The Fiscal-Supply Regime

Abstract

Research summary

Independent research. Views and data are stated as of the research cut-off of 1 September 2026. This is a conditional classification, not individualised investment advice.

01.

An event is not its consequence

Economic commentary is organised around events. A war begins, an election is decided, an oil price rises, a central bank changes its rate, or a government announces a budget. The event is visible and therefore becomes the explanation. Yet identical disturbances have repeatedly produced different economic consequences across periods and jurisdictions. What differs is the state through which the disturbance is transmitted.

A regime fixes many of the constraints that bind before the event occurs: the condition of private balance sheets, the fiscal capacity of the sovereign, the credibility of the monetary authority, the structure of labour supply, the price of duration and the availability of external financing. Those conditions govern whether a subsequent disturbance becomes inflationary, disinflationary, expansionary, contractionary or largely absorbed. The event supplies the impulse; the regime supplies the transmission mechanism. This is the practical implication of a large state-dependent literature in macroeconomics and finance (Hamilton, 1989; Ang & Bekaert, 2002).

The distinction matters because the same observed event can therefore imply very different things for policy and capital. An oil shock encountered with weak demand, ample fiscal capacity and anchored expectations is not the same economic event as an identical oil shock encountered amid persistent deficits, constrained labour supply and a higher sovereign term premium. The physical disturbance may be the same. The economic consequence is not.

A second literature concerns the interaction of fiscal and monetary policy. When monetary and fiscal authorities interact, inflation cannot always be understood from the policy rate alone (Leeper, 1991; Sargent, 1982; Sims, 1994). Bianchi and Melosi (2022) extend that line of reasoning to the post-pandemic United States and assign a role to fiscal conditions in the inflation process. These results do not imply that deficits mechanically determine inflation. They imply that the monetary reaction function, the expected path of primary balances and the public's beliefs about government liabilities belong in the same analysis.

A third literature concerns the financial cycle. Credit and property balance sheets shape the depth and duration of downturns (Borio, 2014; Jordà, Schularick, & Taylor, 2016; Schularick & Taylor, 2012). That literature enters the 2020s with vulnerabilities, but not with the household debt-service burden that preceded 2008: the U.S. household debt-service ratio was 11.16 percent of disposable income in the first quarter of 2026, against 15.85 percent at the 2007 peak (Board of Governors of the Federal Reserve System [FRB], 2026a). A lower payment burden does not eliminate credit risk; it changes where the risk is concentrated and how a policy shock is transmitted.

02.

Identifying the present regime

Definition. Chapra & Co. uses fiscal-supply regime to describe a macroeconomic state in which recurrent supply disturbances interact with persistent sovereign borrowing, an elevated price for duration and reduced monetary-policy flexibility, while private balance sheets remain sufficiently intact to prevent the adjustment from taking the form of a conventional household-led deleveraging cycle. The significance lies in the interaction. None of these conditions, taken alone, constitutes the regime.

No single historical label describes the present configuration well. It is not a conventional demand-overheating episode: by July 2026 headline U.S. CPI inflation had eased to 3.4 percent and core CPI to 2.5 percent, while the Federal Reserve maintained a target range of 3.50 to 3.75 percent. A contemporaneous CPI subtraction is not, however, an adequate measure of the ex-ante real policy stance. The New York Fed's July survey placed one-year household inflation expectations at 3.6 percent, while longer-run measures remained much better anchored: the Federal Reserve reported a five-year/five-year professional-forecaster expectation of 2.1 percent (Federal Reserve Bank of New York [FRBNY], 2026; FRB, 2026e). The more useful distinction is therefore between still-elevated near-term inflation uncertainty and a longer-run nominal anchor that has not broken.

Supply conditions in 2026 are also partly policy-made rather than purely exogenous. CBO expects higher tariffs to raise imported-goods and intermediate-input costs, reduce foreign investment and lower economic efficiency. Lower net immigration, meanwhile, slows labour-force growth. Its price effect is two-sided: weaker population growth reduces demand for goods and housing, but a smaller labour supply raises wage and non-housing-service price pressure in sectors where workers are scarce (CBO, 2026). That combination is important for the regime framework because the same policy can restrain aggregate demand while tightening particular supply constraints.

