Deferred Reconciliation: The System, Then and Now
D.T. FranklyPublished:
Two Tracks
The economy looks contradictory from ground level. Billion-dollar industrial construction sites are operating at full capacity across states that have not seen this kind of capital investment in decades. Semiconductor fabs, pharmaceutical plants, and liquefied natural gas terminals are rising simultaneously with headlines about structural debt stress, commercial real estate collapse, and consumer credit strain. Both observations are accurate. They coexist because they are funded by entirely different mechanisms operating at cross-purposes.
The construction activity is primarily driven by legislated government spending: the CHIPS and Science Act, the Inflation Reduction Act, and the Infrastructure Investment and Jobs Act allocated hundreds of billions into specific industrial categories regardless of the monetary environment. This capital is long-duration and largely locked in. Semiconductor fabs, pharmaceutical manufacturing capacity, and energy infrastructure represent genuine productive additions with asset lives measured in decades. These are the closest modern analog to the New Deal’s Tennessee Valley Authority and Hoover Dam: government investment creating durable industrial capacity against a backdrop of financial stress.
The AI compute buildout is a different category with different physics. Data centers and GPU clusters being deployed by the large technology companies convert liquid balance sheets into specialized hardware with three-to-five-year obsolescence cycles. This is not the TVA analog. The capital is private, the assets depreciate at compressed rates, and the forward accounting obligations this creates are specific and calendar-driven in ways the durable industrial investment is not. The distinction between these two types of capital matters for understanding what the construction boom means and what it does not mean.
The fiscal-monetary bifurcation itself is not new. The same dynamic operated in the 1930s: New Deal investment created visible construction activity while the banking system was in structural collapse and the Federal Reserve tightened into a deflationary spiral. Both tracks existed simultaneously because they were driven by separate mechanisms with separate logics. What appears paradoxical from street level is the predictable output of two policy instruments pointed in opposite directions.
This piece is about the mechanism underneath both tracks.
How Financial Systems Decouple from Reality
A financial system breaks not when something goes wrong but when locally rational behavior produces collectively destructive outcomes. Every actor responds correctly to available signals. The aggregate of those correct responses makes the signals worse. This is the closed loop.
Eight structural mechanics produce this state. They are not unique to any era, technology, or ideology. They emerge from the architecture of credit-based economies and repeat in recognizable form whenever debt, leverage, and price signals interact under stress.
Debt and collateral feed each other. When an asset serves as collateral for a debt, its price and the debt’s serviceability are coupled. Price falls impair collateral, which forces selling to cover the debt, which forces prices lower, which impairs more collateral. Each actor selling to protect themselves accelerates the condition that required selling. Deleveraging individually produces the environment that demands more deleveraging collectively.
Liquidity freezes at a threshold. Below a certain level of confidence, holders of liquid assets refuse to deploy them at any price. Interest rates fall toward zero; money does not move. The hoarding is self-sustaining because the contraction it causes is the same contraction that justifies the hoarding. The system has two stable states: circulating and frozen. Once frozen, price signals alone cannot unfreeze it.
Administered prices replace discovered prices. When prices are set outside the mechanism that would normally clear supply and demand, actors make decisions based on prices that do not reflect underlying reality. This can come from direct government controls, from accounting conventions, or from institutional mandates. The critical point is that the actors believe they are responding to real signals when they are responding to artifacts of the administrative framework. Policy calibrated to these indicators then responds to the administration, not the condition underneath it.
Deflation transfers wealth from debtors to creditors and concentrates it. Fixed nominal claims become more burdensome as the price level falls. A debt of $100 at $1 per unit of output requires 100 units to service. At $0.75 per unit, it requires 133. Deflation systematically transfers resources from those who borrow and spend to those who lend and save, concentrating capital in hands that have already met their consumption needs. The transfer accelerates the contraction.
Individually correct responses compound into collective destruction. Each policy authority sees its own indicator and responds to that indicator correctly in isolation. No single actor is wrong. There is no integrator that sees the compound effect before it lands. Multiple tightening forces operating simultaneously, each calibrated to different indicators, can produce catastrophe that none of the individual decisions would have caused alone.
