The Last Free Price

morning musings 7.14.26- views from the island

‘You can ignore reality, but you cannot ignore the consequences of ignoring reality.’ ~Ayn Rand

So the administration now owns 10% of Intel, 15% of MP Materials, slugs of Lithium Americas and Trilogy Metals, a ‘golden share’ of US Steel, and a running tab north of $21B in ‘strategic’ equity stakes BBG. And this month it started promoting ‘Freedom Fuel’ stations pumping gasoline at $3.479 a gallon, roughly 17 cents under what it costs to pull it out of the ground and haul it to the pump. The White House swears there is no subsidy and no government hand, which would make it the first charity in recorded history incorporated in Delaware with no owner on the filing CNN. Twenty-five stations around Philadelphia, losing money on every gallon by design, and we are asked to believe the invisible hand did it. The man who spent a decade calling everyone else a communist has quietly assembled the commanding heights: the chip supply, the rare earths, the steel, and now the price at the pump. Say what you like about him, he would have run one hell of a Gosplan. (I could not verify the literal ‘best communist’ line against a primary source. The portfolio is real.)

None of which is a digression, because it is the same trade. A state that sets gasoline below cost, tolls the Strait of Hormuz at 20%, and warehouses equity stakes it will never mark to market is a state that has decided the price mechanism is a suggestion. That is the fiscal-overhang, term-premium regime with a flag planted in it. It is why the long end will not behave, why gold is doing something this morning it is not supposed to, and why the bond market is the only honest broker left in the building.

Five banks printing records into a screen bleeding red. IBM off a fifth of its cap on a preannounce nobody flagged. Oil paid, chips squeezed off the mat overnight, and the gold market the entire street buried on Monday quietly bid this morning while real yields sit at a 2023 high. The surface says calm: VIX 17, credit tight, banks minting money on the same volatility that is supposed to terrify everyone. Underneath, the regime is repricing in plain sight. The story is not the red headlines. It is which havens caught a bid and which did not, and what that tells you the market is being made to do.

STATE OF PLAY. The sharpest read of the week on the tollbooth is Jonathan Samuel over at CoarsemanNews, who actually read the Islamabad document instead of the press release CoarsemanNews:

“It did not say Iran must keep the strait free. It said Iran would use its best efforts for safe passage, with no charge for 60 days only. Sixty days. Only. The ceasefire wasn’t a ceasefire. It was a free trial ... The 20% will never be collected as a ‘toll’, it’s legally impossible. It will reappear as war-risk insurance premiums, escort contracts, and ‘security services.’ The toll will be laundered through the premium.”

That is the part the tape has not priced. If the mechanism is right, the Hormuz premium does not live or die on a headline ceasefire, because it was never built to end. It migrates off the front page into Lloyd’s war-risk listings, freight rates and the crack spread, where it becomes a sticky cost that keeps feeding forward inflation long after the cable desk has moved on. That is the whole difference between a geopolitical spike you fade and a supply tax you underwrite. It is why I am short duration and long energy and not waiting for the 4pm deadline to tell me which. And the real tail is not oil at all: it is Beijing filing this away as the precedent for a percentage at the Taiwan Strait. That option is not in a single bank multiple.

---

TL;DR

• Live tape is a term-premium plus oil-supply-shock rotation: real assets bid (Brent $86.22 +3.5%, gold $4,033 +0.7%, copper +1.6%), dollar and long bonds offered (10Y 4.609%, +4bp, near session high; DXY -0.21%). Not a de-gross, not a clean risk-off. FMP

• The non-consensus catch: gold reclaimed a bid into RISING real yields. Every brief on the desk still has gold as a ‘failed haven.’ The tape says the marginal buyer is not spent. Watch $4,150.

• CPI at 8:30 and Warsh at 10:00 are the whole-book gate. One print, core at or under +0.1% m/m that drags 10Y under 4.45%, flips the regime and squeezes everything you are short.

• Banks are a sell-the-strength tape: JPM record profit but -1.8% pre on a raised expense guide; GS record equities $7.42B but crowded with a 2:1 put skew; BAC beat and sold. XLF at 52-week highs into a double binary.

