20.67 received paper dollars that were, by statutory definition, worth 59 cents of their previous gold value. The transfer was complete. It was legal. It was mechanically enforced by criminal penalty. The Supreme Court, in the Gold Clause Cases of 1935, upheld it 5–4, ruling the government’s need to manage monetary policy outweighed contractual obligations to creditors — a legal architecture Edwards (2019) documents in precise detail.
Four structural elements define the 1933 mechanism: (1) government cornered by debt, insolvency, and systemic instability; (2) legal authority invoked to channel assets into a specific vehicle at an administered price; (3) asset repriced against the holder after acquisition; (4) no practical exit.
Every element is present in the current system. The mechanism is slower, more diffuse, and operates through regulatory architecture and accounting conventions rather than criminal statute. The structure is identical.
The Architecture of Directed Savings
The Employee Retirement Income Security Act of 1974 constructed the framework through which the modern repricing operates. The mechanism differs from EO 6102 in speed and visibility, not in kind.
ERISA created the legal architecture for 401(k)s and IRAs. Over the following decades, the defined-benefit pension — a collective structure in which longevity risk was pooled and professional fiduciaries managed assets against defined income obligations — was replaced by individualized defined-contribution accounts. Workers who had held effective rights to a predictable income stream in retirement lost those rights through corporate restructuring and plan terminations. In exchange, they received an account and a menu of approved investment vehicles.
The entry compulsion: employer matching contributions create an immediate economic penalty for non-participation. The exit barrier: early withdrawal carries a 10% federal penalty plus ordinary income tax, making exit expensive enough to function as structural lock-in. The menu restriction: assets are limited to plan-administrator-selected vehicles, overwhelmingly public equity index funds and target-date funds. The fee extraction: intermediaries collect management fees, record-keeping fees, and fund expense ratios at every layer, whether the system performs or not.
Coimbra, Gomes, Michaelides, and Shen, writing in the Journal of Finance in October 2025, quantified what this architecture did to asset prices in aggregate. In the defined-benefit era, pension fund sponsors injected capital when markets fell, creating structural support for the equity premium. Cost of equity ran approximately 7%. In the defined-contribution steady state, individual savers don’t inject capital when markets fall — they withdraw. The equity premium compressed. Cost of equity fell to 1.5–3%. Equity valuations expanded not because underlying corporate value increased but because the architecture of mandatory savings flows inflated prices. Simultaneously, government financing costs more than doubled, because the compression of the equity premium required the risk-free rate to rise to clear markets.
The result is mechanical: the savings of an entire generation were deployed to inflate equity valuations beyond fundamental support. The workers who contributed to create those prices now hold assets that cannot be liquidated at scale without collapsing the price level their contributions were used to construct. The enclosure is self-reinforcing: exit pressure from any significant fraction of the savings base is itself a trigger for the vaporization they are trying to exit.
The passive index vehicle through which most of this capital flows compounds the trap. Market-cap weighting directs capital disproportionately to the largest, highest-priced assets — reinforcing existing concentration rather than distributing across fundamental value. As passive share grows, the active capital that would otherwise identify mispricing and correct it is crowded out. The Green-Krishnan-Sturm SSRN preprint (not yet peer-reviewed) formalizes this: passive dominance attenuates the mean-reversion force in equity markets — the mechanism by which prices return to fundamental value after disruption. Passive share of US mutual fund and ETF assets reached approximately 54% as of October 2025 ($19.1T passive vs $16.2T active), growing at approximately 4 percentage points per year per ICI data. The active bid is being systematically eliminated from the markets where the vaporized values will ultimately have to clear.
The Shadow Layer
Shadow banking, for this purpose: credit intermediation occurring outside the regulatory perimeter of daily mark-to-market pricing and standardized capital requirements. This is where the deferred recognition lives.
