The rates story shifted from the Fed to the Treasury. Bessent has launched a debt-recycling program — selling bills, buying back at least $4 billion of long-dated Treasuries per operation — with 30-year yields at a 19-year high of 5.31%, and one research firm, Citrini Research, describes Warsh and Bessent as acting in coordination to pull the long end down, an interpretation reported by MarketWatch and Crypto Briefing. The question is not only what Warsh says about inflation at Jackson Hole, but whether the Fed’s rate stance and the Treasury’s yield management are pointing in the same direction. They currently are not: Cleveland’s Hammack said “now is the time to act” on hiking, and Kansas City’s Schmid called core PCE at 3.3% sticky and said the current policy rate is not restraining borrowing, while Treasury operations are designed to suppress the long end.
Front-end pricing agrees that a cut is off the table — Kalshi puts a September cut at 1%, with a hold at 71% and a 25bp hike at 31%. Elsewhere, the AI trade broadened but also fractured: Nvidia and Salesforce beat, Marvell fell 8% on guidance. And a cyberattack has reportedly halted Boston Scientific’s global cardiac device shipments, which is the most concentrated single-company operational risk in today’s data.
New Developments
The Treasury is now the marginal buyer of duration
Mechanism: Treasury issues short-dated bills, uses the proceeds to repurchase long bonds, and thereby shortens the public debt’s duration. That reduces the amount of long-duration risk private investors must hold, which compresses term premium at the long end. Front-end funding costs rise with bill supply. The result is curve flattening driven by debt management rather than by monetary easing.
Three things follow. First, $4 billion per operation against roughly $40 trillion of debt is small relative to outstanding stock, which is why Forbes framed it as partly signaling. Its effect depends on whether investors believe the program will scale. Second, if long yields do fall, MarketWatch reports one research house sizing a $326 billion unrealized loss position on bank balance sheets that would improve. I treat that figure as a single-source estimate and not established. Third, UBS raised its long-term Treasury yield forecast, arguing deficits, tariffs and higher defense spending keep structural pressure on the long end regardless of the buybacks. Those two views are not incompatible: the twist can lower yields tactically while the structural drift remains upward.
The larger risk is how markets interpret the program. Post-WWII, the Treasury-Fed dispute over rate pegs ended in a formal grant of Fed independence, and Barron’s makes that comparison explicitly. If markets conclude the Treasury is setting the price of long money while the Fed talks about inflation, the compensation demanded for holding long bonds widens. That would defeat the program’s purpose.
Positioning makes this two-way. Leveraged funds are net short 39.8% of 10-year open interest and 19.1% of the 30-year contract. A credible, scaled buyback into that short base produces a violent squeeze. Do not press a short-duration view here on the structural argument alone.
AI capex faces a permitting constraint, not a demand constraint
Investor’s Business Daily reports a growing number of states enacting data-center moratoriums as governors head into reelection. This is one outlet (two syndicated placements of the same story), so treat it as an early signal.
The transmission is a timing problem. Announced hyperscaler capex converts to revenue for equipment and electrical contractors only after siting, permitting and interconnection. Moratoriums lengthen that interval without reducing the demand behind it. Constrained new supply raises the value of already-sited, already-energized capacity, which favors existing data-center owners and power contracts over the construction supply chain. I am not initiating on a single source, but this is the first policy risk to the AI buildout in this window that runs through local politics rather than tariffs or financing.
Boston Scientific’s shipment halt is a concentrated single-name risk
Reporting describes an August 25 IT compromise that stopped global order processing and shipping for pacemakers, stents, defibrillators and WATCHMAN implants. Implantable devices have a specific feature that makes lost revenue less recoverable than in most supply disruptions: when a hospital substitutes a competitor’s device mid-procedure-schedule, physician preference frequently stays with the substitute. The share loss can outlast the outage. Confirmation gap: both articles come from one outlet and I have seen no company statement or 8-K quantifying the impact. Conviction stays neutral pending that.
Developing Themes
Hawkish faction now on the record. Hammack’s “now is the time to act” is explicit advocacy for raising rates; Schmid’s judgment that core PCE at 3.3% is sticky and that policy is not restrictive signals resistance to cuts rather than advocacy of a hike. Core PCE at 3.3% year over year and rising, CPI at 3.5%, initial claims at 203,000 and unemployment at 4.10% sit alongside a 4-week claims average that rose for a second straight week to about 205k and July payrolls that fell by 23k, so the labor market is resilient but softening, which gives the hawkish faction less cover to hike immediately. The hike-by-year-end probability sits at 60%.
Oil risk premium has largely drained. Goldman estimates Hormuz flows back at two-thirds of pre-war levels, and crude is set for a weekly loss despite the US shifting to a Navy blockade of Iranian exports. Managed-money crude positioning is nearly flat at -10,696 contracts (-1.3% of open interest), so a Hormuz supply shock would not be cushioned by short covering flows into a large existing short base. The escalation channel has migrated to secondary sanctions on third countries — a payments and banking risk, which oil exposure does not hedge.