The fiscal backdrop is unusually important. The Congressional Budget Office projects a federal deficit of 5.8 percent of GDP in fiscal year 2026 and debt held by the public of 101 percent of GDP (Congressional Budget Office [CBO], 2026). Recent Federal Reserve research strengthens the supply channel in this argument. Using high-frequency Treasury auction shocks, Bi, Phillot, and Zubairy (2026) find that debt-expansion shocks raise yields across the curve through higher term premia and tighter financial conditions. Dallas Fed work similarly finds that the supply of Treasury duration is an important driver of the term-rate premium. The relevant claim is therefore not that deficits mechanically set long rates, but that persistent sovereign financing needs can raise the compensation private balance sheets require to absorb duration.

The sovereign financing channel is also a market-structure question

The supply of government debt does not reach the economy through yields alone. It must be intermediated. The International Monetary Fund's April 2026 Fiscal Monitor identifies two changes that matter for this paper: a larger role for leveraged non-bank intermediaries in sovereign-debt markets and an erosion in the safety or convenience premium traditionally attached to U.S. Treasuries. Both make the capacity and composition of the marginal balance sheet more important when public issuance is large (International Monetary Fund [IMF], 2026e).

Debt management itself provides a contemporary illustration. On 19 August, the U.S. Treasury announced that it would at least double the maximum size of liquidity-support buybacks in the 10-to-30-year nominal sectors, from $2 billion to at least $4 billion per operation from 9 September. Treasury described the measure as liquidity support and simultaneously noted strong sponsorship and a significant volume of high-quality offers. The evidence therefore does not support a claim of market dysfunction. It does support a narrower proposition: maintaining liquidity in the long end has become a more active debt-management consideration as the stock and flow of duration rise (U.S. Department of the Treasury, 2026a).

The present is also not a straightforward replay of 1979. Energy remains capable of producing large relative-price shocks, as the 2026 Middle East disruption demonstrated, but the energy intensity of advanced-economy output is lower and labour-market institutions differ materially from the 1970s. Nor is the environment well described by the secular-stagnation configuration of the 2010s. Public deficits are larger, strategic investment in defence, energy and industrial capacity has increased, and demographic pressures are less disinflationary than during the period of expanding global labour supply described by Goodhart and Pradhan (2020).

We use fiscal-supply regime as a compact description of that conjunction: recurrent supply disturbances; fiscal deficits large enough to keep sovereign financing needs material through the cycle; a less reliably disinflationary labour backdrop; and household balance sheets that are less payment-constrained than before 2008 even as leverage remains concentrated elsewhere. The term is useful only if it accounts for observations that competing explanations handle less comfortably. It should therefore be read as a testable classification, not as a claim that fiscal variables mechanically determine inflation or yields.

Counterevidence: the nominal anchor has not broken

A regime claim is useful only if it records observations that sit uncomfortably with it. First, longer-run inflation expectations remain well anchored. In July, the New York Fed reported median household expectations of 3.6 percent at one year, 3.3 percent at three years and 3.0 percent at five years, while the Federal Reserve reported a 2.1 percent five-year/five-year expectation among professional forecasters (FRBNY, 2026; FRB, 2026e). Second, CBO's baseline combines persistent fiscal deficits with a return of PCE inflation to about 2 percent by 2030 (CBO, 2026). Both observations count against the strongest version of the fiscal-supply thesis: persistent deficits do not imply a broken nominal anchor.

Two historical periods are particularly useful as reference regimes. The first is 1946 to 1951, when debt constraints and a close Treasury–Federal Reserve relationship made the real value and distribution of public debt central to the macroeconomic adjustment. The second is 1968 to 1982, when recurrent energy shocks and persistent inflation demonstrated the limits of treating a supply disturbance as a purely demand-management problem. The present shares mechanisms with both periods, but also differs in institutional credibility, energy intensity, labour institutions and private leverage. Those differences are not footnotes; they determine how much weight each comparison can carry.

03.

The record through the regime lens

03A. 2022 to 2025: a change in the joint distribution

The 2022–25 period is notable less for any single event than for the way several relationships changed together. Equities and bonds sold off simultaneously in 2022 as inflation risk pushed the stock–bond correlation positive, and sovereign debt ceased to provide the same portfolio insurance it had provided through much of the preceding decade. A positive term premium re-emerged as fiscal issuance increased and the marginal buyer of duration changed in composition. At the same time, the 2023 banking stress did not develop into a general household-credit crisis, because the private sector entered with better household debt-service capacity than before 2008, even as lending standards remained tighter in particular segments (FRB, 2026a, 2026c).

