Autoresearch: Warsh Jackson Hole Friday + real-yield (not inflation) decomposition of the long end

Generated by /autoresearch on 2026-08-25. Step 1 refresh of warsh-higher-for-longer-to-brokerage-nii-rerate plus step 2 bucket #9 (financials / rate regime). Treat as raw material. Context: vault/projects/stock-market Prompt: What is new since 2026-08-21 on Warsh higher-for-longer, 10-year/30-year real yields, and the SCHW/IBKR NII overhang? Specifically: is the long-end move inflation or real rates, and what is the next dated catalyst?

Summary

The long-end selloff that nick-colas is walking through on The Compound (30-year at 15–20 year highs; inflation expectations "dead flat") is the same decomposition independent sources now print as a year-to-date fact: real rates did the work, not inflation compensation. 10-year TIPS real yield 2.40% on 2026-08-21 (FRED DFII10, updated Aug 24). Cleveland Fed: 10-year real rate +53 bp YTD vs expected inflation +16 bp; 10-year TIPS +44 bp while the 10-year breakeven fell 6 bp. Fed minutes (June and July) already said higher nominal yields "reflected higher real rates."

The dated catalyst is Chair Warsh's first Jackson Hole keynote, Friday 2026-08-28 (Kansas City Fed symposium Aug 27–29; theme "Financial Innovation: Implications for Payments and Policy"). This is a reaction-function / communication event, not a rate decision. Do not graduate warsh-higher-for-longer-to-brokerage-nii-rerate off a speech preview. The load-bearing fork is still whether higher-for-longer survives the next inflation print and the Sept 16 FOMC — Jackson Hole can move the term premium without resolving the SCHW sweep / IBKR 25bp-sensitivity arithmetic.

A second, under-discussed collision: Treasury doubled long-end buybacks (Aug 19; Bessent "could be more than $4 billion per issue") while a Warsh task force is examining a shorter Fed portfolio — Operation Twist run backwards against the buyback window. Not priced as a two-principal trade, per Piedmont Crescent (Aug 23).

Findings

Real yields, not inflation, moved the long end

• 10-year TIPS (DFII10): 2.40% on 2026-08-21, 2.35% on 08-19/08-20, 2.44% on 08-17. FRED updated Aug 24 3:16 PM CDT; next release Aug 25 (FRED DFII10).
• Cleveland Fed decomposition through August: expected-inflation component 2.33% → 2.49% (+16 bp); real interest rate 1.67% → 2.20% (+53 bp). Market TIPS: 10-year real +44 bp YTD, 10-year breakeven −6 bp. Cleveland inflation risk premium unchanged at 44 bp. "This year's rise in long-term interest rates has essentially nothing to do with expected inflation" (Piedmont Crescent Capital, "The Speech and the Curve," 2026-08-23).
• FOMC staff, June and July minutes: higher nominal yields "reflected higher real rates" while longer-term inflation compensation "remained stable and consistent with the Committee's 2 percent longer-run inflation objective" (same source, citing the minutes).
• 30-year: closed at or above 5% for 34 consecutive sessions and touched 5.34% (last seen June 2007) in the week of Aug 18 (same). Arise WM: 30-year above 5.3%, "highest since 2007"; 10-year "highest since January 2025" (Arise, 2026-08-21).
• Three claimants on the real-rate / savings-supply side (Piedmont): federal borrowing ~$2T/year; energy shock; AI hyperscaler debt $220B through Aug 10 vs $12.5B in the same period last year. Neither the funds rate nor a buyback is a tool for that.

This independently corroborates Colas's Compound claim that the 30-year breakout is a real-yield residual with inflation expectations "dead flat" for 15 years. It does not by itself confirm the SCHW/IBKR NII-haircut-still-embedded test.

Jackson Hole Friday 08-28 is the dated catalyst

• Kansas City Fed: symposium Aug 27–29, topic "Financial Innovation: Implications for Payments and Policy" (KC Fed symposium page).
• Warsh keynote: Friday Aug 28. Slot reported as 10:00 a.m. ET by MNI/Fed confirmation recaps; some retail guides say 8:00 a.m. ET — treat time as Friday morning, confirm on the KC Fed agenda Aug 26. Nineteen days before the Sept 16 FOMC.
• What the speech is for: Warsh has withdrawn forward guidance on purpose (no June dot; June statement cut to 130 words from 341; July presser "learning to play the ball, not the referee"). Bessent moved the other way: doubled long-end buybacks Wednesday Aug 19, "could be more than the $4 billion per issue," "yields don't reflect the underlying fundamentals" (Piedmont; Treasury buyback already ingested as 2026-08-21-autoresearch / treasury-buyback-twist-to-hard-asset-debasement Step 1 confirmed).
• Five listen-fors (Piedmont, not a prediction): (1) reaction function as distinct from guidance; (2) maturity composition of the SOMA portfolio (Twist-backwards vs Treasury buybacks); (3) whether he frames the long-end as monetary or fiscal; (4) meeting calendar 8→6; (5) inflation-target / measurement. Score the speech on MOVE >70 through Sept 16 and on forecast dispersion, not on the 30-year print the day of.

What this does to the brokerage-NII hypothesis

warsh-higher-for-longer-to-brokerage-nii-rerate stays hypothesis. The forcing function (no-cut / higher real rates) is stronger on this tape — 10y real 2.40%, 30y 5.3%+, minutes on-record that the move is real rates. The conversion tests are unchanged and still unmet:

• A sell-side NII model that still embeds 2026 cuts (the haircut to remove) — not fetched this pass.
• SCHW sweep-balance / cash-sorting vs IBKR 25bp sensitivity — last resolved at Q2 as lending-mix self-help, not rate-beta; no new print this week.
• Does higher-for-longer survive Jackson Hole + Sept 16, or does a reaction-function clarification bid the long end and put the cut-haircut back in? That's the new dated fork. Do not graduate on Friday's headline.

Piedmont's own call (for scoring, not adoption): hold Sept 16, no cut in 2026, first hike H1 2027, 30-year 5.00–5.40% through September. Arise WM's opposing read: additional upside in yields "limited," room for lower rates as growth tailwinds fade. The hypothesis is still two-sided.

Open questions

• Does Warsh describe a reaction function on Friday, or keep the "play the ball" refusal? MOVE >70 through Sept 16 is the clean test of whether missing guidance is still costing duration.
• Does he say anything about SOMA maturity that would put the Fed in the long end against Treasury buybacks?
• Is a 2026-cut NII haircut still in SCHW/IBKR consensus models, or did Q2 already remove it?

Provenance

• Round 1: search 10-year Treasury yield real yields Warsh Fed August 2026 brokerage NII; search Jackson Hole 2026 Warsh speech date August.
• Fetched: FRED DFII10; Piedmont Crescent 2026-08-23 (primary decomposition + speech preview); Arise WM 2026-08-21; Empower "Between the margins" August 2026 (YTD real vs breakeven table); KC Fed symposium page; XTB / Money.it / Gate News / Regards of Wallstreet for the Friday slot.
• Early exit after round 2: the conversion tests (sell-side NII models, SCHW sweep) were not in the open web this pass; no round 3 pad.
• Priors: skipped (headless).
• Podcast color (not fetched as a web page; same-day transcript): nick-colas on The Compound 2026-08-24 — 30-year decomposed into flat inflation expectations vs rising real residual; hyperscaler IG issuance as a competing bid for duration. Promote that transcript separately; this synthesis does not quote it as a web source.