Asset prices decouple from income when discount rates are forced. Asset prices reflect discounted future cash flows. When the discount rate is artificially suppressed, prices rise without any improvement in the actual cash flows the assets will generate. The spread between asset price recovery and real economic activity (employment, output, wages) is the measurable decoupling. The prices imply a future the real economy cannot deliver.
Institutions optimize for survival over function. The institution whose function is credit intermediation stops intermediating credit and accumulates reserves. This is rational for the institution. It is a system-level failure. The institution persists; the function it was built to perform disappears. The failure is invisible to any metric that measures institutional health rather than institutional function.
Measurement is changed to avoid showing the gap. When the distance between stated value and real value becomes uncomfortable, the measurement is adjusted. This is not necessarily corrupt. It can be accounting policy, regulatory forbearance, or statutory framework. The effect is identical: the gap between stated position and real position is absorbed into the convention rather than forced to reconcile through the market. The loss exists but does not appear.
The 1930s: Each Mechanic Made Visible
The Great Depression is the canonical case because the system was simpler, the velocity was faster, and the mechanics were exposed without the modern suppression architecture that obscures equivalent dynamics today. Each of the eight elements above was present, measurable, and in several cases described in real time by contemporaries watching it happen.
The pre-crash circuit (1927-1929) was a pure version of administered prices replacing discovered prices. Investment trusts, the era’s version of pooled investment vehicles, were organized in pyramidal, cross-owned structures: Trust A held shares in Trust B, which held shares in Trust A or its subsidiaries. Circular ownership inflated reported asset values and earnings without any increase in industrial output or consumer demand. By late 1929, trailing price-to-earnings ratios for market leaders exceeded 30x while industrial production had been flat or declining since June 1929. Call loans to brokers surpassed $8.5 billion, consuming a disproportionate share of total bank credit. The velocity of financial transactions had fully decoupled from the velocity of real-economy transactions. Capital formation stopped funding productive capacity and began funding the acquisition of existing financial instruments.
The post-crash circuit (1930-1933) demonstrated the debt-collateral mechanic at full scale. Irving Fisher formalized the observation in 1933 while it was still happening: liquidation contracted deposit currency, contracted deposit currency fell the price level, the falling price level increased the real burden of remaining debt, and the increased real burden required more liquidation. The system consumed itself through individually rational responses. Wholesale prices fell 33%. The real cost of servicing fixed nominal debt rose by roughly 50 percent. Viable enterprises faced insolvency from liquidity starvation rather than operational failure.
The monetary base, controlled by the Federal Reserve, actually increased or held flat across this period. M2 money supply collapsed by over 30%. The gap between these two measures is the broken multiplier: banks accumulated excess reserves rather than lending them into circulation as deposits. Those excess reserves grew from roughly $500 million in 1933 to over $3 billion by 1936 while lending stagnated and unemployment remained above 14 percent. The banks were performing Element Seven precisely: optimizing for institutional survival by holding reserves against the deflation their collective behavior was perpetuating.
While nominal interest rates approached zero, price deflation reached 10 percent annually. The real interest rate spiked toward positive 10 percent. The price mechanism had inverted: borrowing cheaply in nominal terms was devastatingly expensive in real terms. The Federal Reserve’s signal (near-zero nominal rates) pointed toward expansion. The real condition pointed toward contraction. Policy was reading the administered signal.
The 1937 contraction is the cleanest controlled demonstration. The surface indicators looked like recovery by 1936: unemployment had fallen from 25 to 14 percent, the stock market had rallied substantially, commodity prices appeared to be recovering. Three contractionary forces applied simultaneously, each justified by those surface indicators. The Federal Reserve doubled reserve requirements three times between August 1936 and May 1937. The Treasury sterilized gold inflows, removing approximately $1.3 billion from circulation. The Roosevelt administration cut fiscal spending to approach budget balance. Each decision was defensible in isolation. The compound effect: industrial production fell 30 percent, GDP contracted roughly 11 percent, unemployment jumped from 14 back toward 19 percent within 12 months. The economy was responding to the administration of the indicators, not the underlying condition. When the administration tightened, the underlying condition was exposed.