The other side of the regime was investment rather than deleveraging. Artificial-intelligence compute, electricity and grid investment and strategic industrial policy produced a capital-expenditure cycle financed substantially from the operating cash flows of large firms and the public sector. That composition matters: a cash-financed installation phase can correct sharply in equity prices without creating the same banking transmission as a debt-financed one. The framework therefore has to accommodate inflationary supply constraints and investment simultaneously, rather than assuming that one must immediately extinguish the other.

03B. 2026: reference episodes, not historical scripts

The 2026 record is examined below through historical reference episodes. An analogue is useful only where a transmission mechanism is comparable. It is not admitted because two price charts look similar, and its subsequent chronology is not treated as a forecast. Differences in policy credibility, leverage, trade structure, energy intensity and institutional design determine the weight placed on each comparison.

The Middle East energy shock reopened headline-inflation risk without recreating the 1970s. July U.S. CPI inflation stood at 3.4 percent, while the energy component was 14.7 percent higher than a year earlier (U.S. Bureau of Labor Statistics [BLS], 2026). The relevant 1973 and 1979 comparison is the mechanism — a supply shock that complicates the inflation–growth trade-off — rather than the magnitude. Lower energy intensity and better-anchored expectations reduce the case for a mechanical repetition of the 1970s sequence.

The Federal Reserve held the policy rate while the long end carried more of the adjustment. The July target range of 3.50 to 3.75 percent sat modestly above contemporaneous headline inflation (FRB, 2026d). The tension therefore lies less in a negative front-end real rate than in the coexistence of elevated inflation uncertainty, large Treasury financing needs and a positive duration premium. The 1946 to 1951 period is informative about the fiscal–monetary interaction, but today's independent central bank and market-based debt management make the institutional setting materially different.

Gold continued to receive an official-sector bid. The World Gold Council estimates net central-bank purchases of 289 tonnes in the second quarter of 2026, with buying broad across several reserve managers (World Gold Council [WGC], 2026). This is consistent with a continuing demand for reserve diversification under geopolitical and fiscal uncertainty. It is not, by itself, evidence of imminent monetary disorder; the relevant question is whether the official bid persists as real rates and fiscal expectations change.

Treasury duration remained a distinct source of risk. Recent Federal Reserve research finds that a larger supply of duration to private investors is associated with a higher term-rate premium. This helps explain why long bonds can remain vulnerable even when expectations for the overnight policy rate are stable. The historical comparison with pre-1998 positive stock–bond correlation regimes is useful for portfolio construction, but the modern institutional investor base and collateral architecture make direct replication unlikely.

The artificial-intelligence investment cycle remained unusually capital intensive. The useful historical comparison is not a particular equity index but the installation phase described by Perez (2002): a period in which infrastructure is built ahead of the full diffusion of the technology. The key present-day distinction is financing. To the extent that the largest investors continue to fund capital expenditure from operating cash flow, an eventual digestion of overcapacity can remain concentrated in equity valuations rather than the banking system.

China remained a source of global disinflationary pressure even as it grew. The IMF projects 4.5 percent real growth for China in 2026 while expecting inflation to recover only gradually from a 2025 average near zero. Weak domestic demand, a prolonged property adjustment and a negative GDP deflator keep the Chinese transmission mechanism different from the fiscal-supply configuration in the United States. Export capacity and a depreciated real exchange rate can transmit part of that domestic disinflation to traded goods abroad.

Emerging economies displayed divergence rather than a common beta. Brazil entered the shock with restrictive monetary settings and a fiscal-credibility problem; Mexico retained a sizeable real-rate buffer but remained closely tied to the U.S. cycle; India combined stronger trend growth with vulnerability to imported energy; and China continued to transmit domestic disinflation through traded goods. Two additional cases sharpen the point. Saudi Arabia, an energy exporter with a dollar peg and substantial reserve buffers, receives part of an oil-price shock through fiscal revenues rather than through the terms-of-trade loss faced by India. South Africa sits at the opposite end of the spectrum: modest growth, public debt near 77 percent of GDP and a structurally higher sovereign financing burden leave it more exposed to global risk premia and domestic fiscal credibility. The relevant unit of analysis is therefore the national balance sheet — external position, energy balance, fiscal trajectory, credit structure and real-rate space — not the label “emerging market” (IMF, 2026a, 2026b, 2026c, 2026d).