• Book stays short duration, long energy, long defensives, with the short-semi leg the one fighting a live squeeze today. Two intra-financials expressions below: networks over consumer finance live, IB sell-the-news armed but not triggered.

---

POST-CPI UPDATE · 8:31 ET · live

The print hit the falsifier threshold and the tape refused to confirm it. Core CPI landed 0.0% m/m (2.6% y/y), headline -0.4% m/m / 3.5% y/y, softer than consensus across the board. On paper that is the branch that kills the rates-up book. The market said no. FMP

• 10Y barely moved. Roughly 4.59% inferred off the note future, down 1-2bp, nowhere near the 4.45 falsifier. A soft core that should have bid duration got shrugged. (inferred from ZN; the direct 10Y quote is blocked by the FMP plan tier, so the bp is approximate, the direction is not.)

• Equities did not rally. ES still red, NQ less green than pre-print. No relief rally. A tape that will not buy soft inflation is telling you the regime outranks the data.

• Oil went the other way. Brent $86.79 (+4.2%), WTI over $80. Forward inflation strengthening while the backward-looking June data softens. That is the supply-tax tell. FMP

• Gold held the bid. $4,032, +0.7%, through the one print that could have broken the variant view. Still stand down the NEM short.

Book: do NOT cover the TLT short. There is no relief rally to cover into, and the decision-tree cover trigger (10Y toward 4.45) never printed. Keep the halved SMH short; the squeeze had no CPI accelerant. The XLE add-toward-$55 thesis strengthens on firmer oil.

The swing is now Warsh at 10:00, not the data. If he dismisses the stale June, the regime is confirmed and duration stays offered. If he leans on the 0.0% core as room to wait, that is the only path to the duration rally the book is short. Line in the sand stays 4.45 on the 10Y. Right now it is rejecting lower yields, not chasing them.

MY READ

I think the tape is telling you the opposite of what the briefs are telling you, and the tell is gold.

Consensus this morning, and it is unanimous across every document on my desk, is that gold has failed as a haven: it fell Monday into a live Hormuz escalation, real yields are the highest since 2023, and the reflex ‘crisis equals buy gold’ trade is broken. The broad book is short it through Newmont. The house brief calls it a spent haven. And on Monday’s close, they were right.

They are looking at yesterday. Live, gold is up 0.7% at $4,033, bid into a morning where the 10Y is climbing, not falling. That is not supposed to happen. A non-yielding asset catching a bid while its single biggest headwind (real yields) is pressing higher means the marginal buyer is not reaching for a rate trade at all. It is reaching for a supply-shock and inflation hedge, and it is willing to eat the carry to own it. That is a different, more durable bid than the one the street just buried.

So my variant view: the ‘failed gold haven’ is the crowded trade, not the contrarian one. The consensus got long-dated on Monday’s tape and has not re-marked. I am not chasing gold here, but I am not short it, and I am watching $4,150 as the level that turns a twitch into a reclaim. The rest of the book (short duration, long energy) is aligned with the briefs and with the tape. On gold, I am offside the desk and I think the desk is offside the market.

Confidence: moderate. What kills it: gold closes back below $4,000 with the dollar catching a bid, in which case the real-yield headwind wins and the failed-haven crowd was right after all.

WHAT MATTERS TODAY

• CPI, 8:30 ET. Consensus headline about -0.1% m/m and ~3.8-3.9% y/y on the June gasoline drop, core about +0.2% m/m and ~2.8-2.9% y/y third-party est, verify vs BLS. This is a backward-looking June print landing into a July oil re-spike, so a soft read may get discounted, not bought. The number that matters is core. Above +0.3% m/m revives the hawk and steepens the bear move. At or under +0.1% and the 10Y through 4.45% is the single print that flips the regime.