The New Deal price controls demonstrated what happens when administered prices replace discovered prices across entire sectors. The National Recovery Administration created cartel-like price floors through industry codes covering over 500 sectors. Minimum prices were set administratively, output was constrained, and labor costs were coded. Prices no longer reflected supply and demand; they reflected political negotiation. Actors made decisions based on those prices, producing outcomes that diverged from what normal price clearing would have produced. The NRA was struck down by the Supreme Court in 1935 partly because the empirical record showed prices rising while output and employment did not follow, which no standard price theory could explain within a functioning market.
The Agricultural Adjustment Act paid farmers to destroy crops and slaughter livestock while hunger was documented across the country. This was the most visible expression of the system consuming itself: destroying supply to raise prices to preserve farm income to maintain demand, while the underlying need went unmet. The price system and the welfare system had disconnected. Both were operating. Neither was clearing the actual condition.
The stock market recovery from 1932 to 1937 (roughly 4.7x from the trough) against a backdrop of 14 to 25 percent unemployment is the asset-income decoupling made explicit. Prices implied a future the real economy was not delivering. The discount rate had been forced down by policy and monetary conditions; prices reflected that forcing, not the underlying cash flow trajectory of the economy beneath them.
Today: The Same System, Different Instruments
Every element from the 1930s is operating now. The instruments are different. The velocity is suppressed by a modern architecture specifically designed to prevent rapid crystallization. The mechanics are identical.
Shadow credit is the modern M2. In the 1930s, the broken multiplier appeared as a gap between the monetary base and M2: the Fed held reserves, banks sat on them rather than lending them into deposits. Today, the Federal Reserve’s balance sheet expanded from roughly $900 billion in 2008 to approximately $9 trillion at its peak. M2 expanded with it. But a third aggregate exists that standard measures do not capture: the credit generated by the non-bank financial system through private equity, private credit funds, non-traded real estate vehicles, and collateralized loan obligations. Call it shadow M. This aggregate expanded enormously during the period of zero interest rates as cheap debt funded buyouts, private asset accumulation, and yield-seeking allocations from institutions whose guaranteed return obligations exceeded what public markets could provide. Shadow M is now contracting. Private equity distributions are at their lowest levels since the global financial crisis. Secondary market clearing for LP stakes averages approximately 10 to 15 percent below stated net asset value across strategies. The contraction does not appear in M2 the way bank failures appeared in the 1930s data. The symptoms appear instead as gating events, denominator distortions in institutional portfolios, and a slowdown in PE fundraising. The contraction is uneven: direct lending and private credit origination has expanded as banks tightened underwriting standards, relocating illiquid risk into new structures with the same maturity mismatch properties rather than eliminating it.
Redemption gating is the modern bank run. A 1930s bank run was a maturity mismatch: depositors held demand claims against banks holding illiquid loans. When enough depositors ran simultaneously, the bank had to liquidate illiquid assets at fire-sale prices. The loss crystallized through the run itself. Open-end private credit funds, non-traded REITs, and private equity vehicles with periodic redemption windows are structurally identical. Investors hold liquid-adjacent claims against funds holding assets with no liquid market. When redemption pressure exceeds the fund’s liquidity buffer, gates activate. The difference is timing: a bank run completes in days. An institutional fund gate slows the same run to quarters or years. The maturity mismatch does not resolve; it accumulates behind the gate.
Mark-to-model is the modern NRA price code. Private assets are valued through internal models at discount rates and comparable transactions selected by the fund manager. When a public index falls 20 percent, a private equity fund holding economically equivalent assets marks its portfolio down 3 to 5 percent, citing “operational intervention.” The mark is an administrative price, not a discovered one. Institutions holding these assets make allocation decisions, report funding ratios, and set contribution requirements based on prices that reflect the administrative framework rather than clearing prices. Policy calibrated to these marks then responds to the administration. The parallel to the 1930s price codes is structural, not rhetorical.
