Short Rates Desk
Fri, Aug 07, 2026

Notes for Fri, Aug 07, 2026

Morning brief

· generated 07:06 ET

Keep the 2s10s steepener into payrolls, at half size

Front end rallied to 4.18% on soft labor prints while term premium keeps the long end cheap; today's NFP is binary, so we trim.

Payrolls

Our 2s10s steepener stays on into this morning's 8:30 ET payrolls, carried at half size. Both break tests on record held overnight: funding printed no stress and 2s10s held at 44bp[1], so we carry the stance rather than re-baseline. The front end rallied hardest, with the 2Y at 4.18%[2] versus 4.28% at the July close[3], and soft labor signals sit behind it: ADP came in at 44k[4] and the ISM services employment index fell to 47.4[5]. We attribute that bid to haven and duration flows through the tech selloff rather than to rate-cut conviction, since September cut pricing eased to 54% from 58%[6] over the same window. The long end stays cheap on term premium: the 10Y at 4.63%[7], the 30Y at 5.17%[8], and the Fed's 10-year term premium estimate at 0.87%[9], near its cycle high.

The print is binary, and that is why we trim. A hot number revives the July hike camp, the three dissenters at the 9-3 meeting (Logan, Hammack, Kashkari)[10], and flattens the curve as the front end sells off. A soft number steepens it. We would rather run a smaller steepener through the release than take a view on the number itself.

Funding

Funding is calm and the reserve drain we flagged has not bitten. SOFR's last fix was 3.64%[11], a basis point through the 3.65% IORB[12], with GCF Treasury repo at 3.67%[13]. RRP take-up sits at $1.4bn[14], effectively drained, but with SOFR under IORB it reads as spare cash leaving the facility rather than reserve scarcity. Our standing assumption was that Treasury holds its bill-issuance pace and the settlements drain reserves; reserves instead rose $8.8bn on the week[15] to $2.99tn[16] while the TGA fell $41bn[17] to $929bn[18], the two roughly offsetting. We re-baseline the funding read only if next week's settlements push SOFR firmly above IORB.

Risks

The Iran-Oman framework on the Strait of Hormuz is close but not clean: Iran's draft bans US-linked vessels[19], a non-starter for the White House if enforced. The rates impulse is two-sided. Enforcement that holds floors energy inflation higher and buys front-end premia; a collapse back to shooting pulls a duration bid. Neither is confirmed, so we hold the steepener into the print.

We're wrong on the steepener if payrolls print hot and the front end sells off faster than the long end, flattening 2s10s through the release. The trade is off if the Hormuz talks break to open conflict and the whole curve rallies on a haven bid. Initial claims at 199k[20], near a 57-year low[21], keep the labor backdrop benign into the number, which is why we lean to holding rather than cutting the steepener outright.

