Two Forces, One Fence.

The map of the week

• The Fed held for a fifth time, 3.50 to 3.75 percent, on a hawkish 9-to-3 vote, and the very next morning June core PCE, the inflation measure it actually targets (the Personal Consumption Expenditures index, stripped of food and energy), came in soft at 0.1 percent on the month. The standoff between the two is the week’s story, and the Analytical Takeaway is where it is argued.
• The fence tell. The two-year Treasury, the cleanest read on the Fed’s next move, closed almost exactly between the two lines we drew last week. The full call, with its levels, is in the 90-Second Read above and is scored in the Takeaway.
• Oil is decelerating but still the year’s engine, and the energy dial on the crash gauge stays red. The full oil-to-rates chain is in the Takeaway; the barrel’s week is in The Week That Was.
• The stock market made a fresh record on megacap earnings, sorted on one axis: AI spending that shows up as revenue was rewarded, and AI spending that only shows up as cost was punished. The earnings are in The Week That Was; what the axis means is in The Speed of Now.
• A second gauge is born. The Stack Inversion debuts in the Bubble and Risk Scan, scoring how close the AI hardware shortage is to ending, and the Fed chair spent Wednesday asking its exact question about chip prices.
• Accountability: a call that came out flat. Venture Global reached its thesis horizon, held its thesis, and earned no edge over its sector. Scored honestly in On the Radar.

The week’s closes are in the masthead tiles above and, in full, in the Scoreboard; this page does not repeat them.

The gauge holds at 25, and this week its calm is genuine: the Fed’s hawkish words and a soft inflation print cancelled out, and the fight is not resolved, only paused.

The full working, every signal, its weight, its threshold, this week’s reading and the score it earned, is printed in Appendix A6, so you can recompute this number yourself.

Hold, drifting gently dovish: the hawkish hold and the soft inflation print cancelled, and now the data, not the Fed, breaks the tie.

Reconciliation: the market spent June pricing a September cut, then priced it out as the oil rose, and is wavering again after the soft inflation print. We treat that cut pricing as sentiment to be tested, not an input to follow; our own read comes from the model set out below, which stays at Hold and will not lurch on a single print.

Last week’s tell, scored first. We left a two-line tell on the two-year Treasury after the Fed: above 4.45 percent and the oil shock had beaten the disinflation, a hike genuinely on the table; below 4.20 percent and the meeting had reassured everyone the cut still lived. It did neither. The two-year settled at 4.23, on the fence, while the ten-year rose to 4.68. That gap is the whole message: the long end of the market took the oil-and-inflation warning and lifted, while the front end, which tracks the Fed, stayed pinned near the dovish line because the same-day inflation print was soft. Read plainly: despite three members voting for a hike, the market did not price one, and the autumn cut is not dead. It is on the fence, exactly where the tell said to look.

Two forces, held in balance

Start with the barrel, because it wrote the year even as it cooled this week. Crude is up 34 percent in 2026 and 23 percent over four weeks, and although it fell about 5 percent this week to 84.67 dollars, an oil move of that size feeds into inflation with a lag of a few months, which is why a central bank that fears the pass-through leans hawkish. That is the whole chain, oil to bond: higher crude, higher for longer, a higher two-year Treasury yield and lower long-bond prices, which is why long-dated Treasuries are the worst of the major government-bond and equity benchmarks this year. But the second force pushed the other way this week. June core inflation, the measure the Fed actually targets, came in soft, and the energy that is lifting the headline is the same energy a central bank is trained to look through. So the Fed held and talked hawkish, and the data drifted dovish, and the two-year sat between them. The reader hoping for an autumn cut should understand that this was not a defeat; it was a stay of execution.

The second story sits inside the stock market, which made a new high, and yet the calm is an average of two violent halves: the market paid up for AI with a revenue meter attached and marked down AI that is still only a cost line, while the most leveraged AI-compute names staged a violent one-day bounce that changed no trend. The sorting is told in The Week That Was and The Speed of Now; the bounce, and why it is not a breakout, is the opening scene of our new Stack Inversion gauge in the Bubble and Risk Scan. What matters for the rate story is only this: the record was made on broad participation, not a narrowing top, which is what lets the Fed stay patient.

