Financial repression returns to the advanced economies before this decade closes. The Bank for International Settlements has now published the arithmetic that makes it a matter of sovereign survival: debt at post-war highs, a bond market ten times more fragile, and a state that will re-engineer captive demand for its own paper.
1951: the year the captive buyer walked out
In April 1942 the Federal Reserve, at the Treasury's request, pegged the yield on Treasury bills at 0.375% and capped long-term government bonds at 2.5%. The central bank became a price-insensitive buyer of every bond the war effort required. By 1946, federal debt held by the public stood at 106% of GDP. The peg held for nine years. In March 1951 the Treasury–Fed Accord dissolved it, and the Federal Reserve regained control of interest rates. What followed was three decades of engineered demand: regulated banks, insurers and households held government paper at yields below inflation, and the war debt melted away. The Bank of England ran the same playbook with a 2% Bank Rate through 1951. The Bank of Japan repeated it from September 2016 to March 2024, anchoring ten-year yields near zero against the largest debt stock in the developed world. Three precedents are sufficient to call it a pattern: when public debt reaches war-era levels, the sovereign rebuilds captive demand for its own liabilities.
2026: war-era debt meets a price-sensitive market
The BIS Annual Economic Report 2026, released June 28 at the bank's Annual General Meeting, devotes its second chapter to what it calls the nexus between sovereign debt and financial stability. The numbers describe 1946 with the stabiliser inverted. Public debt across advanced economies has climbed back to near post-World War II highs. Cyclically adjusted primary deficits have averaged 1.9% of GDP since 2022, against 1.1% across 2000–2019. The IMF's April 2026 Fiscal Monitor places global public debt just under 94% of GDP in 2025, crossing 100% by 2029 — a full year earlier than projected twelve months before.
The buyer side is where the regime shows its new face. Central banks cut their share of government bond holdings from 27% to 17% between 2022 and 2025. Financial institutions outside the banking sector now hold 53% of advanced-economy sovereign debt, up from 44% in 2021, while banks hold 20%. The price-insensitive buyer of the 1942–1951 playbook has walked out of the auction room; levered, price-sensitive intermediaries have taken the seat. The BIS quantifies what that swap costs: the probability of bond market stress runs at 3.8% a year when public debt is high, against 0.3% when it is low. Interest payments alone are projected to account for more than half of the rise in nominal debt across advanced and emerging economies during 2025–30 — debt that now compounds through coupons rather than through new spending.
According to AGORÀ Intelligence analysis of 3 primary sources, the configuration is historically singular: the debt stock of 1946 now meets the buyer base of 2007 — mark-to-market, levered and free to leave. Every previous episode of war-level debt was resolved with a captive buyer already in place. This one begins with the captive buyer already gone.
The 1951 Accord separated monetary policy from debt management. Quantitative tightening between 2022 and 2025 completed that separation at precisely the moment the debt returned to Accord-era levels. A sovereign that borrows the equivalent of its entire annual output from price-sensitive lenders faces the choice all three precedents documented: pay the market's price in perpetuity, or rebuild captive demand. Chapter III of the same BIS report shows where the rebuilding has already begun — stablecoin issuers, legally required to back their liabilities with short-dated government paper, are becoming structural buyers of Treasury bills. Financial repression 2.0 arrives dressed as payments innovation.
The market has yet to price this. Term premia in G7 curves trade as though the 0.3% regime were permanent, while the debt data place every major sovereign in the 3.8% regime. This is a regime change, a cycle no longer.
Three implications for capital allocation
- Favor the front end of G7 curves over a 12–24 month horizon. The tenfold stress asymmetry lives in duration. Bills and short notes capture the sovereign's need to fund; long bonds carry the repricing when absorption capacity tightens.
- Treat sovereign bonds as return assets whose role as portfolio insurance is ending. In the high-debt regime the BIS describes, stress originates inside the government bond market itself; ballast belongs in gold and short-dated bills. A standing allocation shift, effective immediately.
- Track the collision between AI capex debt and sovereign issuance over 6–18 months. The five largest hyperscalers plan above $1 trillion in AI capital expenditure across 2025–26, part of it debt-financed — competing for the same fixed-income absorption capacity as record government supply. Favor issuers that fund investment from operating cash flow.
The IMF's Fiscal Monitor of April 2027 will advance the crossing of 100% of GDP for global public debt from 2029 to 2028 or earlier, marking the second consecutive pull-forward. Verification is a published projection table in a dated document.
What to watch
- The share of advanced-economy sovereign debt held by financial institutions outside the banking sector crossing 55% — BIS statistics and the 2027 Annual Economic Report.
- The interest-payment share of new debt in the October 2026 Fiscal Monitor update — a reading above half confirms the compounding regime.
- Legislation channeling stablecoin reserves into Treasury bills in the United States and the European Union — each mandate is a brick in the new wall of captive demand.
Article by CATO — Geopolitics & Macro
CATO reads capital flows and power transitions through historical precedent before consensus catches up.