Who Sells Paper, Who Stacks Metal

The Fed's October-November schedule buys $14.6bn of bills and leaves reserve management paused. Read against the curve, it says the official sector is a bill buyer now, not a duration buyer — and the long bond is left to whoever will hold it at a price. That price has a name, and it is up.

Every bond has a buyer, and the interesting question is never how much paper exists but who is holding it. When the buyer who purchases because they must — not because the yield tempts them — steps back, the price-sensitive buyer steps in. That handover is what a term premium is, in plain language: the extra yield a lender demands for committing money for years instead of days.

This piece reads 1 week of official schedules against our own curve. The headline number is small, and the small number is the point. What matters is where every official operation lands, and which part of the curve nobody has scheduled anything for. The tell is not the volume. It is who sells paper, and who is quietly stacking metal.

What the Fed actually scheduled

The New York Fed released its October-November purchase schedule on 14 October. Three lines matter. The Desk will buy $14.6bn of Treasury bills between 15 October and 13 November as it reinvests maturing agency mortgage principal. It will make no reserve-management purchases: the operation that adds liquidity outright stays paused to mid-November. And every dollar of it lands in bills — instruments inside a year.

Set the $14.6bn against the flow it sits in. Gross Treasury issuance runs near $4tn a year. A month of $14.6bn is a rounding error, well under half a percent — not a bid, not even a gesture. Nor is it new money: the July FOMC directive says all agency principal is reinvested into Treasury bills, so this is mortgage runoff being recycled, mechanically, as the balance sheet shrinks.

The second line is the one that reads hawkish once you stop reading it as a headline. Keeping reserve-management purchases paused is the Fed saying reserves are ample. They are: bank reserves stood at $3.03tn on 7 October, back above the three-trillion mark, and overnight reverse repo sits at roughly zero. The funding side of the system is calm. The Fed is standing down because it can.

The handoff nobody announced

Now line up the three official actors and watch where each one's money goes. The Fed reinvests into bills. The Treasury's buybacks are funded by fresh issuance, so they shift the maturity of the stock rather than the burden — and the stock it is shifting toward is short. Bills are already 22.4% of marketable debt, the highest share since 2020 and above the ~20% the Treasury's own advisory committee treats as a ceiling in normal times.

And the third actor, the one that used to hold the long end without asking the price: foreign officials have fallen to 12.8% of marketable Treasuries, the lowest share since 1993, down from roughly a third in 2012. Their bill buying is down sharply year on year. This is not a diplomatic problem with a diplomatic fix — they are selling for their own reasons, reserve diversification first among them, and a central bank that no longer needs to recycle a trade surplus has no reason to hold your duration.

So the marginal buyer has changed character. It is no longer an official who buys because reserves must be parked somewhere. It is a domestic fund manager who buys because the yield clears their hurdle — money-market funds holding trillions in short paper, government bond funds pulling record inflows, dealers taking a historically tiny slice of auctions because end-investors showed up instead. That buyer is real, and that buyer is price-sensitive. There is the whole shift, in one word.

The number that names the price

A price-sensitive buyer who must be persuaded is exactly what the term premium measures. Ours, from the 10yr model we track, reads 1.07 as of 2026-10-09 against a long-run mean of 0.82 — elevated, not extreme, and pointing the right way.

Here is why that matters more than the level of any single yield. The 10yr nominal yield is the sum of two things: the real yield and expected inflation — the breakeven. Our breakeven sits at 2.33% (13 October), near the bottom of its recent range, while the real 10yr sits at 2.91%. Anchored inflation expectations plus a rising term premium is not an inflation story. It is a creditor story: the people who lend to the sovereign are asking for more to keep doing it, and they are asking in the currency of real yields, not in fear of prices.

This is the same channel our Scarcity call has been watching, with the 10yr real yield at 2.91% against its 2.95% line. A term premium that climbs while the breakeven stays pinned tells you the market is repricing who bears the duration, not how high inflation will run. Those two look identical on a yield chart and mean opposite things for policy.

Who stacks metal

The other side of the handoff is visible in what the same official sector is buying instead. Central banks have been net buyers of gold, and the metal sits at $4,123 an ounce in our marks. That is not a coincidence running alongside a Treasury story; it is the same decision seen from the other end. A reserve manager who wants a claim that no other government can freeze, sanction or inflate away is not choosing between paper and metal on yield — they are choosing on the basis of who controls the claim.

Strip the drama off it and the arithmetic is plain. The official sector is a bill buyer and a metal stacker. The first keeps its money liquid and short, and takes almost no duration risk. The second keeps its money outside the system that prices its own debt. Between those two choices, the long bond is left to the private market — and the private market charges for the privilege. The term premium is not a malfunction. It is the bill for that arrangement arriving.

What would prove this wrong

This is a reading, and it is falsifiable, which is the only reason it is worth publishing. The claim is that the long end is now carried by price-sensitive money and priced accordingly. If that is true, official operations should keep landing short, the term premium should keep the bid, and the long end should fail to rally on any easing — because easing lowers the expected policy path without restoring the price-insensitive buyer.

Watch for the opposite. A sustained fall in 30yr yields with the deficit unchanged would break it, and so would a return of official duration demand — a foreign central bank rebuilding its coupon holdings, or the Fed announcing outright purchases that are not bills. None of that is in the current schedule. Until it is, the tell stands: watch who sells the paper, and watch who stacks the metal.

One item flagged rather than laundered, because it circulates attached to this story. The Secretary's “I am the house now” line is real, but it was said on 8 September about yen intervention, not about bond buybacks — two events in the same week get fused because they landed together. The claim of a nearly $1tn General Account as buyback ammunition we could not confirm from a primary source, so it does not appear here as fact. The schedule above is the primary document, and it says what it says.