Rearmament and industrial policy continued to place a public bid under selected capacity. The historical comparison with the early Cold War is useful at the level of procurement persistence, not as a claim that today's fiscal institutions are equivalent. Multi-year defence orders, grid investment, semiconductor incentives and trade barriers all raise the probability that parts of capital expenditure remain less sensitive to the normal business cycle.

Commercial property and bank credit remained a slow adjustment rather than a single clearing event. The July 2026 Senior Loan Officer Opinion Survey showed that standards remained relatively tight in several commercial-real-estate categories even as some standards eased at the margin. The relevant historical reference is therefore a drawn-out recognition and refinancing process, closer to the savings-and-loan and early-1990s commercial-property episodes than to a single systemic event. That comparison should be used to understand the mechanism of slow balance-sheet repair, not to promise a particular investment vintage.

04.

Conditional paths from the regime

The historical references narrow the set of plausible mechanisms; they do not determine the sequence of future events. We therefore treat the following as conditional paths rather than inherited forecasts.

1. Inflation remains more volatile around target than in the 2010s. Recurrent energy and trade shocks can lift headline inflation quickly, while the fiscal stance and labour-market structure can slow the return to target. The key distinction is between temporary relative-price shocks and a broadening into expectations, wages and services inflation. A sustained rise in longer-run inflation expectations would materially strengthen the regime call; continued disinflation in core measures despite large deficits would weaken it.
2. Monetary policy can ease without restoring the old duration regime. A lower policy rate does not automatically imply lower long-term yields if the supply of duration, inflation uncertainty and fiscal risk remain elevated. Cuts may therefore coexist with a term premium that stays structurally above its 2010s range. The empirical test is the behaviour of long-rate term-premium estimates and auction absorption as the policy rate changes.
3. Fiscal pressure is more likely to appear through prices and portfolio constraints than through an announced policy of repression. Reinhart and Sbrancia document historical episodes in which negative real returns and regulation reduced the real burden of public debt. In a modern market, the relevant mechanisms would be subtler and should not be labelled financial repression without evidence. Changes in collateral rules, liquidity requirements, debt-management composition or the tax treatment of savings matter only insofar as they demonstrably alter the structural demand for sovereign liabilities.
4. The investment boom can be both inflationary in construction and disinflationary in its eventual productivity effects. Compute, power, transmission and industrial capacity increase current demand for scarce physical inputs. If the resulting capital stock raises productivity sufficiently, however, the same investment cycle becomes an exit route from the regime rather than confirmation of it. The timing of that transition is more important than the headline level of capital expenditure.
5. Emerging economies separate by macroeconomic structure. China's weak domestic price dynamics can export tradables disinflation. Brazil and Mexico retain comparatively high nominal and real policy rates that can support disinflation but also expose fiscal and growth trade-offs. India retains a stronger growth profile but is more sensitive to imported energy costs. Across emerging markets, the relevant variables are therefore not the label “EM” but the current account, energy balance, fiscal trajectory, domestic credit cycle and real policy rate.
6. Credit adjustment remains sectoral unless leverage becomes systemic. Current U.S. household debt-service ratios and the absence of a broad bank-lending contraction argue against treating every refinancing problem as 2008. Commercial real estate and parts of non-bank credit can still suffer material losses. A transition from localised stress to a system-wide deleveraging cycle would be a regime change, not a minor variant of this paper's base case.

05.

Sector and asset-class sensitivities

The following are regime sensitivities rather than recommendations. Each is paired with a revision condition that would require the interpretation to be reconsidered.

Energy and commodities

Supply disturbances matter most when spare capacity is limited and new supply has long lead times. The regime is therefore relatively supportive of businesses whose cash flows are linked to scarce production, transport or storage capacity, but the case weakens as investment restores spare capacity and the marginal cost curve flattens. Revision condition: a sustained rebuilding of spare capacity and lower marginal production costs across the principal energy markets.

