〔Overnight Brief〕Goldman Defers Second Hike to December on 3.0% Core PCE, Yet Long Yields Stay Sticky and the Energy-Rate Link Snaps — Earnings Buffer vs Bond Velocity

0. Overnight Arc

The dovish repricing dominated the wires: Goldman moved its second hike call from October to December after core PCE printed 0.2% m/m, with economists Jan Hatzius, David Mericle and Alec Phillips writing that the FOMC may ultimately see no further hikes ^[1][2][3]. Yet the long end barely budged — global bond yields hover near multi-decade highs, the 10-year US Treasury touched 5.29% in September, and Goldman's One-Delta desk says the energy-rate link is breaking ^[4][5][6]. The split: a dovish Fed call against a sticky long end, with the equity buffer narrowing if bond velocity picks up ^[4].

1. Fed Path Repricing

• **[NEW] Goldman Sachs:** pushed the second hike from October to December after Wednesday's inflation data and New York Fed President John Williams' remarks; Q4 core PCE forecast at 3.0%, well below the FOMC median of 3.4%; 'high' probability the FOMC ultimately sees no further hikes ^[1][2].
• **[NEW] Morgan Stanley China (Chief Economist Xing Ziqiu):** shifted to one hike each in December 2026 and March 2027; will track liquidity impacts and notes foreign capital favors Chinese tech ^[7].
• **[NEW] Oxford Economics:** a hawkish Fed will not deliver 'bullish hikes' — body text unavailable, flagged as headline-only ^[8].
• **[ONGOING] SPDB International read on the September Fed move:** 25bp to 3.75%-4.00%, dot plot showing possibly one more hike this year, 10-year US Treasury touched 5.29% ^[6].

2. Bonds Aren't Buying It

• **[NEW] Core PCE 0.2% m/m vs 0.3% expected, annual 3.0%** ^[3]. Caveat: the unrounded monthly print was 0.247, 'about as high as it could be without rounding up to 0.3', and the annual step-down from 3.3% came from a PCE methodology change, not new data collection; not all of the methodology change is transparent ^[3].
• **[NEW] Goldman One-Delta (head Rich Privorotsky):** the cross-asset signal of widening credit spreads, falling rates and declining oil was the 'most unsettling cross-asset signal day of this crisis'; the regime has shifted from energy shock to fiscal financing, with high real yields the 'clearing price' sovereigns must pay to attract private capital ^[5].
• **[ONGOING] Physical market:** Hormuz Strait shipping 'basically recovered'; another ~40 million barrels released from the Strategic Petroleum Reserve; spreads and freight still elevated ^[5].

3. The Stock-Earnings Buffer

• **[NEW] Goldman Chief Global Equity Strategist Peter Oppenheimer:** earnings have stayed healthy and absorbed some valuation pressure, but the buffer is not infinite; the risk is the magnitude and speed of any further bond move, not the absolute yield level — 'in that case, I think equities will become fragile' ^[4].
• **[NEW] UBS Global Wealth Management strategists:** since 1994, hedge funds posted cumulative positive returns in every Fed tightening cycle, but high rates do not guarantee strong performance; recommends diversified exposure to select hedge fund strategies ^[9].
• **[NEW] Bernstein upgrades (Outperform):** Coherent Corp, $350 target; Lumentum, $1,220 target; Celestica, $520 target ^[10][11].

4. Falsifiable Tests and Sourcing

• The dovish Goldman call rests partly on a methodology-driven annual revision to 3.0%, not new data; the next PCE print is the harder test ^[3].
• The Caixin 'modern apocalypse' framing of the Treasury market is headline-only, flagged thin ^[12].
• The Federal Reserve Board finalized changes to its stress test to enhance transparency and reduce volatility in stress-test-related capital requirements ^[13].
• CME FedWatch: October hike probability at 64.20% as of Sep 25 — a single pre-print, pre-Goldman-call snapshot, treat as stale ^[14].
• SPDB International flags Q3 earnings as the next key verification window for AI-led equities; long-end rise was driven by real-rate gains and a higher policy-rate center, not just inflation surprises ^[6].