The rate read, governed by the model. Our internal rate model, a Taylor-Rule-anchored panel we use to keep the rate call honest, reads Hold, and this week its composite fell from 14 to 8, a mild dovish drift. What moved it was the cooler June core PCE pulling the inflation block down. The discipline matters as much as the direction: the model changes its view only when two of its three blocks, inflation, labour and markets, agree, and this week inflation eased while the labour block stayed soft and the market block held, so the drift is real but gentle. The Taylor Rule benchmark it anchors on (a mechanical formula for where the policy rate should sit given inflation and unemployment) still prescribes a rate about 1.2 percentage points above the funds rate, which means a cut is not yet rule-justified, however the market prices it. The plain translation: the Fed is talking tough to make the market do its tightening for it, but the data is quietly giving the Fed room it says it does not want.

The competing explanation, named. We have told this week’s equity strength as a rotation into metered AI and away from unmetered spend, and we should say what else it could be: a broad melt-up that would leave the market more fragile, not less. The tell that separates the two is breadth. A rotation shows improving participation as money moves within the market; a melt-up shows the same narrow leaders carrying an ever-thinner tape. This week breadth improved even as the crowded names fell, about two-thirds of the market above its long-term trend, which is why we read it as a rotation. If next week breadth narrows while the index keeps rising, the melt-up read takes over, and we will say so.

Where I could be wrong (the standing self-check): First, the model we anchor on is fed by June-vintage inflation data and the oil is a July-into-August event, so if the coming prints re-accelerate the model is late by construction; the tell is a two-year yield breaking above 4.45 percent, which has not happened. Second, I have read the equity strength as healthy breadth over a wobbling top. If the megacap earnings were the last of the good news and the worry about AI spending without a return spreads, the tell is high-yield credit spreads widening through 350 basis points from today’s 284; that has not begun, and until it does the rotation read holds.

2026 thesis check-in

Seven editions on, the Repricing thesis is easy to show with a number rather than assert. Across our five reference asset classes, US shares are up around 9 percent this year, high-yield credit up around 2, the aggregate bond index down around 5, gold down around 7, and crude oil up around 34. Best to worst, that is a spread of roughly 41 percentage points, with three of the five up and two down, which on our own tightened standard, a spread above 35 points with at least two moving each way, is a strong confirmation. The honest caveat, and it matters: much of that width is still the oil shock. Strip crude out and the remaining four run from plus 9 to minus 7, a spread of about 16 points, which is only moderate. The gaps between asset classes are real and wide, exactly as the thesis said they would be; this week, one gap is doing much of the work, though less than a month ago as the oil cools.

The calm was an average of violent halves. The S&P rose about 1 percent to a fresh record and the Nasdaq 100 half a percent, but the flat weekly prints hid round trips that would have made a reader dizzy. The Nikkei traded in a 5.7 percent range to finish almost unchanged. Crude swung nearly 7 percent inside the week to close down 5 percent. Three of the AI names our readers hold, Arista, USA Rare Earth and MP Materials, each covered a 13 to 14 percent range and ended roughly where they started. When the weekly number is flat and the intraweek range is enormous, the story is not in the close; it is in the churn.

The barrel cooled but still leads. Crude fell 5 percent to 84.67 dollars as stalled peace talks took a little risk premium out, yet it is up 23 percent over four weeks and 34 percent for the year, still second on the Scoreboard behind only the Baltic Dry shipping index. An oil price here is not only an energy story; it is a rate story wearing an energy costume, and it is the reason the Fed leaned hawkish.

Megacap earnings split the market on one axis. Microsoft added roughly 450 billion dollars of market value in a day, the largest one-day gain on record, after its Azure cloud grew 43 percent. Amazon’s AWS grew 37 percent, its fastest in eighteen quarters, and yet the shares fell because it raised its 2026 spending plan to 220 billion dollars. Apple beat for a thirteenth straight quarter but its services slowed, and it fell. The axis the whole market sorted on was simple: does the AI spending show up as revenue, or only as cost?

The rand and silver were the quiet movers. The South African rand actually strengthened about 2 percent against the dollar, unwinding last week’s slide, and silver fell about 2 percent while copper rose, the industrial metals telling a firmer story than the precious ones. Bonds were soft, long-dated Treasuries again the worst major benchmark of the year, and the two big cryptocurrencies slid with the AI names they have come to track, Bitcoin to about 62,800 dollars and Ethereum to about 1,860.

A few of the eight readings, the MOVE index, factory orders and insider selling, are carried from earlier dates, and we flag them rather than let you find them; every carry is asterisked in Appendix A6.