AI dispersion. Nvidia’s beat and Salesforce’s 48% revenue growth with raised guidance sit alongside Marvell falling 8% on fiscal 2028 guidance despite 37% revenue growth. Within-AI selection now matters more than sector beta.
Corporate margins at a post-war extreme. The FT reports pre-tax earnings at their highest share of national income since just after WWII, while labor’s share weakens. Real PCE remains above 2%. That combination supports index earnings but sharpens the political vulnerability of margins — the same dynamic visible in fuel surcharges exceeding cost pass-through and in the White House summoning refiners.
Continuing Themes
Housing transmission remains negative: the 30-year fixed rose to 6.66%, starts are down 13.5% year over year, and existing sales at 4.06 million continue to fall. Separate research warning of roughly 3 million excess homes is single-source and unverified; treat it as a hypothesis, not a supply thesis.
ECB accounts show some Governing Council members would not have opposed a July hike, with eurozone corporate lending at a three-year high of 4.0%, a mildly hawkish tilt with no rate action yet.
What to Watch
Treasury Launches Debt-Recycling Buybacks as 30-Year Yield Hits 19-Year High of 5.31%
Treasury is selling bills and buying back at least $4B of long-dated Treasuries per operation to suppress long-end yields, with research houses describing this as coordinated Fed-Treasury policy and UBS raising its long-term yield forecast on structural deficit, tariff and defense pressures.
FIRST-ORDER EFFECTS
Buying long-dated paper with bill issuance shortens the average maturity of federal debt and mechanically compresses long-end yields relative to the front end.
Higher bill supply raises the effective floor on front-end funding, making the curve flatter through the long end rather than through easier policy.
SECOND-ORDER EFFECTS
Lower long yields would reduce unrealized losses on bank held-to-maturity and AFS bond books, improving tangible capital and buyback capacity at large lenders.
If markets read the twist as fiscal dominance rather than debt management, term premium and inflation compensation can widen again, offsetting the intended yield relief.
TICKERS
⚪ TLT — Direct beneficiary of long-end buybacks, but exposed to reversal if the market prices fiscal dominance into term premium.
⚪ BAC — Cited as among the lenders whose large bond-portfolio marks improve most if long yields fall, though this rests on one research note.
⚪ GLD — Treasury-led yield suppression alongside above-target inflation supports the debasement bid, but speculative gold longs are already crowded at 34.9% of open interest.
Warsh’s Jackson Hole Debut Lands With Hammack and Schmid Openly Pushing to Hike
Fed Chair Kevin Warsh delivers his first major speech Friday with long yields near two-decade highs, while Cleveland’s Hammack said ‘now is the time to act’ on raising rates and Kansas City’s Schmid called core PCE at 3.3% sticky and current policy not restrictive.
FIRST-ORDER EFFECTS
Named FOMC participants advocating a hike removes a cut from the near-term distribution and keeps the front end anchored near current levels.
A framework-only speech that avoids the rate path would leave term premium elevated, since the bond market’s stated need is an inflation plan.
SECOND-ORDER EFFECTS
Persistent hawkish rhetoric combined with Treasury yield suppression puts the two arms of policy in visible tension, raising the premium markets demand for Fed independence risk.
Rate-sensitive equities priced for eventual relief — utilities, homebuilders, small caps — face repricing if the hawkish faction gains a majority.
TICKERS
🔴 XHB — Homebuilders need lower mortgage rates; the 30-year fixed rose to 6.66% and the hawkish faction makes near-term relief less likely.
🔴 IWM — Small caps carry the most floating-rate debt and near-term options vol is already rich at 22.6% against 18.6% realized.
🟢 CME — Unresolved policy path with documented hike advocacy sustains rate-futures and options volume.
AI Earnings Split: Nvidia and Salesforce Beat, Marvell Falls 8% on Weak Forward Guidance
Nvidia rose about 7% on an earnings beat and Salesforce reported 48% revenue growth with raised guidance, lifting SaaS names including Figma, while Marvell fell 8% premarket as its fiscal 2028 outlook disappointed despite 37% revenue growth.
FIRST-ORDER EFFECTS
Nasdaq and S&P 500 gains were driven by AI-linked megacaps, pulling the index back toward its August high.
Marvell’s guidance miss shows that within-AI dispersion is widening: revenue growth of 37% was insufficient against embedded expectations.
SECOND-ORDER EFFECTS
Custom-silicon and networking suppliers now trade on multi-year design-win visibility rather than current growth, so guidance becomes the binding variable for the second tier.
Strong enterprise software results support the view that AI spend is converting into software revenue, not only hardware capex, broadening the sector’s earnings base.
TICKERS
🟢 NVDA — Reported beat and a roughly 7% move confirm continued AI accelerator demand from company results rather than commentary.