Sources read

8 sources read

  • Commentary items: 8

Citations

  1. [1]2s10s held at 44bp (0.44) — FRED T10Y2Y · Aug 6, 2026
  2. [2]the 2Y at 4.18% (4.18) — FRED DGS2 · Aug 5, 2026
  3. [3]4.28% at the July close (4.28) — FRED DGS2 · Jul 31, 2026
  4. [4]ADP came in at 44k (the ADP report showing employment growth of 44k in July (vs 65k expected)) — Commentary · zerohedge.com
  5. [5]the ISM services employment index fell to 47.4 (the employment component fell to 47.4 in July (vs 51.2 expected)) — Commentary · zerohedge.com
  6. [6]September cut pricing eased to 54% from 58% (pricing of a September Fed rate cut eased from 58% to 54%) — Commentary · zerohedge.com
  7. [7]the 10Y at 4.63% (4.63) — FRED DGS10 · Aug 5, 2026
  8. [8]the 30Y at 5.17% (5.17) — FRED DGS30 · Aug 5, 2026
  9. [9]the Fed's 10-year term premium estimate at 0.87% (0.86810) — FRED THREEFYTP10 · Jul 31, 2026
  10. [10]the three dissenters at the 9-3 meeting (Logan, Hammack, Kashkari) (9-3 no hike; three dissenting votes for hikes came from Lorie Logan, Beth Hammack, and Neel Kashkari) — Commentary · peterboockvar.substack.com
  11. [11]SOFR's last fix was 3.64% (3.64) — NY Fed SOFR · Aug 5, 2026
  12. [12]the 3.65% IORB (3.65) — FRED IORB · Aug 7, 2026
  13. [13]GCF Treasury repo at 3.67% (UST 3.67000 / MBS 3.68400) — DTCC GCF repo · Aug 6, 2026
  14. [14]RRP take-up sits at $1.4bn (1.42900) — FRED RRPONTSYD · Aug 6, 2026
  15. [15]rose $8.8bn on the week (+8,779M w/w) — FRED WRESBAL · Aug 5, 2026
  16. [16]$2.99tn (2,993,349M) — FRED WRESBAL · Aug 5, 2026
  17. [17]the TGA fell $41bn (-41,117M w/w) — Treasury General Account · Aug 5, 2026
  18. [18]$929bn (929,325M) — Treasury General Account · Aug 5, 2026
  19. [19]Iran's draft bans US-linked vessels (Iran's draft bans US-linked vessels, a non-starter for White House if enforced) — Commentary · zerohedge.com
  20. [20]Initial claims at 199k (Initial jobless claims subdued at 199k, slightly below the 205k estimate) — Commentary · peterboockvar.substack.com
  21. [21]near a 57-year low (basically hovering at its lowest since 1969) — Commentary · zerohedge.com

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

Close brief

· generated 16:24 ET

Payrolls miss guts the September hike case; we hold the 2s10s steepener

July payrolls fell 23k versus an 80k median; front-end hike pricing collapsed, and we carry the 2s10s steepener into next week's CPI.

Payrolls

Establishment payrolls fell 23,000 in July against an 80,000 median, a five-sigma miss[1], and the prior two months were revised down a combined 103,000[2]. Average hourly earnings rose 0.1% on the month and 3.2% year over year[3], both under estimate, and that wage roll is what decides the call for us: a hike needs wage pressure, and there is none. The unemployment rate ticked to 4.1%, but it fell because participation dropped to 61.4% from 61.8% over two months[4], not because the labor market tightened. Ahead of the print, futures held roughly 60% odds of a September quarter-point hike[5]; the miss erased that, and the front end now leans toward cuts by year-end[6].

Positioning

We carry the 2s10s steepener at half size. The break test on record was that the trade comes off if the Hormuz talks break to open conflict and the whole curve rallies on a haven bid. That did not happen: the Hormuz reopening narrative tightened around Bessent's 30-to-60-day ceasefire timeline[7], and Brent fell about 6.9% on the week[8], so there was no haven bid and the test held. The payrolls miss pulled the front end lower and steepened 2s10s in our favor. We keep it at half size rather than press: Warsh's September hike trial balloon has only lost traction[9], and if CPI runs hot next week, Warsh keeps a September move live and the front end sells back off.

Counter-case

The bear case is that the number is noise rather than signal. The unemployment drop is a participation artifact, the household survey lost 87k and the birth-death model added 235k[10], so part of the establishment weakness could reverse. Amarnath reads the miss as mostly seasonal government and education churn[11] that does not shake the Fed's September conviction. We weigh that and still land on the dovish side: minus 23k, another 103k of downward revisions, and 0.1% wage growth is too much to fade on an optical unemployment tick, and the wage roll is not a seasonal story. We are wrong if CPI prints hot next week and the front end reverses; that is where we cut the steepener rather than add to it.