The week AI had to show its receipts. For two years the artificial-intelligence story ran on promise. This week it ran on invoices, and the market read them closely. The dividing line was whether the spending has a meter attached. Microsoft and Amazon could point to cloud revenue growing more than a third, spending that shows up as someone else’s bill; Apple and Meta could point mainly to their own. Underneath the earnings sat a colder number: Alphabet’s quarterly free cash flow turned negative for the first time in about twenty years, because the capital going into AI now outruns the cash coming back. None of this says the technology is failing. It says the market has stopped paying for the story and started pricing the return, which is a healthier and a more dangerous phase at once.

(The Alphabet cash-flow figure reached me through a family-office contact; the framing rests on the four companies’ own quarterly releases.)

The Strait of Hormuz, the sea lane through which roughly a fifth of the world’s oil passes, is barely open, and this week a second front appeared. The Iran ceasefire has collapsed, Washington insists the strait is open while Tehran calls it closed, and a threatened charge on cargo and a tanker squeeze have kept crude whipsawing between roughly 81 and 92 dollars all month before it closed the week at 84.67 dollars. The new element is the Red Sea: reports that Houthi forces have declared a blockade against Saudi shipping, which together with Hormuz would put 30 to 35 percent of global oil flows within reach of two simultaneous chokepoints rather than one. Two taps, not one, are now within a hand’s reach of being turned.

The discipline we hold on this story is to watch the insurance market, not the headlines, because war-risk premiums and force-majeure clauses move before prices do. The non-consensus point worth giving a reader is this: the more sophisticated view among the desks we track is not that oil is heading to 110 or 120 dollars in a demand-destroying spike, but that a partial, grinding closure that keeps crude in the mid-eighties is something else entirely, a margin tailwind for oil producers rather than a catastrophe. When underwriters raise premiums or insurers invoke force majeure, the shock is real; when they do not, the map is scarier than the market.

Beyond the barrel, the flashpoint that is not about oil is water. Munich declared a water-shortage state of emergency on 2 July, and low levels on the Rhine and Danube forced European refineries, nuclear plants and chemical works to trim output through late July. It is a slow, physical constraint on the same industrial economy the oil price is squeezing, and it belongs on the risk map even though no headline will shout it.

The cross-check that keeps the story honest. A genuine supply catastrophe would show up in the long end of the bond market and in credit spreads. Neither has ruptured: the 30-year sits around 5.21 percent and high-yield spreads are still calm at 284 basis points. The market is treating the two chokepoints as a priced friction rather than a rupture, which is also exactly what the pricing would look like the week before it was wrong, which is why the insurance clauses, not the indices, remain the tell. The rest of the chain, how a chokepoint becomes a petrol price becomes a rate decision, is the oil-to-rates argument already made in the Takeaway; the lesson of 1979, in one line, is that the central bank’s response is usually harder on markets than the spike itself.

The founder, in one archetype: the Jungle Explorer. Some founders optimise a known plot of land; the Jungle Explorer opens ground the incumbents assumed could not be a business at all. Kraken’s bet is that the uncrewed warship, a boat with no one aboard, is not a science project but a category, and that a small British firm can build one at scale before the big defence contractors decide it is worth building.

The problem, up front. A navy that wants to watch a stretch of ocean has, for a century, had two options: send a crewed warship, which is expensive and puts sailors in harm’s way, or send nothing and hope. Kraken Technology Group, a British builder of small, fast, uncrewed surface vessels, exists to sell a third option, a boat that can patrol, hunt mines and carry sensors for a fraction of the cost of the ship it replaces, with no crew to lose. The demand is not theoretical: the same Hormuz and Red Sea chokepoints elsewhere in this edition are exactly the kind of contested water where a navy would rather risk a machine than a crew.

The bet, and the tell that it is being made. A promising defence startup becomes a real one at a specific moment, and Kraken reached it this year: a Series B raise, and its first senior American hire, a US chief executive, the classic signal that a British firm is about to try to sell into the world’s largest defence budget. Building the boat was the easy part; drones at sea are proven. The hard part is the one every category-opener faces, persuading a conservative buyer to trust a machine with a job that used to require a crew, and doing it at scale before a better-funded rival arrives.

The moment of real uncertainty, and the number under it. The strategic knife-edge is not whether the technology works; it is whether Kraken can build at the scale a real contract demands. And here sits the uncomfortable fact, stated plainly because a case study that hides it is a press release: the company’s own stated build rate has been reported at both roughly one thousand and roughly two thousand vessels a year, and the gap between those two figures is the difference between a workshop and an industry. We will not paper over it with a number we cannot yet verify against the company’s filings; the build rate is the whole thesis, and until it is pinned down the thesis is a promise, not a fact.