HTM accounting locks trillions in place. During the zero-rate period, commercial banks bought long-duration Treasuries and mortgage-backed securities at peak prices. When rates rose, the market value of these bonds fell substantially. Under accounting regulations, banks moved these assets to held-to-maturity portfolios, removing the requirement to report unrealized losses. The constraint: selling a single bond from an HTM portfolio forces marking the entire portfolio to market, which would trigger insolvency for institutions where the unrealized losses exceed tier-one capital. The result is trillions of dollars locked onto bank balance sheets, earning below-market yields, contributing nothing to credit creation, and unable to be released without triggering the loss recognition they were classified to avoid. This is Element Eight as accounting: measurement changed to avoid showing the gap. The Reconstruction Finance Corporation performed the same function in 1933, explicitly permitting banks to carry underwater agricultural and real estate loans at book value to prevent the loss recognition that would collapse the system. The 1930s deferred reconciliation was state policy; the modern version is accounting classification.
NAV loans are the modern investment trust pyramid. To pay the 7 to 8 percent distributions that pension fund mandates require, private equity managers borrow against the collective net asset value of their portfolio companies and distribute the proceeds to limited partners. The pension receives cash that it treats as investment return. The debt sits on the underlying operating companies, which are already running on compressed margins at elevated leverage. The distribution masks the fact that the underlying businesses cannot generate sufficient organic cash to distribute. Trust A is borrowing against Trust B to pay Trust A’s investors, and Trust B is borrowing against its operating companies to pay Trust B’s investors. The circular leverage structure is the same. The instruments are different.
What the Institutions Do When They Need Cash
Pensions, life insurers, and endowments owe nominal cash. A monthly retirement payment is a specific dollar amount, not a proportional share of a private equity fund. When the system they hold cannot generate that cash, they follow a predictable liquidation sequence.
The first buffer is incoming contributions: employee and employer payroll deductions flowing into the fund each cycle. When those contributions cover outgoing benefit payments, nothing moves in markets. When they fall short, as they increasingly do in mature public plans with growing beneficiary populations relative to active workers, the institution sells.
Treasuries move first because they trade with near-zero market impact at institutional scale. Investment-grade corporate bonds follow. Public equities come next, and here the denominator effect activates: because private asset marks are stale and have not fallen proportionally with public assets during a market correction, the private allocation appears to have grown as a percentage of the total portfolio. Investment policy mandates require rebalancing toward targets. The institution is therefore required by its own governance framework to sell the assets that have fallen in order to maintain exposure to the assets that are pretending they have not. This is not irrational at the institutional level; it is mechanically required. Collectively, it transmits the hidden stress of the shadow banking system directly into the pricing of public markets.
Capital calls continue regardless. When a pension commits $500 million to a private equity fund, the fund manager draws on that commitment when deals are available, not when the pension has liquidity to spare. Unfunded commitments across a large public pension’s alternatives program can equal 15 to 20 percent of total assets, and they are legal receivables, not optional obligations. During the current period, PE distributions to LPs have slowed dramatically while capital calls have continued. The pension is simultaneously paying beneficiaries and funding capital calls, receiving less from PE than at any point in the past decade, and covering the difference by selling public securities.
When public market liquidation is insufficient, institutions turn to the secondary market for private assets. LP stakes in PE and venture funds trade at discounts to stated NAV. At the 2022–2023 stress trough, buyout fund secondaries cleared at roughly 80 to 85 percent of NAV and venture capital at 60 to 65 percent. Per the Jefferies Global Secondary Market Review for H1 2025, pricing has since partially recovered: buyout secondaries at approximately 94 percent of NAV, venture at approximately 78 percent, with a transaction-weighted average discount of approximately 13 percent across strategies. The institution accepts that haircut to obtain immediate liquidity. The official book valuation does not update. The realized loss occurs, and the gap between model value and clearing price accumulates with each transaction without appearing in the reported marks of the remaining portfolio.
The final layer is financial engineering through NAV loans. The logic above: borrow against the fund to manufacture a distribution, distribute the borrowed cash, maintain the mark, service the debt from the underlying companies. The pension uses the distribution to pay its beneficiaries. The debt is now embedded in the operating companies. Each iteration of this cycle reduces the real collateral supporting the model mark while the model mark does not move.
