Alan Greenspan said there was zero probability of default because the United States can always print the money. He was right, and it was not news. Thirty-six days earlier the agency that was about to downgrade the country had published a paragraph conceding the point in advance — and explaining, in writing, that its rating was about a choice rather than a capability.
“Replace ‘print’ money with ‘create’ money in the form of deposits created by the Fed for the banking system and you’ll see that the Maestro agrees with #MMT. A monetarily sovereign country cannot run out of its own currency. Just wait and see.”
Four moving parts. One of them is true, and is the least interesting thing on the page. Two of them misread the man being quoted. The fourth — “just wait and see” — treats a settled question as an open one, fifteen years after it was settled in writing by the party on the other side.
The clip is real. On Sunday 7 August 2011, two days after Standard & Poor’s cut the United States from AAA to AA+ for the first time in the history of the rating, David Gregory put a question to Alan Greenspan on Meet the Press. The question was not about how to solve the national debt. It was about safety. Behind him, the chyron read U.S. DEBT DOWNGRADE, and carried the new rating: AA+.
What follows is an argument about money that almost nobody is actually having, because both sides agree on the proposition they are shouting about. The disagreement is one layer down, it is written out in plain English in a document that was already public when Greenspan spoke, and it has a number: paragraph 121.
Taken in the order they appear.
A sovereign credit rating is not a measurement of whether a government can produce the currency. It never was. It is a measurement of whether, at some point, it would rather not. That is not a hostile reading imposed from outside; it is what the agency published, in its own criteria, under a numbered paragraph, five weeks before it used them on the United States.
121.One might ask why sovereign local-currency ratings are not all rated ‘AAA’ given sovereigns’ extensive powers within their own borders, including the ability to print money. While the ability to print local currency gives the sovereign tremendous flexibility, heavy reliance on such an expansionary monetary stance may fuel the risk of very high inflation or even hyperinflation, which may cause more serious political and economic damage than rescheduling of local-currency debt. In such instances, sovereigns may opt to default on their local-currency obligations.
Read it slowly, because the order of the sentences is the argument. The objection comes first and it is granted without qualification. The power is real. The flexibility is tremendous — their word. And then the turn: the alternative to default is not free, the alternative has its own price, and a government that has to pick between two bad outcomes may pick the one the bondholder does not like. Default stops being a thing that happens to a state and becomes a thing a state does. Hence the verb. Opt.
That single word is the whole distance between the two sides of this argument. One side is describing a capability and is correct. The other side is pricing a decision and has never claimed otherwise. They are not contradicting each other. They are not even talking about the same kind of thing: one is a statement about what is possible, the other is a probability attached to a choice, and no quantity of the first ever resolves the second.
The same paragraph carries a second concession that gets lost. Local-currency ratings are higher than foreign-currency ratings precisely because of the printing press — S&P allows up to two notches of uplift for it, conditional on the sovereign having an independent monetary policy. The power is not ignored by the model. It is an input with a stated numerical value. It is simply not infinite, because the thing it buys you is the ability to inflate, and inflating is itself one of the outcomes the rating is trying to warn you about.
Nothing in paragraph 121 says the United States is likely to choose inflation over payment, and nothing here argues it. The paragraph is about why the scale exists at all, not about where on the scale any particular country sits. The United States sits one notch below the top at all three, for reasons each of them stated separately, and those reasons are in the next section.
Three agencies have now taken the United States off the top rung. Their press releases are public and short. Not one of them says the country cannot pay. Read in order, they say the same thing three times in fifteen years, in slightly different words, and the words are about process.
| Action | What the agency actually said |
|---|---|
| S&P 5 Aug 2011 AAA → AA+ |
The downgrade reflects the view that “the effectiveness, stability, and predictability of American policymaking and political institutions have weakened” at a time of fiscal and economic challenge, and pessimism about the capacity of Congress and the Administration to bridge the gulf between the parties on fiscal policy. |
| Fitch 1 Aug 2023 AAA → AA+ |
Expected fiscal deterioration, a high and growing debt burden, and “the erosion of governance” over two decades “that has manifested in repeated debt limit standoffs and last-minute resolutions.” |
| Moody’s 16 May 2025 Aaa → Aa1 |
The rise over more than a decade in debt and interest payment ratios to levels well above similarly rated sovereigns; successive administrations and Congresses have “failed to agree on measures to reverse the trend.” The same release records the country’s default history as none since 1983. |
| S&P 26 Jun 2026 AA+ affirmed |
Held at AA+ with a stable outlook, on the strength of a resilient economy, broad revenue collection including tariffs, and credible monetary policy. Fifteen years after the cut, the rating has not come back. |
The Moody’s release contains the cleanest proof available that none of this is about currency capacity, and it is buried in the technical paragraphs. A sovereign gets an issuer rating — what the government itself is worth as a borrower — and the country gets a ceiling, which is the agency’s view of the currency, transfer and convertibility risk sitting over everything inside the border. On 16 May 2025, Moody’s cut the issuer rating to Aa1.
It left the local- and foreign-currency country ceilings at Aaa.
The gauge that measures currency risk did not move. The only thing that moved was the government’s own standing as a borrower. If the downgrade had been a judgement about dollars, the ceiling is where it would have shown up, and it is exactly where nothing happened. The release even explains why the local-currency ceiling stayed perfect: a small government footprint in the economy and “extremely low risk of currency and balance of payment crises.”