Defence and industrial capacity

Multi-year public procurement can make selected industrial order books less sensitive to the conventional business cycle. The relevant exposure is to funded backlog and capacity expansion rather than political announcements alone. Revision condition: a durable reversal in bloc-level procurement and industrial-policy commitments.

Capital goods, grid and power infrastructure

Compute and electrification shift part of the technology cycle into physical infrastructure. Equipment, transmission, engineering and cooling therefore inherit both the demand of the digital investment cycle and the constraints of the physical supply chain. Revision condition: a sustained decline in aggregate data-centre, grid and power capital expenditure among the principal financiers.

Financials and private credit

A steeper curve and tighter bank constraints can create room for lenders able to price complex or illiquid risk. That does not make non-bank credit structurally superior: the thesis depends on underwriting quality and on losses remaining commensurate with the additional spread earned. Revision condition: non-bank credit losses persistently exceeding comparable bank losses after adjusting for borrower quality and seniority.

Technology

The financing structure matters as much as the product narrative. Cash-funded capital expenditure can sustain an installation cycle longer than debt-funded expansion, while unprofitable long-duration businesses remain more exposed to the discount rate. Revision condition: the principal infrastructure investors become materially dependent on external debt or equity issuance to sustain capital expenditure.

Real estate

Where financing costs remain structurally higher, operating income, lease structure, scarcity and capital structure matter more than simple cap-rate compression. The regime therefore increases dispersion across property types and operators. Revision condition: broad cap-rate compression accompanied by durable net-operating-income growth and normalised credit availability.

Consumer and labour-intensive services

Persistent labour scarcity raises the value of pricing power and productivity. Businesses that cannot substitute labour or pass through wage costs are more exposed to the regime than firms with flexible cost structures or higher output per employee. Revision condition: sustained unit-labour-cost growth near levels consistent with the inflation target.

Fixed income

The central question for fixed income is whether long-duration yields are driven primarily by expected policy rates or by a persistent term premium. If sovereign duration supply remains large, the risk-adjusted behaviour of long bonds can continue to differ from the 2010s even during periods of policy easing. Revision condition: a durable compression in term-premium estimates alongside a credible decline in net duration supply.

Gold and reserve assets

Official-sector gold demand is consistent with a desire for reserve diversification under geopolitical and fiscal uncertainty. The case is strongest when fiscal credibility is questioned and real rates are not sufficiently high to offset that concern; it weakens as policy credibility improves. Revision condition: a sustained decline in official-sector demand alongside higher ex-ante real rates and an improving fiscal trajectory.

Emerging markets

Emerging markets are best treated as a dispersion problem. China, India, Brazil, Mexico, Saudi Arabia and South Africa occupy materially different positions with respect to energy, external balance, fiscal capacity and monetary space. A common global shock can therefore produce a fiscal windfall in one economy, a terms-of-trade loss in another, and a sovereign-risk repricing in a third. Revision condition: a broad convergence in external and fiscal trajectories sufficient to make cross-country transmission materially more uniform.

Digital assets

Under the firm's standing framework, digital assets remain primarily a liquidity, adoption and risk-appetite exposure operating inside the broader macro regime. The fiscal-supply thesis may influence the liquidity backdrop, but it does not by itself establish digital assets as a hedge against fiscal risk. Revision condition: a sustained contraction in global liquidity and risk appetite.

06.

Revision conditions

The central interpretation would require material revision under four observable developments. First, underlying inflation could return durably to levels consistent with the Federal Reserve's two-percent objective while large fiscal deficits persist, long-run inflation expectations remain anchored and the term premium compresses. CBO's baseline explicitly contains much of that combination; if realised data converge to it, the stronger version of the fiscal-supply thesis would be weakened. Second, a systemic private-credit contraction in the United States or Europe could shift the economy into a deleveraging regime in which demand destruction dominates the supply and fiscal channels described here. Third, the current investment cycle could deliver productivity growth quickly enough to permit strong real activity alongside falling unit labour costs and declining inflation. Fourth, Treasury market liquidity and the sovereign convenience premium could prove largely insensitive to rising debt supply, weakening the proposed link between fiscal arithmetic, intermediation capacity and private discount rates.

The paper is dated because the interpretation is conditional. Subsequent editions should identify which observations have strengthened, weakened or retired the framework rather than treating every new event as confirmation.