HTM accounting is a regulatory permission to carry securities at a price that is not their current price. This is not a technical limitation — market prices for these securities are observable every trading day. The decision to permit amortized cost accounting for HTM is a policy choice made by regulators who understood that requiring mark-to-market on the entire HTM portfolio would immediately reveal insolvency at multiple institutions. Basel III’s Endgame rules are now eliminating the AOCI opt-out for larger regional banks — forcing available-for-sale unrealized losses into CET1 capital calculations — while preserving the HTM carve-out. The incentive structure this creates pushes banks to migrate securities from AFS to HTM: this locks up more balance sheet liquidity but protects reported capital ratios. Every migration widens the gap between reported and real value.
Private credit and PE marks. The Apollo/Athene, Blackstone/F&G, and KKR/Global Atlantic structures have concentrated over $600 billion in combined assets inside life insurance general accounts, with heavy allocation to private credit originated by affiliated PE managers. The portfolios are marked using Level 3 inputs. The NAIC, in regulatory initiatives beginning in 2025, has attempted to address the arbitrage that allowed equity-like risk to carry investment-grade capital treatment — reclassifying instruments from Schedule D (bond treatment, lower risk-based capital charge) to Schedule BA (other assets, higher charge) where they lack genuine debt characteristics. This addresses the edges. The core valuation methodology — self-administered, quarterly, not market-observable — remains intact. PIK doubling and default rates above 9% are the market’s reading of the gap. The reported marks are not.
Commercial real estate adds a third observable stress channel already visible in the shadow layer. CMBS delinquency rates reached 7.28% as of Q1 2026 per MBA, with office properties hitting 12.34% in January 2026 per Trepp — well into crisis territory. Regulated bank CRE exposure shows only 1.24% delinquency per the same MBA Q1 2026 report because banks extend rather than foreclose, preserving the loan as “performing” on the balance sheet while the underlying collateral has declined 20–30% in value. The dual reading — office CMBS above 12%, regulated banks at 1.24% — is the same shadow layer structure as HTM and PE marks: observable stress in the segment that must price to market, masked stress in the segment that controls its own accounting. When forbearance capacity exhausts, forced sales in bank-held CRE clear at prices already established by the CMBS market.
The basis trade. This is the least visible node and the most immediately dangerous catalyst. Hedge funds financing Treasury arbitrage at 20–50:1 leverage through bilateral repo agreements operate in a space where no real-time aggregate exposure reporting exists publicly. The CFTC’s leveraged trader short positions in Treasury futures are a proxy, not a direct measurement. The Federal Reserve’s Financial Stability Reports describe the systemic risk but cannot fully observe the bilateral repo financing supporting these positions. The BIS Quarterly Review has flagged the trade in multiple editions as a structural vulnerability. The Office of Financial Research Short-Term Funding Monitor tracks repo volumes but not the identity or leverage of participants behind the flows. The SEC’s Treasury clearing mandate, phased in through 2025–2026, increases transparency for centrally cleared transactions. Bilateral repo — where the most extreme leverage concentrations sit, at near-zero haircuts — remains largely unobservable by regulators in real time.
The shadow layer is not a metaphor. It is a set of specific accounting categories and regulatory perimeters inside which losses that already exist are not required to be stated as losses.
The Transmission
The mechanism by which localized value vaporization becomes systemic is the flow of liquidity demand through passive-dominated markets with insufficient active absorption.
A recognition event at any single node — a PE gate cascade forcing pension funds to sell public equities to meet cash obligations; a repo market seizure forcing simultaneous basis trade deleveraging; a bank compelled to sell HTM securities, triggering the GAAP reclassification rule that immediately forces the entire HTM portfolio onto the balance sheet at market value — creates a sudden large selling flow in markets where the correction mechanism has been atrophied.
Passive funds do not add to the active bid during a sell event. They provide price-insensitive buying proportional to existing index weights during inflows, and symmetric price-insensitive selling during outflows. The Haddad-Huebner-Loualiche study in the American Economic Review (2025) demonstrates that active investors currently offset only 50–60% of passive flows, and that this offset capacity declines as passive share rises. The selling hits markets where 40–50% of the marginal flow has no fundamental price anchor.
