Fed & Macro 2026-09-07 中文 PDF

NIGHTLY INTELLIGENCE BRIEF

〔Overnight Brief〕162k August Payrolls Flip September Hike Odds to ~60% and Push 2Y to ~4.42%, While 30Y Holds 5.27% on Term Premium, Not the Fed - Quiet Period and CPI Decide

The August nonfarm payrolls print landed at 162k versus a ~56k median expectation, with unemployment steady at 4.1% and average hourly earnings at +0.3% m/m and +3.1% y/y, and within an hour the S&P sold off, gold fell more than 1%, bitcoin slid from roughly $82k to the $79k range, and the 2-year Treasury yield jumped to about 4.42% [1]. CME Fed Funds futures lifted the implied September hike probability from around 49% to about 60% [1][2]. The Fed has now entered its pre-meeting quiet period ahead of the Sept. 16 FOMC [4], yet the long end refused to follow the front end down - the 30-year sits near 5.27% and the 10-year touched 4.79% [3][2]. Treasury Secretary Bessent has expanded liquidity buybacks on 10-30Y paper from $2B to $4B+ per operation, a move that briefly clipped the 30-year from 5.34% to 5.18% before half the move reversed [3]. With term premium driven by supply and fiscal, not policy, the next decisive test is next week's CPI, where an August oil rebound may lift the headline [2][5][6].

0. Weekly Arc

Stronger-than-expected August payrolls reset the policy debate just as the Fed entered its quiet period: CME-implied September hike odds climbed from about 49% to roughly 60%, and the 2-year yield tagged ~4.42% [1][2]. Yet the long end barely budged in the dovish direction - the 30-year is back near 5.27% and the 10-year touched 4.79% on Sept. 1, signaling that supply and term premium, not the policy rate, set the back end [3][2]. The setup: a hot jobs print, a quiet-period Fed, and a CPI print this week that markets are treating as the last swing factor before the Sept. 16 decision [4][5][6].

1. Policy Narrative

  • **[ESCALATED] Hawkish repricing from the data, not the Fed:** the 162k August payrolls print (vs. ~56k expected) is the proximate driver of the September hike repricing, with CME Fed Funds futures lifting the implied probability to ~60% within an hour of the release [1][2].
  • **[ONGOING] Fed in pre-meeting quiet period:** the Sept. 16 FOMC is now less than two weeks out, and the FOMC has formally entered its quiet window - the next guidance vector is the post-meeting statement and dot plot, not inter-meeting commentary [4].
  • **[NEW] Debate on whether a hike would even help the long end:** the 10-year decomposes into short real rate, inflation expectations, and term premium; with term premium sticky and fiscal-driven, neither hiking nor holding cleanly tames the back end - the "hike camp" and the "hold camp" disagree on which works [7].
  • **[NEW] Desk view - hike plus balance-sheet expansion:** a sell-side weekly notes that a hike-plus-QE combination is now a live tail, arguing it could rescue the US financial capital account and absorb AI and Treasury funding liquidity, with a preference for high-dividend, big-finance, and supply-constrained cyclicals [8].

2. Key Data and Market Read

  • **[NEW] August nonfarm payrolls, well above consensus:** +162k vs. ~56k median expectation; unemployment 4.1% (unchanged); average hourly earnings +0.3% m/m, +3.1% y/y [1]. The 12-month average monthly print is about 31k, so 162k is a clean upside outlier [1].
  • **[NEW] Cross-asset reaction within an hour:** major US equity indices lower, gold down more than 1%, bitcoin from roughly $82k toward the $79k range, 2-year Treasury yield sharply higher to about 4.42% [1]. Read: "good news is bad news" - stronger jobs collapse the dovish path [1].
  • **[NEW] Long end under separate pressure:** 30-year tagged 5.34% on Aug. 18 (highest since 2007), the 10-year hit 4.79% on Sept. 1, and the 30-year is back to 5.27% - moves that the Treasury's expanded buyback only partially relieved [3][2].
  • **[NEW] Fed operational schedule:** the Fed is set to buy $2.12 billion in Treasury bills next week - small in size, but a reminder that the SOMA desk is still active in the bill complex even as QT continues [9].

3. Long-End Mechanics

  • **[NEW] Buyback recalibration, not a policy easing:** Treasury Secretary Bessent raised the per-operation liquidity buyback ceiling on 10-20Y and 20-30Y nominal paper from $2B to more than $4B, effective Sept. 9 through Nov. 4, with Bessent signaling room to expand further [3]. The 30-year dropped to about 5.18% the same day and recovered roughly half the move the next session - the buyback capped a spike, it did not redirect the trend [3].
  • **[NEW] Why the long end is sticky:** term premium is the residual, anchored by Treasury supply, fiscal trajectory, and global liquidity, and is largely outside the Fed's reach on a single meeting; 10-year = short real rate + inflation expectations + term premium [7]. Translation: a Sept. 16 cut-or-hold decision moves the front end, the back end is a fiscal story [7].
  • **[ONGOING] The pre-FOMC window is volatility-prone at both ends:** Bloomberg flags that US bond investors are positioned for two-way swings across the maturity spectrum into the FOMC, with the long end and the front end carrying different catalysts [10].

4. What Decides Next

  • **[NEW] This week's CPI is the swing variable:** multiple outlets explicitly tie the September decision path to the upcoming inflation print - Fortune, FT's Market Questions, and IndexBox all name CPI (with PPI and Oracle earnings as secondary tests) as the decisive input [11][5][12][6].
  • **[NEW] Oil is a one-sided risk to the August CPI:** a sell-side weekly notes that August oil's staged rebound is likely to put upward marginal pressure on the August inflation print, complicating the "soft-data cooling, hard-data hot" narrative [2].
  • **[NEW] Calendar:** Fed quiet period in effect [4]; Treasury buyback expansion effective Sept. 9 [3]; FOMC decision Sept. 16 [13][14][15].

5. Contrarian and Tail Risks

  • **[NEW] The "history says" warning:** a recurring wire piece flags that across 36 years of Fed rate hikes the equity market has typically reacted poorly, at least initially - a tail-risk framing for the Sept. 16 base case [13][14][15][16][17][18].
  • **[NEW] Source-quality flag:** the 60% September hike probability is from one wire (CME Fed Funds futures via Chinese-language summaries); the 2-year at ~4.42%, 10-year at 4.79%, and 30-year at 5.27% are similarly reported in Chinese-language wires and should be cross-checked against Bloomberg/Reuters tape before being quoted as point figures [1][2].
  • **[NEW] Open contradiction:** the policy-direction debate is genuinely split - one desk argues hike-plus-QE is now a live scenario to fix the term premium, while the long-end analysis argues term premium is outside the Fed's reach and that even a successful hike would not durably compress 10s/30s [7][8]. Both views sit in the packet; treat as live, not resolved.

SOURCE TRAIL

Citations

18 records

  1. [1]
  2. [2]

    新浪财经 · 券商研报索引(vReport 宏观+策略)策略周报:外部加息扰动未消 短期以守为攻 ↗

    relevance 0.66

  3. [3]
  4. [4]
  5. [5]

    Financial Times — Global EconomyWill US inflation data persuade the Fed to raise rates this month? ↗

    relevance 0.65

  6. [6]
  7. [7]
  8. [8]

    新浪财经 · 券商研报索引(vReport 宏观+策略)周观点:美联储加息扩表可能性或已浮现 看好供给约束型周期品 ↗

    relevance 0.68

  9. [9]
  10. [10]

    Bloomberg — MarketsBond Traders Brace for More Swings at Both Ends of US Yield Curve ↗

    relevance 0.59

  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]