〔Overnight Brief〕Barclays Maps 6% 30Y as 10Y Crests 5.29% and Mortgage Hits 7.58% — Williams' Late Bid the Sole Relief

0. Overnight Arc

A second session of long-end selling pushed 10-year yields to a 5.29% intraday high and the 30-year fixed mortgage gauge to 7.58% — the highest since November 1, 2023 — before NY Fed's John Williams sparked a late round-trip bounce ^[1][2]. The driver is a coordinated upgrade of long-end forecasts from Barclays and Wells Fargo, both building the same productivity-led r* story that wants the Fed cut path erased ^[3][4][5]; J.P. Morgan layers a bang-ending call on top ^[6], while Citi still argues a soft September core CPI could pull the October hike off the table ^[7]. The split is now structural: hawkish long-end vs. a Williams-defended front end.

1. Rates and the Curve

• **[NEW] Barclays Capital US rates research head:** 30Y fair value could reach 6% if productivity growth keeps accelerating — a level last breached in June 2000 — with term premia absorbing a higher r* and "no reason" for the Fed to cut back to current forward levels ^[3].
• **[NEW] Wells Fargo:** lifts end-2027 10Y call to 5.25-5.75% from 4.50-5.00% and end-2027 fed funds to 4.75-5.00% from 4.00-4.25% — two parallel shifts, one story ^[4][5].
• **[NEW] J.P. Morgan:** Fed hikes go out "with a bang" this year ^[6].
• **[NEW] Citi:** a soft September core CPI could stop the October Fed rate hike — the dovish counter ^[7].
• **[ESCALATED] NY Fed's John Williams:** late-day comments triggered the relief bounce in the cash market ^[1].

2. Cash Market and Mortgages

• **[NEW] 10Y at 5.29%:** the intraday peak came mid-session before the Williams bid; the on-the-run level sits at the next technical resistance band ^[1].
• **[NEW] 30Y mortgage at 7.58%:** up from 7.50% the prior day, the highest since November 1, 2023 ^[2]. Mortgage rates are "doing much better than their often-cited Treasury benchmarks" — a stretched basis, not a mispriced Treasury market ^[2].
• **[ESCALATED] Quarter-end flow:** the mirror-image stocks/bonds correlation is consistent with a quarter-end rebalance dynamic — a one-time technical, not a regime change ^[1][8].
• **[ONGOING] Oil shrugged off:** a "respectable drop in oil prices" had "little to no bearing" on the bond move; the market is pricing a productivity premium, not an energy cycle ^[1][8].

3. Wall Street Calls

• **[NEW] Morgan Stanley's Adam Jonas:** issued "the most compelling endorsement to date" of Tesla FSD — 2,713 miles of FSD driving in the past 12 months (33% of his Model Y mileage), FSD enabled on 100% of trips with 98-99% of miles actually system-driven, and a 28-mile round trip from his residence to MS's Times Square HQ with zero human interventions ^[9][10]. Equal-weight, $400 target, current $357.45 ^[10].
• **[NEW] Meta / "Muse" cluster:** Citi (Buy, $800) flags 2027 compute capacity at ~14GW, with MongoDB CEO CJ Desai hired to commercialize ^[11]; Jefferies (Buy, $875) on a "killer" Muse product and 2028 EPS at 19x ^[12]; Deutsche Bank raises target to $820 (from $750) on a bull case for 2030 revenue above $36bn after Muse hit 3m US installs in 16 days ^[13].
• **[NEW] GF Securities on Lumentum (single-source / unverified relay):** Nvidia's CPO rollout is "faster than expected"; Lumentum CPO + NPO laser revenue could scale from $427m + $501m in 2027 to $750m + $2.55bn in 2028 ^[14].
• **[NEW] European offshore wind (single-source / unverified record-keeper note):** order gap may end in Q4 2026 with 2027 FID nearly doubling; a "rigid" domestic capacity gap in one sub-segment ^[15].

4. Sector Rotation and Commodities

• **[NEW] Wells Fargo sector shift:** downgrades S&P 500 IT to Neutral (from Overweight) and upgrades Industrials to Overweight (from Neutral) — a barbell away from AI infrastructure toward real-economy capex ^[16].
• **[NEW] Wells Fargo gold cut:** 2027 year-end range lowered to $5,200-5,400 (from $5,400-5,600), consistent with the higher-for-longer fed-funds call ^[17][5].

5. What Would Falsify It

• The September core CPI is the live gate: Citi explicitly anchors a soft print to killing the October hike ^[7]; Barclays' 6% 30Y call needs sustained productivity acceleration, not a one-quarter blip ^[3].
• Source quality control: J.P. Morgan's hawkish call is single-source via a Yahoo headline ^[6]; the productivity/r-star thesis is consistent across Barclays, Wells Fargo and the mortgage gauge ^[2][3][4], but the data path is unchanged until CPI prints. Treat the long-end forecast cluster as a band, not a point ^[3][4][5].
• Contradiction is content: hawkish long-end (Barclays 6%, Wells Fargo 5.25-5.75%, J.P. Morgan bang-ending) ^[6][3][4] vs. dovish counter (Citi soft-CPI kills October hike, Williams' late bid) ^[1][7]. Print both.