And then the market got a vote, which is the part most worth sitting with. S&P cut the United States after the close on Friday 5 August 2011. The ten-year Treasury yield that day was 2.58 per cent. The following Monday it was 2.40. By Tuesday, 2.20. On 18 August, 2.08. By 22 September it was 1.72 per cent. In the seven weeks after the United States was told it was a worse credit, the world lent it money for a third less.
2011 was a flight to safety in the middle of the European sovereign crisis, which Greenspan himself spent the rest of that same interview talking about. The pattern does not repeat. After Fitch cut on 1 August 2023 the ten-year went from 3.97 to 4.20 within two sessions and 4.34 by 21 August. After Moody’s cut on 16 May 2025 it went from 4.43 to 4.58 within three sessions. Yields are not a verdict on a rating; too much else moves them. What 2011 shows is narrower and sufficient: a downgrade of the United States is fully compatible with the world paying more, immediately, to hold its paper. Whatever the agencies were pricing, the market was not pricing non-payment.
On 29 September 2026, the last business day of the fiscal year, total public debt outstanding closed at $40,096,954,633,566.68, of which $32.37 trillion is held by the public. The ten-year yield that week was 5.26 per cent, the highest it has been at any point in this story. Nobody is refusing to lend. They are repricing.
There is a counter-example to “zero probability,” it is small, it is well documented, and it proves the point rather than breaking it. In late April and early May 1979 the United States Treasury did not pay holders of maturing Treasury bills on time.
The amount was about $120 million. The first delayed payments were due on 26 April 1979. The cause was clerical: a debt-limit fight had driven an unusual volume of activity through the Treasury’s redemption desk, the paperwork backed up, and the word-processing equipment that produced the cheques failed. Investors were paid, late.
Terry Zivney and Richard Marcus went looking for the price of it a decade later, in a paper whose title says the finding — “The Day the United States Defaulted on Treasury Bills,” The Financial Review, 1989. They found Treasury bill rates rose roughly 60 basis points at the first delay and stayed elevated for months afterwards. A paperwork failure on a tenth of a billion dollars put a measurable, persistent premium on the whole curve.
The United States did not run out of dollars in 1979. It ran out of working cheque printers during a debt-limit backlog.
That is the shape of the real risk, and it is the shape every one of the three downgrades describes. The binding constraint on paying a dollar-denominated debt in dollars is never the supply of dollars. It is the machinery of authorisation that stands between the obligation and the payment — and in the American system that machinery is statutory, it is self-imposed, and it has been switched off on purpose more than once.
Put the two facts together and the sentence everyone is arguing about becomes almost beside the point. The United States cannot run out of its own currency; the United States can be prevented by its own statute from using it. Both are true at once. The first is a fact about money. The second is a fact about law, and it is the one with a date on it.
If the capacity to create dollars is not what a rating measures, it is fair to ask what is. Moody’s answered that in its own release and gave the measure: federal interest payments as a share of federal revenue. It put the figure at about 18 per cent in 2024, up from about 9 per cent in 2021, and projected around 30 per cent by 2035.
That measure can be rebuilt from the Treasury’s own books. Net interest comes from the Monthly Treasury Statement, budget function 900; receipts come from the same statement’s summary table. Both are published by the Department of the Treasury and both are machine-readable. Reconstructed that way, FY2024 comes out at 17.9 per cent and FY2021 at 8.7 per cent — which is how you know this is the same quantity Moody’s was talking about.
So: two gauges. The left one is the thing the claim is about, and it is pinned. The right one is the thing the rating is about, and it has not been pinned for a decade.
Move the fiscal year. One needle is a statement about law. The other is a division.
There is no fiscal year in which the United States lacked the legal power to issue the currency its debts are denominated in. This needle has no history. That is the whole of what the claim asserts, and it is correct.
—
One further number, kept separate because it is a different measure and mixing the two is how this subject gets fudged. Moody’s also reported general government interest — federal, state and local together — at 12 per cent of revenue in 2024, against 1.6 per cent for the sovereigns still rated Aaa. That is not the same quantity as the gauge above and it does not belong on the same scale. It is here because it is the comparison the agency actually made, and because 12 against 1.6 is the gap that a notch of rating is standing in for.
None of this says the gap is a crisis, and none of it says it is not. The point is narrower, and it is the point of the whole volume: every figure above is compatible with the United States being able to create every dollar it owes, forever, and none of them is improved by that fact.
Each line is a statute, a published criteria document, a rating action or a Treasury series.
Read down the left column and the dates sort themselves into two kinds. The 1935, 1942, 1981, 2025 and 2027 entries are about permission — who may issue, who may buy, up to what number, until when. The 2011, 2023 and 2025 entries are about judgement — whether the people holding those permissions can be relied on to use them in time. Neither kind of entry is about whether the dollars exist.
The claim says a sovereign cannot run out of its own currency.
The agency wrote that down first, granted it in full, and called it tremendous.
Then it wrote the next sentence, which is the one about choosing.
Visual direction after the oceanographic films of Jacques-Yves Cousteau and the crew of the R.V. Calypso, 1943–1996. The saucer is the SP-350 “Denise”. No affiliation; the debt is the point.