• Warsh, 10:00 ET. First testimony as Chair, House Financial Services. He killed forward guidance on principle, which is elegant right up until you need to guide, and now the market gets to read tea leaves off a man who told it there would be no leaves. Waller did the hawkish work for him Monday (’if core stays hot, the FOMC will need to consider tightening’). July hike odds near 50% from under 10% a week ago. The awkward part nobody on the panel will say out loud: the Fed is being asked to fight an inflation the fiscal side is actively manufacturing, one Hormuz toll and one below-cost gas station at a time. PDB, BBG

• The bank tape, pre-open. JPM, GS, WFC, BAC reported, C on deck at the 11:00 call. Records across trading and IB, and the group is being sold anyway. That reaction is the information, not the beats.

• Oil into the 4pm Hormuz deadline and 4:30 API. Brent pinned one tick off the session high, WTI lagging at sub-$80. The premium is being paid. Washington is simultaneously tolling the strait at 20% and selling gas below cost at home, which is a fascinating way to fight inflation: tax the barrel on the water, eat the loss at the pump, and let the CPI print land on Warsh’s desk. A verified de-escalation is the only thing that collapses the premium, and there is not a whisper of one.

MOVERS THAT MATTER

• IBM, -17% to -23% pre-market. Preliminary Q2 $17.2B vs $17.9B, infrastructure -7%, Krishna citing client capex reprioritization in June. So what: this is not an IBM story, it is an enterprise-IT budget story. The read-through hammered the whole software complex pre-market (ACN -8%, WDAY -7.3%, NOW -6.8%, CRM -5%, ADBE -4.8%, MSFT -3.1%) on the fear that AI infrastructure spend is cannibalizing legacy software budgets, not adding to them. That is the bifurcation the tape has been groping toward for two weeks, now with a number attached. BBG

• Goldman, +2.2% pre on a record $7.42B equities haul; JPM -1.8% on a record profit. So what: same quarter, opposite reactions, and the difference is the guide. JPM raised the FY expense outlook to ~$107.5B from ~$105B and the tape sold the record. When a group at 52-week highs sells its own best print, the beat was in the price and the swing factor is forward guidance, not the quarter. That is the definition of a sell-the-news tape.

• SK Hynix +7% in Seoul, memory stabilized overnight. So what: the leg of the book that is short semis (XLV/SMH) is fighting this. Monday was a 20%-plus two-day drawdown in Hynix and a 4%-plus SOX day; this morning is the bounce. The AI-demote is a weekly thesis, but pressing a crowded short into a live stabilization is how you bleed on a day the rest of the book is working. Size accordingly.

FINANCIALS

• The driver of the day is not the beats, it is the rejection of the beats. Five banks delivered records or near-records on trading and IB, and the group is being sold: JPM -1.8%, BAC -1%, with XLF sitting at 52-week highs ($56.07, year high $56.59) into a double binary of CPI plus five prints. FMP, Fin Daily When a sector this extended sells its own record quarter, the message is that the good news was fully owned and the marginal buyer has left. The swing factor from here is NII trajectory and expense guides, not the EPS line, which was a near-certain beat.

• The dispersion is the whole point, and it is wide. Payment networks led again (’V +2.5%, MA +2.1%’ Monday, live V $357.75, MA $537.70), the clean no-credit secular expression with zero exposure to what a higher-for-longer rate path does to a card book. FMP, Fin Daily At the other end, the credit-sensitive complex carries the tail: office CMBS delinquency at a record 12.3%, $936B of 2026 CRE maturities, BCRED cutting its July distribution to $0.18 (second cut in nine months) and prorating redemptions at the 5% gate after requests hit ~10%, BDC non-accruals up ~40% q/q to ~2.0%. Fin Daily That is the private-credit gating tail that is not in a single bank multiple, and it sits under the regionals and consumer-finance names, not the networks.

• The level that matters: the 10Y at 4.61% is the pivot for the whole group. Below ~4.75%, the back-up is a NIM tailwind and the diversified banks earn on a stable-to-steeper curve (2s10s +36bp, modest steepening). FMP Through 4.85%, it flips: AFS/HTM marks start biting the regional cohort, AOCI becomes the story, and the bank-long thesis breaks. We are 24bp from that line. So the read is not ‘own financials.’ It is own the subsector dispersion: quality diversified and no-credit networks over the crowded, priced-for-perfection IB and the CRE-and-deposit-exposed regionals. Both financials expressions in the paid section are intra-sector, sized separately from the XLF beta leg.

