〔Overnight Brief〕Bessent's 'Quasi-QE' Doubles Long-Dated Buybacks as 30Y Holds Above 5%; Dalio Cuts Bonds, Puts 10-15% in Gold, Calls a US Debt Crisis at 3 Years ±2 — Warsh at Jackson Hole Decides

0. Overnight Arc

The Bessent trade — at least doubling long-dated buybacks in the 10-20Y and 20-30Y buckets on August 19, financed by new issuance, not bank reserves, and not a true QE — is the overnight mechanism ^[1]. BofA's Michael Hartnett calls it 'quasi-QE' and the latest installment in a string of 'Bessent put' interventions ^[2]. The 10Y has already run from 4.44% on July 1 to 4.72% on August 18 (+28bp), with the curve bear-steepening, so the policy put is colliding with a duration premium the Fed cannot fully fix ^[1]. Net: Treasury buybacks in the offer, term premium in control.

1. The Bessent Trade and Long-End Resistance

• **[NEW] Buyback expansion (Aug 19):** Treasury at least doubled liquidity-supporting buybacks in the 10-20Y and 20-30Y buckets, with no reserve injection and no reduction in net financing need; Sino research notes the Treasury still has buffer room from average duration and the short-bond recommended share ^[1].
• **[ESCALATED] BofA's Michael Hartnett:** if Bessent cannot drag the 30Y under 5%, expect USD weakness and a build of short bets against AI hyperscalers, private credit, and financials before the November midterms; 'policy panic' to 'fix' the fixed-income market can cap further yield rises but cannot actually lower them ^[2][3].
• **[NEW] Credibility question:** Jefferies' Modupe Adegbembo and Morgan Stanley's Marina Zavolock debate whether Bessent already has a bond-market credibility issue on Bloomberg Television ^[4].
• **[NEW] Drivers beyond the Fed (TD Securities, Molly Brooks):** 'asymmetric risk' around Fed Chair Kevin Warsh's Jackson Hole speech; term premium, supply, and the Warsh reform plan are doing the heavy lifting ^[5].
• **[ONGOING] Warsh as the trigger:** Aug 28 Jackson Hole is flagged as the moment Warsh either muddles or coordinates with Treasury ^[6][1][5].

2. Gold, Duration, and the Dalio Hedge

• **[NEW] Dalio allocation call:** underweight bonds, 10-15% portfolio in gold plus 'a small amount' of Bitcoin to hedge a US debt crisis he pegs at 3 years ±2; the fiscal gap is ~$2 trillion (revenue ~$5.5tn, spending ~$7.5tn), interest ~$1 trillion, and ~$10 trillion of debt needs refinancing ^[7].
• **[NEW] Goldman (Lina Thomas et al.):** gold call volumes have built a 'two-way price amplifier'; the $4,900/oz end-2026 forecast carries significant upside risk ^[8][9].
• **[NEW] UBS CIO:** end-2026 $4,600/oz retained, new September 2027 target $5,400/oz (raised by $200); the call rests on USD weakness, lower real-rate expectations, and rising investment demand ^[10].
• **[NEW] BIS (Amaral, Guerra, Salcedo, Tomasini, Upper):** high public debt in the Americas has non-linear implications for risk premia and inflation expectations ^[11].
• **[ONGOING] HSBC's Max Kettner:** 'a streak of a couple of weeks uninterrupted of negative economic surprises' can bring back a duration bid in US Treasuries ^[12].
• **[ONGOING] FT framing:** 'The US economy is running hot', with a few cool data points unlikely to mark weakening demand; separately, FT flags productivity-adjusted labour costs as the key inflation signal ^[13][14].

3. AI Debt, AI Capex, and the 'Hyper' Short

• **[NEW] JPMorgan (Chong Sin, Terrell Bobb):** floating-rate SASB CMBS data center spreads have widened even though primary data center SASB issuance is only $4.9bn YTD (~4% of private-label, down from 7.1% last year); the pressure comes from data center and adjacent financing elsewhere, so AAA data center selection is now more selective as leverage has risen ^[15].
• **[NEW] Hartnett's short list:** if the Bessent plan fails, AI hyperscalers and private credit are first on his short list, financials too ^[2][3].
• **[ONGOING] A-share AI tape:** the sector remains a momentum story — Yuanda's weekly notes the adjustment does not change the medium-term thesis, even as the industry-chain report (Sina index) flags the market is in a 'strong reality, weak expectations' divergence phase on AI ^[16][17].

4. Asia, China, and the Midterm Cross-Read

• **[NEW] SK Hynix:** announces 40 trillion KRW share buyback-and-cancellation over three months, the largest in Korean history; equal to 24 million shares / 3.3% of float at the 1.66 million KRW Tuesday close; 2025-27 cumulative FCF return target raised from 'up to 50%' to 'over 50%' ^[18].
• **[ONGOING] A-shares weekly (Yuanda):** resources and dividends lead — petroleum/petrochem +4.02%, agriculture +3.90%, banks +2.69%, transport +1.60%, coal +1.45%; tech lags — computer -5.04%, media -4.83%, defense -4.78%, building materials -3.78%, telecoms -3.77% ^[16].
• **[ONGOING] China policy tape:** August LPR unchanged for a 15th month at 1Y 3.0% and 5Y+ 3.5%, with Q3-end RRR/cut expectations rising; July IP +4.5% y/y with high-tech and equipment strong, mining and traditional sectors weak ^[16][19].
• **[NEW] SPDB International midterm preview:** since 1978 across 12 midterms, VIX rises pre-vote and eases post; Nasdaq leads equities, long Treasuries beat short, USD weakens post-vote, gold's hedge emerges gradually, oil lags ^[20].
• **[ONGOING] Sino sector tracker (Aug week 3):** electronics up on continued memory price gains and mid-to-high PCB export growth; oil/coal up sharply (Brent +5.9% w/w, Qinhuangdao Q5500 +2.4% w/w, stock -2.4% w/w); domestic demand still soft, typhoon a drag on travel ^[21].
• **[ONGOING] China monetary transmission (CEPR/VoxEU):** contractionary Chinese monetary policy announcements drive significantly lower subsequent returns in country and firm-level stocks with greater exposure to Chinese mutual funds, via a retail-led portfolio rebalancing channel ^[22].

5. What Falsifies This

• Warsh at Jackson Hole on Aug 28 is the cleanest single test: a coordinated message lowers duration; a muddled one leaves the 30Y above 5% and forces Hartnett's AI / private-credit shorts ^[2][6][5].
• The US-Iran path is the swing variable for Q4: a Trump 'TACO' would ease inflation uncertainty and curve pressure; an escalation could send the 10Y toward 5% and revive the 'K-shaped' domestic pressure flagged in this week's macro write-ups ^[1][23][21].
• Source-quality flags: Jefferies/MS, TD Securities, BofA, HSBC, and BIS items are direct bank/research attributions ^{[2][5][4][3][12][11]}; the Bessent 'credibility' framing is one Bloomberg TV segment and should be treated as a view, not consensus ^[4]. The '30Y under 5%' is Hartnett's trigger, not a market level ^[2]. The China policy list is a single Guoxin Securities weekly ^[19].