The equilibrium-restoration counter-argument — Grossman and Stiglitz (AER, 1980) — holds that active capital scales up as passive distortions grow, restoring equilibrium. The Haddad study directly tests this at current passive scale and finds it does not hold: active offset is incomplete and declining. The Grossman-Stiglitz mechanism requires sufficient active capital willing to take contrarian positions; that capital is being structurally competed away.
The ERISA concentration compounds this. Because savings are channeled into a small number of vehicles by the architecture described above, a market-wide stress event hits every participant simultaneously. The concentration that was operationally convenient for plan administrators is, under stress, the mechanism by which a local recognition event becomes a systemic one. Every 401(k) holder is in the same vehicles. Every exit request arrives at the same market at the same time.
April 2020 was the test run. A global flight to cash produced repo market stress severe enough to cause Treasury market dysfunction — the benchmark safe asset became temporarily unable to trade at any price. The Federal Reserve intervened with emergency repo operations over several weeks. The Bank Term Funding Program was established as a permanent backstop post-SVB. That program expired March 2024. The next repo stress event — which, at current basis trade leverage and in the absence of any backstop, can be triggered by any sufficiently large credit stress event — meets a system with none of the emergency absorption capacity that contained the 2020 and 2023 episodes.
The XIV collapse in February 2018 provides the mechanical analogy in miniature. XIV was an inverse VIX ETF — a vehicle whose structure meant it had to buy volatility when volatility spiked, mechanically amplifying the very move it was supposed to profit from in normal conditions. A single-day spike in VIX triggered a forced mechanical unwind that destroyed nearly 100% of the fund’s value in one session. The passive equity market — currently at 53–54% and approaching the ~65% instability threshold Green-Krishnan-Sturm identify as the point of sharply increasing volatility — is already manifesting this reflexive structure: an architecture that amplifies the move in the direction of the shock, with insufficient and declining offsetting capacity.
The Historical Pattern
This is not novel. It is the documented and replicated response of governments with high debt-to-GDP and monetary sovereignty when the carrying cost of that debt becomes fiscally unsustainable.
Reinhart and Sbrancia’s 2015 paper in Economic Policy — and the earlier 2011 paper in Finance and Development — documents the mechanism across the post-WWII US and UK cases with precision. Between 1945 and approximately 1975, the United States ran real interest rates below zero approximately half the time. The mechanism: Regulation Q placed explicit ceilings on deposit interest rates; mandatory Treasury bond holdings for banks and insurance companies created a captive domestic audience; capital controls under Bretton Woods prevented domestic savers from accessing higher-yielding foreign assets. Combined with post-war inflation, the result was a systematic transfer from domestic creditors to the sovereign at approximately 3.2% of GDP per year — not through explicit default, not through austerity, but through the engineering of a negative real rate environment that held nominal rates below inflation. US debt-to-GDP fell from 121% in 1946 to approximately 30% by the mid-1970s. Real economic growth contributed. Financial repression did most of the structural work. The UK ran a more aggressive version: approximately 3.6% of GDP per year in implicit wealth transfer from savers to the sovereign over the same period.
This was not accidental. It was the chosen mechanism for managing a debt overhang that could not be addressed through growth or austerity alone. It operated through the architecture of the financial system — rate ceilings, mandatory purchase requirements, capital controls — rather than through legislative appropriation. The transfer was real and large. It was not publicly announced as a transfer.
EO 6102 was the instantaneous, explicit version. Post-WWII financial repression was the slow, distributed version. Both are responses to the same structural condition: a sovereign carrying more debt than it can service at market rates, with monetary sovereignty, choosing to use that sovereignty to transfer the carrying cost onto domestic savers rather than default on external creditors or impose politically legible tax increases.
