Your Vote Doesn't Price. The Auction Calendar Does.
An election is supposed to be the moment a country decides what it will spend. The bond market is supposed to price that decision years in advance. This week both of those ideas took a hit. The spread between the 30yr and the 10yr Treasury — the part of the curve that prices long-run borrowing rather than the next policy meeting — sits near the bottom of its range for the year, and it got there while the government was moving $59bn of cash out of the banking system in a single week. If the market were pricing an election that changes the fiscal path, the long end would be charging for it. It isn't. It is charging for nothing happening, and charging very little for that. So the question is not who wins. The question is whether a vote can still move a price at all — and the answer this tape gives is uncomfortable for everyone who plans to vote.
Closing time, and nobody is arguing about the economy
The pub is loud and the tape is quiet, which is the wrong way round for a week with an election in it. On the screen behind the bar: the 30yr at 5.29%, down 14 basis points on the week. The 20yr at 5.32%, down 16. The 10yr at 4.96%, down 14. The 5yr at 4.78%. The 3-month bill at 4.22%. Every tenor lower, which is what a market does when it stops worrying about supply and starts worrying about growth.
And then the number nobody at the bar looks at. The gap between the 30yr and the 10yr closed at 33.0bp on 19 Oct. On 2 Jan it was 67.0bp. Out of 200 sessions this year it has closed lower on 4. A curve that is supposed to price thirty years of fiscal risk is pricing the next six months instead, and it is doing so at the flattest end of its own distribution.
The Blazer has the election coverage on his phone and a pint he has not touched. He is not watching the polls. He is watching the spread, and he has decided what it means.
The Blazer says both parties are running on a programme neither intends to fund
The Blazer: Look at that. Four times this year the market has been willing to pay less for thirty years than for ten, and it is nearly a fifth of the way down from January. You know what that is? That is the market telling you it has read both manifestos and priced neither.
The Second Opinion: That is one reading. The other is that the market is pricing a growth scare and the election is a coincidence.
The Blazer: A growth scare does not flatten the long end while the government is borrowing at this pace. A growth scare steepens it, because the central bank cuts and the long end stays put. What flattens it is the market deciding the fiscal path is already set, that whoever wins will run the same deficit, and that the only variable is which party gets blamed for it in two years. Both of them are running on a programme neither intends to fund. Vote for whoever. The auction calendar is the only manifesto that gets implemented.
He puts the phone face down, which is his way of making a point.
The Second Opinion brings the auction calendar and refuses the easy version
The Second Opinion: I will give you the deficit is bipartisan. I will not give you the vote does not matter. There is one scenario where this election prices, and it is not the one either party is campaigning on. It is a unified government that actually passes something — a tax bill, a spending package, anything with a number in it. Then the term premium moves, the long end steepens, and your gridlock trade is the one that breaks. You are not long a view. You are long the absence of a view.
The Blazer: Then show me the last time a vote moved the 30yr more than a Fed meeting did.
The Second Opinion: 2022. The UK. A budget with no costing behind it moved the gilt curve 100 basis points in a week and took the pension system with it. That is the counter-example, and it is the only one that matters, because it proves the market will price a fiscal decision the moment someone actually makes one. The reason it is not pricing this election is not that elections cannot move bonds. It is that neither side has offered it anything to price.
He turns the laptop round, which he has been waiting all evening to do, and shows them the term premium against the 30s10s. The two lines have been telling the same story since the summer.
The Third Man says the interest bill decides the deficit, and nobody votes on the interest bill
The Third Man has not said anything for twenty minutes, which usually means he is about to.
The Third Man: You are both arguing about which party can change the deficit. Neither can. The discretionary budget is not where the money goes. The number that decides the deficit now is the interest bill, and the interest bill is set by the stock of debt and the curve. The curve is set by the central bank. So the midterms are a referendum on the Fed, and the Fed is not on the ballot.
The Second Opinion: That is a bit neat.
The Third Man: It is arithmetic. You can vote to spend more or spend less. You cannot vote on the coupon. And this week the coupon did something that should worry anyone who thinks the vote is the lever: the Treasury took $59bn out of the banking system in seven days and the price of overnight money did not move. SOFR cleared 3.88 against a 3.90 administered rate. No scarcity premium. No funding squeeze. The sovereign moved the largest cash call of this sequence by announcement, and the market absorbed it without a flicker. That is not a market pricing political risk. That is a market that has stopped charging for it.
The Blazer: So what does it charge for?
The Third Man: Gold fell $143.20 this week and Bitcoin was flat. Ask yourself which of those two is pricing the fiscal story and which one is pricing the liquidity story. Then ask which one you would rather own if the answer is neither.
The four closes that make the case
Here is the year, one line. The spread opens at 67.0bp on 2 Jan, works its way down through the spring, and closes at 33.0bp on 19 Oct. The full range is 29 to 68. The four circled sessions — 23 and 24 September at 29.0, 25 and 28 September at 32.0 — are the only ones that closed below where it sits now.
Read the shape rather than the level. This is not a spike and it is not a crash; it is a grind lower that has survived every attempt to steepen it, including a week in which the government borrowed heavily and the central bank did nothing. A market that wanted to charge for fiscal risk had an open goal this week and declined to shoot.
The counter-reading is honest and should be stated: a flat 30s10s can also mean the market expects the policy rate to fall, which compresses the whole curve from the front. But that reading requires the long end to fall less than the front end, and this week every tenor fell between 6 and 16 basis points with the 20yr falling most. That is not a front-end story. That is the middle of the curve carrying the repricing.
The transfer nobody charged for
The second number is the one that would have caused a scene in any other decade. The Treasury's general account closed at $945.2bn on 21 October, from $885.8bn a week earlier — a build of $59.4bn. Reserve balances fell $59.5bn over the same seven days to $2962.6bn. The two match within $0.2bn, with $0.4bn parked at the reverse repo facility.
A transfer, not a drain. The money did not leave the system; it changed hands inside it, from the banking system's balance at the Fed to the Treasury's balance at the Fed. In 2019 a move a fraction of this size took repo to 10%. This week secured overnight money cleared 2 basis points under the administered rate and the 3-month bill rallied 6 basis points. The plumbing has been rebuilt well enough that the sovereign can confiscate $59bn of bank reserves by announcement and nobody charges for it.
Hold both numbers at once and the argument writes itself: the market will not price a fiscal decision it has not been given, and it will not price a liquidity shock it has been trained not to fear. What is left to price is the interest bill. Which is the one thing on the ballot that nobody is voting on.
Where this breaks
First, the curve. A 30s10s back above 68bp — the January level — would mean the market has found something to price. That is the clean invalidation of the whole argument: it would say the vote does matter and the long end is charging for it.
Second, the funding. SOFR or EFFR at or above 3.90% returns the administered rate to the conversation and means the next cash call does have a price. Watch the reserve print back above $3,022.1bn with the account still above $945.2bn: that would say the drain was not the mirror it looks like.
Third, the politics. A unified government that passes a fiscal package with a costing behind it breaks the gridlock trade immediately, and it breaks it in the direction that hurts: higher term premium, steeper long end, and every position built on 'nothing happens' unwound at once.
Until one of those three prints, the honest description of this market is that it has voted already, it voted for nothing to happen, and it is charging almost nothing for the privilege. That is not a forecast. It is the price.