Notes for Tue, Apr 28, 2026

No fresh prints, no Tier 1 updates; priors carry. Stay short SOFR/EFFR basis, lean front-end higher on Fedspeak tilt.

Overnight take

Nothing forces a re-baseline. No new observations, no Tier 1 desk updates, and the commentary that landed overnight is cycle-narrative material (Apollo PE/secondaries/software, Noahpinion AI/industrial policy, BlackRock mega-forces) — not a rates input. File and move.

The two live theses both carry: (1) funding squeeze fades as TGA bleeds reserves; (2) Fedspeak channel has tilted hawkish ahead of the FOMC blackout. Yesterday's close stands as today's open.

Funding — priors carry

Last marks: SOFR 3.66% on 04-24 sits +2bp through EFFR at 3.64%^[2][1]; GCF UST 3.687% / MBS 3.709% on 04-24^[3] confirms collateral bid but no escalation. SOFR is 9bp above the 04-09 trough of 3.57%^[4] — squeeze residue, not regime. TGA $970B on 04-23 is $68B off the $1,038B peak^[5]; the mechanical drain keeps feeding reserves and compressing the basis.

Trade: Stay short SOFR vs EFFR basis. No reason to take it off without a print that contradicts the drain.

Front end — asymmetry still skewed up

Per Amarnath's 4/20 monitor, three regional presidents (Goolsbee, Hammack, Musalem) cracked the door to hikes^[6]; his April FOMC preview reads the Committee as internally more hawkish than the dot path. Nothing overnight argues against that. Romanchuk's Middle East de-escalation read is the offsetting factor — orderly energy is the only thing keeping the hawks from getting louder.

Trade: Don't be long the front end on cut hopes alone. Lean 2Y drifts higher over the month-ahead window as blackout-window communication and the next CPI/NFP test the rhetoric. Conviction medium — data still has to validate.

Noise filter

Apollo (secondaries, software reset, gold, $20s vs $100s, 3% vs 60% private credit), Noahpinion (AI, industrial policy, consciousness), BlackRock mega-forces, Atlanta Fed shutdown survey-quality piece — none of it is a rates input today. File.

Citations

1. [1]SOFR 3.66% on 04-24 sits +2bp through EFFR at 3.64% (3.66%) — NY Fed SOFR · Apr 24, 2026
2. [2]EFFR at 3.64% (3.64%) — NY Fed EFFR · Apr 24, 2026
3. [3]GCF UST 3.687% / MBS 3.709% on 04-24 (UST 3.687% / MBS 3.709%) — DTCC GCF repo · Apr 24, 2026
4. [4]SOFR is 9bp above the 04-09 trough of 3.57% (3.57%) — NY Fed SOFR · Apr 9, 2026
5. [5]TGA $970B on 04-23 is $68B off the $1,038B peak ($970B) — Treasury General Account · Apr 23, 2026
6. [6]three regional presidents (Goolsbee, Hammack, Musalem) cracked the door to hikes (Goolsbee, Hammack and Musalem all expressed some openness to the idea of rate hikes this year) — Commentary · employamerica.org

Generated by Short Rates Desk. Informational only. Not investment advice.

Plumbing thesis pays out as TGA bleeds and SOFR holds +2bp; Skanda's Warsh piece adds hawkish-vol risk to the front end.

Funding — basis trade pays its rent

Today's prints validate the plumbing call. SOFR printed 3.66% on 04-27^[1], +2bp through EFFR's 3.64%^[2] — basis intact, no escalation. GCF UST 3.689% / MBS 3.712%^[3] on the same tape confirms collateral is well-bid but not stressed. Reserves took the hit: WRESBAL dropped to $2,902B on 04-22^[4] from $3,130B on 04-15^[5], a $228B drain that maps directly onto Treasury's run from $703B on 04-10^[6] to the $1,038B peak on 04-20^[7]. RRP is empty at $0.36B on 04-27^[8] — no soft tissue left, every dollar Treasury releases hits reserves directly.

Now the cycle reverses. TGA closed at $974B on 04-24^[9], $64B off peak, and the mechanical refill of reserves is the next leg of basis compression. Stay short SOFR vs EFFR basis. Risks the view: month-end on 04-30 — if SOFR doesn't print 3.66% or lower into Friday despite the reserve refill, the squeeze residue is structural, not transient, and the thesis breaks. Pipeline note: TGA closes for 04-25/04-27 not yet filed in our system; Treasury Daily Statement drops 4 PM ET with a one-business-day lag — re-check before tomorrow's open.