Life insurance companies carry this same dynamic with an additional forced-selling mechanism. Variable annuity writers with guaranteed living withdrawal benefits (contracts promising 6 to 7 percent income floors regardless of portfolio performance, written in large volumes between 2000 and 2012) hedge those guarantees dynamically. When equity markets fall, their hedge requires selling equity futures or purchasing puts, which amplifies the downward pressure on equity prices at exactly the moment when other institutions are also liquidating public equities. The mechanism is procyclical and non-discretionary. Several large insurers have already sold these legacy liability blocks to private equity-backed reinsurers, transferring the unreconcilable obligations into the same shadow structure where pension assets are parked. The problem has changed hands, not resolved.
Where the Burden Goes
Four vectors absorb the gap between what these institutions have and what they owe, and they operate simultaneously rather than sequentially.
Taxpayers bear the public pension shortfall directly. When a public pension’s funding ratio falls, the actuarially determined employer contribution increases. That contribution is a mandatory line item in the municipal or state budget. To pay it, the government diverts funds from discretionary spending. Infrastructure maintenance defers. Public services reduce. As services degrade, the productive tax base (businesses and high-income households with geographic mobility) migrates, eroding the revenue that must now cover both services and a larger pension obligation. The spiral is self-compounding. Municipal fiscal stress transmits to state governments, which face their own pension obligations and cannot legally declare bankruptcy under current federal law. The federal government absorbs as a last resort through infrastructure grants, emergency appropriations, or central bank municipal facilities.
Beneficiaries absorb through prospective modification rather than nominal reduction. Direct benefit cuts are legally and politically difficult. Instead, cost-of-living adjustments are capped below realized inflation, retirement ages for current workers are extended, contribution requirements for active employees increase, and benefit formulas for new employees are restructured. The nominal contract is maintained; its real value is eroded over the remaining working and retirement years of current participants.
Inflation erodes real obligations without formal recognition. Fixed nominal pension liabilities become less burdensome in real terms as the price level rises. This is an unacknowledged debasement of the pension contract: the government or sponsor does not formally reduce the obligation; inflation reduces it without requiring any institutional decision. The beneficiary receives the nominal amount promised and purchases less with it each year.
The federal balance sheet is the terminal absorber. Fed liquidity facilities (municipal liquidity lines, emergency swap arrangements) can swap illiquid assets for cash to prevent a freeze. They cannot rewrite the underlying cash flow deficit of a pension that has promised 7.5 percent returns and owns assets that will not deliver 7.5 percent. The gap climbs from mid-market corporate loan to private equity portfolio company to pension fund to municipal budget to state budget to federal intervention, at each stage requiring the level above to absorb what the level below cannot resolve. The sovereign must eventually backstop what the private system cannot absorb, at a moment when the sovereign’s own balance sheet is operating at debt-to-GDP ratios that constrain the traditional tools of fiscal expansion.
Three Elements Without 1930s Precedent
Three structural features of today’s system have no direct historical parallel. They do not change the destination; they change the path, velocity, and appearance of how the system arrives there.
The automated passive bid. Passive indexation accounts for over 55 percent of US equity fund assets as of year-end 2025 (Morningstar), drawing the dominant share of equity inflows in every year since 2014. Every two weeks, automated 401(k) contributions enter the market and are required to purchase the largest capitalization stocks in the index, completely independent of valuation, macro stress, or underlying cash flow. This creates a permanent structural bid concentrated at the top of the market capitalization pyramid. As the shadow banking system freezes and mid-market companies face credit pressure, liquidity collapses inward toward the core. Capital fleeing private assets and regional banking stress flows into the largest, most liquid public companies. Because those few companies weight the major indices heavily, the index rises or holds while the broader economic fabric decays beneath it. Public indices are the last thing to fall. They will fall when the forced selling from non-discretionary obligations (pension payouts, required minimum distributions, capital calls) exceeds the incoming automated passive flows. Until that crossover, the index functions as a lagging illusion sustained by mechanical inflows.