🔴 MRVL — Forward guidance below expectations despite 37% growth signals margin and share pressure in custom AI silicon.
🟢 CRM — 48% revenue growth with raised full-year guidance is company-reported evidence that AI monetization is reaching enterprise software.
Growing Number of States Impose AI Data-Center Construction Moratoriums
Investor’s Business Daily reports a rising number of US states are implementing data-center construction moratoriums as public backlash meets AI-driven buildout, with governors facing reelection.
FIRST-ORDER EFFECTS
Permitting delays lengthen the time between announced AI capex and delivered compute capacity, pushing revenue recognition right for construction and electrical suppliers.
Siting constraints concentrate new capacity in permissive jurisdictions, raising land, power and interconnection costs there.
SECOND-ORDER EFFECTS
If capacity delivery slows while demand holds, existing data-center and power contracts gain pricing power, favoring incumbents with sited, energized capacity.
Local political resistance introduced ahead of midterms creates a policy risk channel to AI capex that is independent of demand or financing conditions.
TICKERS
⚪ VRT — Data-center thermal and power equipment orders are directly tied to construction permitting timelines.
⚪ DLR — Owners of already-permitted, powered capacity benefit from supply constraints on new competing sites.
⚪ PWR — Electrical infrastructure backlog conversion slows if state-level moratoriums delay project starts.
Cyberattack Halts Boston Scientific Cardiac Device Shipments Globally
Reporting says an August 25 compromise of Boston Scientific’s IT systems halted global order processing and shipping of pacemakers, stents, defibrillators and WATCHMAN implants, with analysts projecting weeks of disruption and hundreds of millions in quarterly revenue impact.
FIRST-ORDER EFFECTS
Hospitals lose access to a primary supplier for implantable cardiac devices, deferring elective procedures and shifting orders to competitors.
Revenue lost during an order-processing outage in implantables is only partly recoverable, since procedure slots and physician preference cards move to substitutes.
SECOND-ORDER EFFECTS
Competitors with available inventory can convert emergency substitution into durable share, because implant choice is sticky once a physician switches.
A successful attack on a Class III device manufacturer raises the probability of FDA and hospital-system cybersecurity requirements that increase compliance costs across medtech.
TICKERS
⚪ BSX — Halted shipments of its highest-margin cardiac franchises threaten the current quarter, though the reporting comes from a single outlet without company confirmation.
⚪ MDT — Largest alternative supplier of pacemakers and defibrillators positioned to absorb emergency hospital demand.
⚪ ABT — Cardiac rhythm and structural heart portfolios overlap directly with the disrupted product lines.
US Shifts to Naval Blockade as Hormuz Flows Recover to Two-Thirds of Pre-War Levels
Six months into the conflict the US has moved from airstrikes to a Navy blockade of Iranian oil exports and secondary sanctions threats, while Goldman Sachs estimates Hormuz flows have recovered to two-thirds of pre-war levels; crude is set for a weekly loss and Trump is meeting refiners over gasoline prices.
FIRST-ORDER EFFECTS
Physical flow recovery through Hormuz removes the acute supply premium, and crude is on track for a weekly decline despite the blockade escalation.
Fuel surcharges continue to raise landed costs for US retailers and manufacturers, and some freight carriers are earning margin on those fees.
SECOND-ORDER EFFECTS
Secondary sanctions on countries still trading with Iran move the escalation risk from oil supply into cross-border payments and banking, a channel oil prices do not hedge.
Rising pump prices ahead of midterms increase the political probability of refiner jawboning or export restraint, which would compress crack spreads rather than crude.
TICKERS
⚪ XLE — Flow normalization at two-thirds of pre-war volumes removes the acute geopolitical bid while sanctions keep upside tail risk alive.
⚪ MPC — Refining margins are wide on high fuel prices but face direct political pressure from the announced White House meeting with refiners.
⚪ ODFL — Reuters reports carriers are profiting from fuel surcharges that exceed underlying cost increases.
Options markets show calm at the index level but stress at the margins, with HYG’s near-term implied volatility rich at 7.8% against 4.1% realized even as credit spreads stay tight, and TLT’s 10.7% implied volatility offers no edge over 9.7% realized. The premium section works through what these signals mean for positioning in TLT, CME, XHB, NVDA, CRM and MRVL, and how the squeeze and fiscal-dominance risk scenarios could unfold if Warsh’s Jackson Hole remarks are read as accommodating Treasury’s yield-suppression program. Full options positioning analysis, portfolio playbook, and risk scenario framework below for subscribers.
This publication is for informational and educational purposes only and does not constitute financial, investment, or trading advice. The analysis, opinions, and commentary presented here should not be interpreted as a recommendation to buy, sell, or hold any security. Always conduct your own research and consult a qualified financial advisor before making investment decisions. Past performance does not guarantee future results.