Funding

Front-end plumbing stayed quiet, which is what we want with the day's action in duration. SOFR fixed at 3.65% for August 6[12], level with IORB at 3.65%[13], and EFFR held at 3.63%[14], so the floor is intact. Overnight RRP take-up was $1.4bn on August 6[15], a rounding error against reserves. GCF Treasury repo printed 3.67%[16]. One issue, the February 18 2027 bill, cleared the SOMA lending program at a 29.4bp fee versus the 5bp minimum[17], which is issue-specific demand and nothing systemic. Coupon supply lands next week, a $58bn 3-year Tuesday and a $42bn 10-year Wednesday[18], which front-loads near-term funding needs without threatening the floor. Our standing assumption that reserves stay ample as Treasury issues carries; nothing in today's fixings challenges it.

Sources read

8 sources read

  • Commentary items: 8

Citations

  1. [1]Establishment payrolls fell 23,000 in July against an 80,000 median, a five-sigma miss (Nonfarm payroll employment declined by 23,000 in July, missing the 80,000 median estimate by over 100,000 (characterized as a 5-sigma miss).) — Commentary · zerohedge.com
  2. [2]the prior two months were revised down a combined 103,000 (totaling 103,000 lower than previously reported.) — Commentary · zerohedge.com
  3. [3]Average hourly earnings rose 0.1% on the month and 3.2% year over year (Average hourly earnings rose only 0.1% month-over-month, missing the 0.3% estimate and falling well short of the 3.5% year-over-year median estimate at 3.2%.) — Commentary · zerohedge.com
  4. [4]participation dropped to 61.4% from 61.8% over two months (Unemployment rate drop to 4.1% is pure participation collapse (61.8% → 61.4% in 2mo), not labor tightness) — Commentary · zerohedge.com
  5. [5]futures held roughly 60% odds of a September quarter-point hike (around 60% of a quarter-point rate hike priced in for Fed's September policy meeting) — Commentary · zerohedge.com
  6. [6]the front end now leans toward cuts by year-end (Front end now pricing in cut odds >70% by year-end) — Commentary · zerohedge.com
  7. [7]the Hormuz reopening narrative tightened around Bessent's 30-to-60-day ceasefire timeline (Hormuz reopening narrative tightens with Bessent's '30-60 day ceasefire' timeline) — Commentary · zerohedge.com
  8. [8]Brent fell about 6.9% on the week (Brent crude fell 0.7% to below $82 a barrel, bringing its decline this week to 6.9%) — Commentary · zerohedge.com
  9. [9]Warsh's September hike trial balloon has only lost traction (Warsh trial balloon on September hike loses traction post-NFP miss) — Commentary · zerohedge.com
  10. [10]the household survey lost 87k and the birth-death model added 235k (household survey loss of 87k ... Birth/death model alone adds 235k, dwarfing actual prints) — Commentary · zerohedge.com
  11. [11]Amarnath reads the miss as mostly seasonal government and education churn (mostly seasonal gov/edu churn and doesn't shake Fed confidence in a September hike, per Amarnath's read) — Commentary · employamerica.org
  12. [12]SOFR fixed at 3.65% for August 6 (3.65000%) — NY Fed SOFR · Aug 6, 2026
  13. [13]IORB at 3.65% (3.65) — FRED IORB · Aug 7, 2026
  14. [14]EFFR held at 3.63% (3.63000%) — NY Fed EFFR · Aug 6, 2026
  15. [15]Overnight RRP take-up was $1.4bn on August 6 (1.42900) — FRED RRPONTSYD · Aug 6, 2026
  16. [16]GCF Treasury repo printed 3.67% (UST 3.67000) — DTCC GCF repo · Aug 6, 2026
  17. [17]cleared the SOMA lending program at a 29.4bp fee versus the 5bp minimum (912797TV9 (B 02/18/27) at 29.4 bp) — Observation · observation:seclend_observations:912797TV9:2026-08-07
  18. [18]a $58bn 3-year Tuesday and a $42bn 10-year Wednesday (Treasury supply calendar (including $58bln 3y Tuesday, $42bln 10y Wednesday)) — Commentary · conks.plumbing

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