The silica trap. When the large technology companies deploying tens of billions annually in AI infrastructure spend that capital on specialized computing hardware and data centers, the financial physics are unlike any prior era’s capital expenditure. A railroad built during the 1930s held structural value for decades with minimal maintenance until demand returned. GPU-based AI infrastructure becomes technologically obsolete within three to five years regardless of economic conditions. The depreciation is non-discretionary and front-loaded. These companies are converting liquid free cash flow into rapidly depreciating physical assets on a compressed cycle. When enterprise demand contracts in a recession, the fixed costs of operating these facilities do not flex with revenue. The depreciation charges arrive on schedule regardless of utilization. By late 2027, the capital deployments of 2024, 2025, and 2026 will be generating peak depreciation run rates simultaneously. If enterprise AI adoption has not produced sufficient revenue to absorb those charges, the income statement impact will be un-hideable. Free cash flow, already suppressed by ongoing capital commitments, will turn negative for those whose core businesses cannot subsidize the infrastructure buildout.
The sovereign ceiling. The Reconstruction Finance Corporation’s deferred reconciliation strategy in 1933 operated against a US federal debt-to-GDP ratio below 40 percent. The sovereign had enormous balance sheet headroom to issue debt, absorb toxic assets, and wait for nominal growth to close the gap between book values and reality. It required nearly two decades of financial repression, New Deal spending, and wartime industrial mobilization to complete, but the headroom existed. Today, the same terminal backstop must operate against a debt-to-GDP ratio above 120 percent. Annual federal interest expense is consuming a structurally significant share of tax revenues and competing directly with entitlement obligations. The Federal Reserve can create liquidity facilities to prevent a system freeze. It cannot create solvency. It cannot rewrite the pension’s actuarial gap. Expanding the central bank balance sheet to absorb private losses accelerates currency debasement, which erodes the real value of every fixed-income obligation including the sovereign debt itself. The traditional escape valve (print, inflate, and wait) operates under much tighter constraints than in any prior debt-deflation episode.
The Bond Market and the Velocity Sink
Capital under institutional stress does not hold in commercial bank accounts. Above the FDIC insurance threshold, counterparty risk is real. Institutional capital fleeing private assets and stressed equity positions moves into short-term US Treasuries.
Once it enters the sovereign bond market at scale, it stops circulating in the productive economy. It earns nominal interest, isolates from credit risk, and withdraws from the rolling credit renewals that leveraged operating companies require to survive. This is the modern version of 1930s hoarding: not physical currency under mattresses but electronic capital in sovereign debt, functionally equivalent in its extraction from productive circulation.
The bond market serves as a leading indicator not through traditional yield movements but through the velocity of capital within it. When institutional flows into short-term Treasuries increase while the broader credit system is contracting, the sovereign debt market absorbs liquidity that will not return to productive use until conditions change. Bank balance sheets are already operating a version of this: trillions in held-to-maturity Treasuries and mortgage-backed securities locked in place, earning below-market yields, contributing zero to credit creation, unavailable without triggering the loss recognition that would force institutional insolvency. The money went in. It will not come out until those bonds mature, measured in years to decades.
The Terminal Transition
The system does not reach its terminal transition through a single event. It arrives through the sequential failure of the mechanisms that have been suppressing velocity and deferring recognition.
Private funds maximize gates until the underlying companies can no longer service the debt piled on them through NAV loan cycles. The leverage limit is mathematical, not political. When it arrives, distributions stop entirely and capital call requests may cease, but the marks have not moved and the pension’s stated asset value still includes positions that have no path to cash at anything near book value.
The automated passive bid decelerates as layoffs affect the white-collar sectors that generate the bulk of 401(k) contributions. Job losses at technology companies and professional services reduce the bi-weekly inflows that have been sustaining index prices. As inflows slow, the concentrated index buffer becomes thinner and the forced selling from non-discretionary institutional obligations (pension payouts, required minimum distributions, capital calls funded by public market liquidation) starts to exceed the incoming bid. Indices correct not because sentiment shifts but because the mechanical flows cross over.
The corporate refinancing wall arrives as debt issued at near-zero rates during 2020 and 2021 matures or reprices. Mid-market companies moving from 3 percent debt to 8 percent debt on the same revenue base face a cash flow cliff that is immediate and arithmetic. Default rates in leveraged loans rise. When they rise sufficiently, overcollateralization tests in CLO structures fail automatically. CLO managers are then required by contract to sell loan positions to rebuild coverage ratios. This is non-discretionary, contractual, and has no gate mechanism. It is the most visible mechanical indicator of mid-market corporate stress because leveraged loan prices trade in a semi-transparent market with regular reporting. This is the canary: when CLO coverage tests begin failing at scale, the shadow M contraction has reached a stage that cannot be managed by accounting conventions alone.
