← Contrarian Codex hub  ·  Research hub  ·  Research library

Contrarian Codex — Macro update: the Fed's fiscal backdrop

"The market wanted Warsh to say who controls the long end. He told it nobody at the Fed does." — a Treasury running the long end with a $4bn hose and a checking account, a Chair who will manage only the short rate, and 65 straight months above target.
2026-AUG-28 · Contrarian Codex · macro update (the Fed's fiscal backdrop) · ~6-page written report · read ↗ PDF · post ↗ · actionable insights
One-line take: a Jackson Hole write-up that is really an argument about who now sets the long end. (1) The buyback rescue is a rounding error. The 30-year has come down ~14bp since he last wrote and trades near 5.17% (10-year ~4.64% after tagging 4.75%, its highest since early 2025) — but "the deficit did not improve, no new buyer appeared for 30-year paper, inflation did not roll over either." Treasury said it would at least double liquidity-support buybacks in the 10-20yr and 20-30yr sectors, from $2bn per operation to at least $4bn, effective early September; the 10-year closed down almost 6bp and the 30-year 9 on the news, then inside two sessions the 30-year was back above 5.27% and the whole move handed back, "because $4 billion an operation against $2.1 trillion of deficit financing is a rounding error dressed up as a rescue." Worse, the purchases may be funded straight out of the ~$950bn General Account; Bessent calls it a "Treasury Twist," which is "wonderful branding for something that historically required a central bank to execute" — and he can bolster the TGA with higher gold prices too. (2) Positioning has flipped, the structure hasn't. Swap spreads narrowed to their tightest since February (hedge-fund positioning in that trade ~$305bn vs under $50bn in 2022), long-bond call skew has run up against puts while short maturities sit neutral — "the fear has inverted," from yields grinding higher without end to being caught short into an intervention — and a large liquidity provider's trend model has ultra-long duration at a −2.28 z-score, the most stretched short since 2024, with 64 analogs since 2003 showing yields lower 71% of the time over 120 days (avg −25bp). His verdict: "a crowded short describes who owns what today and says nothing about who funds $2.1 trillion next year… the structural problem is untouched either way." (3) The term-premium thesis picks up a second engine. IG issuance is near $1.5tn YTD, up ~36% and on pace to eclipse 2020; the biggest tech borrowers alone are ~$200bn of it — about 25% of Treasury's net note-and-bond issuance to private investors and ~5x last year — with a single multi-tranche deal near $53bn and somebody selling a century bond "into this"; hyperscaler capex near $800bn this year and above $1tn from 2027, AI-related issuance estimated $300–570bn, and one estimate puts ~30bp of this year's 10-year move on corporate and mortgage supply. (4) Fiscal dominance, in the plumbing. Treasury bought yen alongside Japan's MoF (first joint operation since 1998) funded by selling euros so nobody had to liquidate Treasuries, then Bessent publicly asked the Fed to upsize the $60bn-per-counterparty FIMA repo line — which needs an FOMC vote — with Japan holding ~$1.1tn of USTs and its intervention estimated at $60–80bn. "Every element of that architecture is engineered around one objective, making sure nobody has to sell a Treasury… Does that sound like an independent central bank to you?" (5) The inflation stack is not cooperating: core PCE 0.2%/3.3%, headline 0.2%/3.7% — the 65th consecutive month above target — policy parked at 3.50–3.75% for five meetings, three regional presidents dissenting to hike in July, cuts not on the menu, and July's energy component −1.5% so the crude move lands in August/September instead. Layer on 50% tariffs live on $20–28bn of Canadian goods with Canada matching dollar-for-dollar in September (Carney using the word war out loud; the auto concession narrowed to light vehicles, putting a GM and a Ford plant in Ontario in the firing line, and the Canadians walked), and secondary-sanction threats on Iran that run through Chinese firms and banks weeks before a Trump-Xi meeting. (6) Hormuz optionality is being sold too cheaply. The IRGC claims a revenue-sharing deal with Oman; both foreign ministries describe only an interim transit framework; Iran and the western naval group have designated different corridors; Trump says the mines are gone and 10m barrels transited in a day while the IMO head says the strait is not open and the mines are unconfirmed; the EIA sees Middle East output below pre-conflict levels until early 2027. "The risk premium has deflated on a framework nobody has signed, that would not reopen the waterway by itself, and that depends on concessions nobody has agreed to make." (7) Jackson Hole: Warsh gave no timetable, no forward guidance and no reaction function; the Fed will "have work to do" if it cannot be confident inflation is heading to 2%; price stability is the "predominant focus." The 2-year sold off 7–9bp to ~4.30% and September hike odds went from ~35% to 46% to above 50% inside the hour — while the 30-year finished roughly flat near 5.16% and the 10-year barely budged around 4.67%. "A hawkish Chair delivered a hawkish speech… and the long end shrugged at him. The market has fully separated the policy rate from the term premium." Not one word about the buybacks, FIMA, or the long end — asked implicitly whether the Fed would participate in Treasury's operation, "he answered by declining to acknowledge that the operation exists." Meanwhile he calls financial conditions hard to describe as restrictive and reaches for AI productivity optimism — "the grow-our-way-out-of-it argument… now coming out of the Fed" — while Bessent has skewed issuance hard to bills, so a September hike is "mechanically expensive for a Treasury that has parked itself there on purpose": a Chair "boxed in by an issuer who needs him not to move." (8) Positioning: three weeks that hand over a near coin-flip rate decision (52% September, 56% October), 50% two-way Canadian tariffs, an unsigned Hormuz framework and a Treasury managing the long end alone — any one of which can produce a violent 2–3 day move — so the 10–20% cash position stands, held as "optionality on somebody else's forced selling," letting the gold core stay untouched when the front end reprices and metals go heavy, "which tends to be the exact moment people sell the thing they should be adding to." Remember your sleeping level; reassess in three weeks.

Key points

Why another macro update — and what it is not

The long end: lower is lower, and that is all it is

The buyback: $4bn an operation against $2.1tn of financing

The "Treasury Twist" — funded out of the General Account

Swap spreads — a light backstop gets priced in

The fear has inverted

The crowded short — and why it settles nothing

Corporate supply — the term premium's second engine

The yen operation — funded by selling euros

FIMA — asking the Fed to resize its balance sheet for an allied government

65 months above target

Canada — 50% tariffs in both directions

Iran sanctions run through Chinese banks

Hormuz — two corridors on the same map

The deflated risk premium — optionality sold too cheaply

Warsh at Jackson Hole — hawkish, and deliberately empty

The price action — and the part that didn't move

What he refused to acknowledge

Both arms of policy pulling opposite ways

The issuer who needs him not to move

Positioning — the 10–20% cash stands, reassess in three weeks


Macro-only report — no securities carry an argued stance in this letter, so there is no stocks table (GM and Ford appear only as the location of two Ontario plants inside the tariff negotiation; officials, indices and the EIA/IMO are in the key points). Analysis distilled from the Contrarian Codex macro update "the Fed's fiscal backdrop" (PDF linked above). For personal study — not investment advice. Source material © Contrarian Codex / "Mart".