SECTORS

Monday close, the tape setting up today. Futures this morning hold the same axis: energy bid, tech heavy but memory bouncing.

• XLE +3.01%. The clean winner. Oil supply premium flowing straight to energy earnings; the only sector fully aligned with the live regime.

• XLU +0.68%. Defensive bid, rate-proxy caveat: works as ballast, vulnerable if the 10Y keeps climbing.

• XLF +0.65%. At 52-week highs into its own earnings binary. Strength is the setup for a sell-the-news, not a reason to add.

• XLP +0.56%. Staples doing their job in a rising-real-yield tape. Boring and correct.

• XLRE +0.56%. Green on the tape but carrying the CRE-office overhang; do not confuse the bounce with the all-clear.

• XLV +0.35%. Defensive rotation leg, though HCA cutting FY guidance is a reminder the group is not uniformly safe.

• XLC -0.04%. Flat, masking a heavy IBM-driven software read-through underneath in the growthier names.

• XLB -0.61%. Copper bid but the sector soft; the metal is trading the China-export beat, not the equities.

• XLI -0.85%. Industrials give back; Fastenal beat but the group leans risk-off with the tape.

• XLY -1.02%. Discretionary heavy, the higher-real-yield tax on long-duration consumer cash flows.

• XLK -2.42%. The IBM epicentre. Enterprise-software de-rate is the single largest company-specific drag on the index.

FLOWS & POSITIONING

• The positioning setup is the risk, not the fundamentals. BofA’s fund manager survey has cash at an ‘uber-low’ 3.6% and US equity positioning at the highest since December 2024, a contrarian warning with no dry powder to cushion a shock. BBG Breadth keeps narrowing: the top 10 S&P names are now 40%-plus of index cap, which is exactly why an IBM-style single-name air pocket can bleed the tape. Retail net flows have rolled to their lowest since Covid, so the marginal price-insensitive bid has thinned at the same time positioning is stretched. PDB, Vanda via desk color

• And then the one that should stop you cold: BofA’s July fund manager survey has ‘long global semiconductors’ as the most crowded trade at 82%, the single most crowded reading for any trade in the history of the survey. BofA FMS Sit with that. Not the most crowded of the year, of all time. When four out of five managers are on the same side of one boat, the tape does not need bad news to break it, it needs a reason to reduce, and IBM just handed everyone one. That is the structural case for the short-semi leg even on a morning the memory names are squeezing higher: the daily bounce is noise, the record crowding is the fuel. You do not press the short into the squeeze, but you do not cover the thesis either.

• Under the surface the de-gross is specific, not broad. Foreign investors sold a record net $32.4B of Korean equities in June, the largest monthly outflow in the series, and chip ETFs still pulled $21B last week even as the memory names whipsawed. BBG Bond vol is repricing for real, MOVE ~77.8 and +8.2, which tells you the rate move is a genuine term-premium repricing, not a quiet drift desk/trade sheet. Credit is the one thing still asleep: HY OAS ~269-284bp, tight, calm through the whole oil escalation. Fin Daily, web est Calm credit into a triple risk set is the cheapest thing on the board to fade. Gradually, then suddenly.

PAID BRIDGE

• Behind the wall today: the full book marked to live, re-underwritten leg by leg under the term-premium regime, with the exact levels. The cross-asset expressions (short TLT as the primary duration short, long energy sized against the arm, the defensive-over-AI pair and where to halve it into the squeeze), the two intra-financials trades (networks over consumer finance live, the Goldman sell-the-news armed with its trigger), a CPI decision tree with both branches priced, the conviction ranking, and the watchlist including the gold reclaim level that would flip me off the desk consensus. Plus the one leg the red team says to change today.

morningmusing.com