Required minimum distributions arrive on a non-discretionary calendar. Retirees at the required age must withdraw a statutory percentage of their tax-deferred accounts annually, distributed as cash. The demographic wave of the baby boom generation moving through this threshold is not a forecast; it is arithmetic. The cash must come from liquid assets. The liquid assets are public equities and short-term Treasuries. The selling is mandatory and calendar-driven. It will happen regardless of market conditions. In years when other institutional selling pressure is also elevated, the non-discretionary RMD flows amplify the move.
The impairment cliff arrives on the accounting calendar, not the economic one. Corporations that capitalized billions in AI infrastructure in 2024, 2025, and 2026 will carry those assets on the balance sheet until annual impairment testing requires projecting future cash flows from those assets. If the macro environment has contracted demand for the services those assets were built to provide, auditors will require write-downs. The charges are non-cash but reduce reported net income immediately and entirely. Companies that have been sustaining earnings per share through buybacks while capitalizing operational expenditures will find both levers constrained simultaneously: buyback capacity reduced by cash flow pressure, and capitalized assets returning to the income statement as impairment charges.
What Is Different About This Time
The 1930s system broke fast. The visible collapse ran roughly three years (1929 to 1932) before the RFC, the bank holiday, and the New Deal restructured the terms of the breakdown. The mechanism was identical to what operates today. The velocity was unmediated.
Today’s system has four suppression layers that the 1930s lacked: legal redemption gates prevent forced selling of private assets, mark-to-model accounting prevents forced recognition of impairment, FDIC prevents deposit runs that would otherwise force rapid bank balance sheet reconciliation, and central bank facilities can provide emergency liquidity to prevent interbank freezes. Each layer slows the velocity of recognition without changing the underlying condition. The deferred reconciliation runs over decades rather than years. Municipal balance sheets degrade slowly. Funding ratios drop in increments. Service quality declines imperceptibly until it becomes undeniable. The loss distributes across time and across millions of individual retirement income shortfalls rather than appearing as a single dramatic crystallization.
This is the meaningful structural difference: the mechanism of deferred versus acute reconciliation. The 1930s exposed the loss rapidly and violently, then rebuilt from cleared prices. The modern architecture suppresses exposure indefinitely, absorbing the loss through degraded outcomes across the institutions that hold the assets and the populations that depend on those institutions. The loss is real in both cases. The 1930s case was visible; the modern case is designed to remain invisible for as long as the suppression architecture holds.
The three novel elements (passive bid, silica trap, sovereign ceiling) do not change this fundamental dynamic. The passive bid extends the period over which public indices appear stable. The silica trap creates a specific accounting-calendar event in late 2027 when capitalizations must return to the income statement. The sovereign ceiling constrains the terminal backstop in ways that make the traditional inflate-and-wait resolution harder to execute without accelerating the currency debasement that is itself one of the burden distribution vectors.
The system is not approaching a condition that resembles the 1930s. It is in that condition. The recognition is lagging because the instruments designed to suppress recognition are functioning. When forced selling from non-discretionary obligations crosses the automated passive bid, when CLO coverage tests fail at scale, when the corporate refinancing wall produces visible default rates in leveraged credit, the same mechanics that were visible on stock tapes and in bank windows in 1931 will be visible in whatever instruments carry price discovery today.
The difference is the time it has taken to arrive. The mechanism has been the same throughout.
This analysis describes structural patterns, mechanisms, and their historical antecedents. It is not investment advice, financial guidance, or market timing recommendation. No content should be interpreted as suggesting specific investment actions. Readers should consult qualified financial professionals before making investment decisions.
Analysis draws on Fisher’s Debt-Deflation Theory (1933), Friedman and Schwartz’s Monetary History of the United States (1963), and Eichengreen’s Golden Fetters (1992). Monetary aggregate and reserve data from Federal Reserve H-series statistical releases. Pension funding data from GASB financial disclosures and public plan actuarial reports.
— Free to share, translate, use with attribution: D.T. Frankly (dtfrankly.com)
§