Fiscal Stress Board
Sources dated where they differ
Log your own below — saved in your browser
Twenty-one indicators, four systems
Ordered by what breaks first. Funding plumbing fails before sovereign debt, which fails before risk appetite, which fails before the currency. Each row carries the level that would trip it.
1 · Funding & plumbing — breaks first, most reliable
2 · Sovereign debt stress — where it is already firing
3 · Risk appetite — the complacency check
4 · Currency & the hard-asset bid — the slow signal
The credit-card arithmetic
Call it financing a mortgage on a credit card. Here is that metaphor in actual numbers — and it holds up better than I first gave it credit for.
With ~$10 trillion rolling every year, each 100bp adds roughly $100bn to annual interest within twelve months — about a tenth of the entire current interest bill, arriving in year one rather than spread over a decade. Net interest is already $1.0tn in FY2026, running 10.6% above last year through ten months, and consuming 19 cents of every federal tax dollar. CBO has it reaching $2.1tn by 2036.
That is the credit card. The Treasury is saving roughly 100–150bp today by funding at the short end instead of locking in 5.21% for thirty years — and in exchange it has made the national debt a floating-rate liability that reprices against the Fed every few weeks.
Debt/GDP stabilises when the interest rate on the debt stays below nominal growth. Right now the average rate is ~3.3% against nominal GDP growth of roughly 4.5% — still favourable, which is the strongest argument against imminent crisis. But the average is converging upward on the ~4.4% marginal rate, and CBO has growth moderating. Shortening the maturity is what sets the speed of that convergence. A long maturity would let the old low coupons dilute the new high ones for a decade. Bills do it in about three years.
So the maturity choice does not change whether r crosses g. It changes when — and it moves the crossing from beyond the forecast horizon to inside it.
In 2024 Nouriel Roubini and Stephen Miran published Activist Treasury Issuance, arguing that tilting issuance toward bills was “stealth QE” — roughly $800bn of it — because it strips duration out of the market through the same channel as the Fed's asset purchases, and that Treasury was thereby usurping monetary policy. Miran now chairs the Council of Economic Advisers, and bills have gone from 2025 supply of ~$360bn to ~$827bn in 2026 while Treasury buys the long end back.
Whether you call it monetisation or debt management, the consolidated government is shortening the duration of its liabilities — which is exactly what QE does. The Fed is not doing it. The Treasury is.
What I actually think breaks, and in what order
1. Funding markets go first. The RRP buffer is spent, so every Treasury cash build now drains bank reserves one-for-one. Bessent is reportedly considering the ~$1tn Treasury General Account to fund buybacks. A large TGA rebuild landing on quarter-end, alongside heavy bill settlement, is structurally the September 2019 repo setup. SRF usage is the tell, and it is currently quiet — watch it, don't assume it.
2. The transmission is duration, not defaults. The usual crash script runs through credit losses. This one doesn't have to. A 42.5 CAPE with 40% of the index in ten long-duration names is a bond-substitute portfolio wearing an equity label. The long end re-rating is a discount-rate shock to exactly those assets. You can get a severe equity drawdown with high-yield defaults never rising at all.
3. The dollar is the last domino, not the first. A hawkish Fed with 2.4% real yields defends the currency. The FX break comes only if the Fed is forced to cut into inflation to protect the funding market — fiscal dominance arriving through the Treasury's maturity structure rather than through the Fed's balance sheet. That is the scenario where the debasement thesis becomes exactly right.
30-year yields rising while the dollar falls and equities fall, on the same day, three or more times in a month. Right now those three move independently — which is why this is a watch, not an alarm. When they start moving together, the market has stopped treating Treasuries as the risk-free asset and started treating them as the risk. That correlation is the whole signal. Everything else on this board is context for it.
What would prove me wrong, and I want to say it plainly: if breakevens stay near 2.3%, auctions stabilise once the buyback programme scales, and reserves hold above $2.75tn, then this is an ordinary term-premium normalisation after fifteen years of suppression — uncomfortable, not systemic. That remains the more likely single outcome. The board exists because the tail is fat, not because the tail is the base case.
Bear — Robin Brooks (Brookings senior fellow, formerly IIF chief economist), writing 21 August, two days after the buyback expansion: capping long yields by unconventional means does not touch the fiscal problem, it moves the stress into the currency. “Markets are primed for Dollar debasement to resume and — as Japan shows — it can be next to impossible to stabilize a currency once it enters a devaluation spiral. The U.S. is playing with fire with this buyback.” His template is Japan: 200%+ debt/GDP, suppressed JGB yields, chronic yen depreciation. A debt crisis becomes a currency crisis.
Bull — Jonas Goltermann (chief markets economist, Capital Economics), in the same piece: debasement worries are “overblown.” He expects the dollar to strengthen, reading the recent drop as ordinary yield-differential movement rather than lost inflation-fighting credibility — with the caveat that “if the steady stream of unconventional policy ideas continues, that may well change.” Lawrence Gillum (LPL Financial) makes the third-way case: the long-end selloff is normalisation after fifteen years of suppression, not dysfunction.
The strongest fact against my own bear lean: if the world were genuinely walking away, you would see failed auctions, not merely a high term premium. Global investors absorbed roughly $2.5tn of US deficits over the past year, and 10y and 30y bid-to-cover has stayed above 2.5 across most of the cycle. August's 2.39 is soft, not a strike.
On timing: I can give you ordering and thresholds. I cannot give you dates, and neither can anyone else — the 2019 repo crisis, to take the closest analogue, was invisible three weeks before it happened and obvious three days after. What this board buys you is the difference between three days and three weeks.
Log your readings
The page can't fetch prices, so the source links above open the primary data. Enter what you find here — readings are saved in this browser only.