Front end — Warsh wildcard layered on hawkish drift

Skanda's piece today flags methodological risk in Warsh's Trimmed Mean PCE pivot^[10] — the Dallas Fed measure aggressively trims 55% of the basket^[11], leaving the dovish signal hostage to a narrow window. If a confirmed Warsh abandons it once a more robust 16% symmetric trim shows acceleration toward core, the front end has to price hawkish-volatility risk on top of last week's drift (Goolsbee/Hammack/Musalem cracking the door to hikes^[12]).

Alternate read considered: DGS2 closed at 3.78% on 04-24^[13], 5bp lower on the day — could read as the front-end discounting cuts despite Fedspeak. Weighting against: EFFR/SOFR bedrock unchanged, 10Y/2Y steepened to 0.57bp on 04-27^[14] (a touch of bull steepening, not a flight-to-cuts signal), and the asymmetry case is about the next CPI/NFP, not Friday's tape. Lean 2Y higher into the May data window. Conviction medium — needs validation.

Assumes Warsh confirmation proceeds and the trimmed-mean methodology debate plays out in public; if confirmation stalls or he sidesteps the question, the wildcard goes back in the deck and the call is just hawkish-drift carry.

Noise filter — material flags only

Two pieces worth surfacing, neither a rates input today:

• Nelson on the Fed's GMS stress-test methodology: small percentile shifts (95th vs 99th percentile shocks produce massive spread moves^[15]) — capital implications for dealer banks, worth tracking through the stress-test cycle but not a desk action.
• Concoda on green-energy credit reality ($140B EV losses, ~125 solar bankruptcies^[16]) — desks carrying green-book exposure should reassess for policy-reversal whipsaw; not a rates trigger.

Everything else — Apollo cycle pieces, ten St Louis Fed structural notes, NY Fed BHC primer, Romanchuk on Canadian gasoline, Noahpinion on AI/consciousness — file. None are rates inputs today.

Citations

1. [1]SOFR printed 3.66% on 04-27 (SOFR 2026-04-27: 3.66000%) — NY Fed SOFR · Apr 27, 2026
2. [2]EFFR's 3.64% (EFFR 2026-04-24: 3.64000%) — NY Fed EFFR · Apr 24, 2026
3. [3]GCF UST 3.689% / MBS 3.712% (2026-04-27: MBS 3.71200 / UST 3.68900) — DTCC GCF repo · Apr 27, 2026
4. [4]WRESBAL dropped to $2,902B on 04-22 (WRESBAL 2026-04-22: 2901825 ($M)) — FRED WRESBAL · Apr 22, 2026
5. [5]$3,130B on 04-15 (WRESBAL 2026-04-15: 3129588 ($M)) — FRED WRESBAL · Apr 15, 2026
6. [6]$703B on 04-10 (TGA 2026-04-10: $703218M closing) — Treasury General Account · Apr 10, 2026
7. [7]the $1,038B peak on 04-20 (TGA 2026-04-20: $1038035M closing) — Treasury General Account · Apr 20, 2026
8. [8]RRP is empty at $0.36B on 04-27 (RRPONTSYD 2026-04-27: 0.363 ($B)) — FRED RRPONTSYD · Apr 27, 2026
9. [9]TGA closed at $974B on 04-24 (TGA 2026-04-24: $974251M closing) — Treasury General Account · Apr 24, 2026
10. [10]Skanda's piece today flags methodological risk in Warsh's Trimmed Mean PCE pivot (Warsh's pivot to Dallas Fed Trimmed Mean PCE looks like goalpost-shifting: the measure aggressively trims 55% of the basket) — Commentary · employamerica.org
11. [11]the Dallas Fed measure aggressively trims 55% of the basket (the measure aggressively trims 55% of the basket, making it hostage to a narrow window of components) — Commentary · employamerica.org
12. [12]Goolsbee/Hammack/Musalem cracking the door to hikes (Goolsbee, Hammack and Musalem all expressed some openness to the idea of rate hikes this year.) — Commentary · employamerica.org
13. [13]DGS2 closed at 3.78% on 04-24 (DGS2 2026-04-24: 3.78) — FRED DGS2 · Apr 24, 2026
14. [14]10Y/2Y steepened to 0.57bp on 04-27 (T10Y2Y 2026-04-27: 0.57) — FRED T10Y2Y · Apr 27, 2026
15. [15]95th vs 99th percentile shocks produce massive spread moves (Nelson shows tiny percentile shifts (95th vs 99th) in primary shocks produce massive downstream spread shocks) — Commentary · bpi.com
16. [16]$140B EV losses, ~125 solar bankruptcies ($140B cumulative EV losses since 2022, ~125 US solar bankruptcies in 3 years) — Commentary · bpi.com

Generated by Short Rates Desk. Informational only. Not investment advice.