Current US debt-to-GDP exceeds 120%. Annual interest cost has crossed $1 trillion. The conditions that generate financial repression as a sovereign survival response are present. Itskhoki’s 2025 working paper on financial repression under sanctions demonstrates that this toolkit remains actively deployed globally; it did not expire with Bretton Woods. The policy choices that constitute its modern form — maintaining HTM accounting treatment, preserving PE mark-to-model conventions, sustaining the ERISA concentration architecture, not requiring central clearing of the basis trade — are being made continuously by regulators who understand their consequences. The question of intent is analytically irrelevant. The structure produces the transfer regardless of the intent behind maintaining it.
The Policy Vacancy
The mechanism requires a policy vacancy to operate at scale. That vacancy is not passive neglect. It is the active maintenance of conditions under which recognition is deferred.
HTM accounting treatment is a regulatory choice. The Federal Reserve, FDIC, and OCC could require mark-to-market on HTM portfolios. They have not done so because doing so would immediately reveal $280–300 billion in losses concentrated in regional and mid-sized institutions, several of which would require recapitalization or resolution. The choice not to require it is a choice to defer the recognition. It is not passive. The S&L crisis of the 1980s provides the direct precedent: regulatory forbearance allowed insolvent savings and loans to continue operating for years, accumulating additional losses, until the resolution cost had grown to approximately $160 billion. The mechanism is structurally identical to the current HTM situation. The lesson was absorbed: explicit forbearance was recognized as a policy failure. The modern version operates through accounting classification rather than explicit forbearance, but the functional consequence is the same.
PE mark-to-model is permitted under GAAP and statutory accounting rules that Congress has not revised in response to the growth of private credit from a niche asset class to a $1+ trillion exposure inside life insurance general accounts. The NAIC’s 2025 regulatory initiative is a partial correction proceeding slowly against significant industry opposition, targeting only the most egregious regulatory arbitrage without addressing valuation methodology.
The basis trade operates in bilateral repo markets carrying no mandatory real-time reporting of positions, leverage, or counterparty exposure at the individual fund level. The SEC’s Treasury clearing mandate, phased in through 2025–2026, increases transparency for centrally cleared transactions. Bilateral repo — where the most extreme leverage concentrations exist — remains largely opaque to regulators. No legislation has mandated its comprehensive reporting.
ERISA’s 10% early withdrawal penalty and investment menu restrictions have not been revised since the DC transition was substantially complete, despite the systemic consequences of the savings concentration they enforce.
None of these regulatory failures is partisan. ERISA was constructed under Ford; the HTM carve-out, PE mark-to-model conventions, and basis trade opacity have been actively maintained or passively allowed to persist across every subsequent administration of both parties. The pattern is institutional and structural. It is not a political failure of one party — it is the sustained preference of the financial system across administrations.
Each of these is a specific regulatory decision that has not been made. None requires malice to explain. All that is required is that the cost of not making the decision falls on savers rather than on the institutions and government entities that benefit from continued deferral. That condition is met in every case. The fee structures of intermediaries, the capital positions of banks, the reported solvency of insurance companies, and the reported debt-service capacity of the sovereign all depend on the recognition being deferred. The incentive to maintain the policy vacancy is structural and continuous.
What This Is
The structure is this: savings are directed by law and tax architecture into a set of vehicles where the repricing has already occurred, the losses already exist, and the accounting confirms them last. The government and regulatory infrastructure that built and sustains this architecture has maintained it because recognition is worse for system stability than for the savers carrying the exposure. This is the same structure as EO 6102, the same structure as post-WWII financial repression, and it produces the same result: a transfer from domestic savers to the sovereign and financial intermediary system, executed through the architecture of the financial system rather than through legislative appropriation.
The transmission mechanism — liquidity demand flowing through passive-dominated markets with an atrophied active bid — ensures that when recognition begins at any node, it propagates through the entire enclosed savings population simultaneously, because the ERISA architecture has placed everyone in the same vehicles with the same exit constraints.
The vaporization is not a future event. It is a current condition. The accounting is the last thing that changes.
— Free to share, translate, use with attribution: D.T. Frankly (dtfrankly.com)
§