What it teaches. The transferable lesson is the shape of a category-opening bet. The pioneer’s advantage is not a better product; it is arriving before the buyer knew it wanted one. The pioneer’s risk is that being early and being too small are the same failure: a first mover that cannot build at scale simply teaches the incumbent where the market is and then loses it to them. Kraken’s story is not yet resolved, which is exactly why it is worth watching now rather than after the fact.

A note on the numbers: this profile is deliberately light on figures because the load-bearing one, the annual build rate, is reported inconsistently and has not yet been reconciled against Kraken’s own filings; we would rather flag the gap than publish a figure we cannot stand behind. This is a strategic case study of a category-opening private company, not a recommendation, and the publication holds no position.

This new column takes one phrase that does more work than it looks and holds it to two questions: what is the mechanism underneath it, and what is the number. This week’s phrase is the one people reach for to wave away the cost of tariffs: a tariff is a one-time rise in the price level, not inflation. Concede the point, because on the narrow definition it is true. Inflation is the rate at which prices keep rising; a tariff lifts import prices once and, by itself, does not create a self-feeding spiral, so a central bank can rationally look through a one-off shock. Most economists, the Federal Reserve and the Congressional Budget Office included, would accept that sentence.

The mechanism, and the number. The trouble is that “one-time” only holds if the tariffs themselves are one-time, and in 2025 they came in round after round: stack the steps and you get a staircase, not a step. Worse, the cost arrives slowly. Federal Reserve research finds it builds over roughly seven months to full dollar-for-dollar pass-through, a dollar of tariff becoming a dollar on the shelf, and through that whole year-plus window it is indistinguishable from inflation to anyone paying the prices. The measured 2025 figures are not the small 2018 ones, because the tariffs are far broader: core goods prices up 3.1 percent, which the Fed found to be the entire excess in goods inflation over the pre-pandemic trend; about 94 percent of the cost borne by Americans rather than foreign exporters, on the New York Fed’s estimate; roughly 0.7 to 0.8 points added to headline inflation, so that a reading near 3 percent last autumn would have been about 2.24, essentially at target, without them; and a cost to the average household of about 1,000 dollars a year, with estimates for the fuller proposed programme running past 1,700.

Where the risk was relocated. The phrase hides two things. First, the promised payoff, reshoring, has not landed: tariff revenue roughly doubled and companies pledged over 1.5 trillion dollars of investment, but a Kearney study found no significant near-term reshoring and manufacturing employment was contracting. Second, the politics of the claim run backwards from how it is usually told. In June 2024 sixteen Nobel economists warned that the tariff-and-tax programme would reignite inflation, and the administration publicly attacked the New York Fed study that measured the cost, so “the experts say tariffs are not inflationary” is the opposite of the record. The steelman is still real: if the tariffs stop escalating and the Fed refuses to accommodate them, the impulse does fade within a year, which is why the CBO expects the effect largely spent after 2026.

The transferable lesson. Whenever you meet a sentence of the form “it is just X, not Y”, ask two questions: is X really one-time, and over what window does it feel like Y anyway? A true definition can still describe a false comfort. That is the whole method of this column, and it works on a fund’s pitch as neatly as on a policy’s defence.

(Every figure here is drawn from the primary Federal Reserve, New York Fed, St. Louis Fed and Yale Budget Lab sources; the household-cost range reflects scope, the enacted tariffs at the low end and the fuller proposed programme at the high.)

Some of the most useful books about risk are not about markets at all. Jon Krakauer’s Into Thin Air is a first-hand account of the 1996 Everest disaster, in which eight climbers died in a single storm, and read sideways it is the finest short course in decision-making under pressure a market participant can find.

Three lessons transfer cleanly. Summit fever, the pull to push on for a goal that is finally in reach, is the mountaineer’s name for the sunk-cost trap and for greed, and it kills more climbers than weather does. The turnaround-time rule, a fixed hour after which you descend whether or not you have reached the top, is a disciplined stop-loss written down before emotion can argue with it, and on Everest the deaths clustered among those who broke it. And disaster, Krakauer shows, is almost never one catastrophic error; it is a cascade of small, individually defensible decisions that only look reckless once they have compounded.

A reader who has watched a portfolio unravel will recognise every beat: the position held too long because selling would admit a mistake, the rule abandoned because this time felt different, the small concessions that each made sense and together proved fatal. The mountain does not care about your thesis, and neither does the market. What saves you in both places is the discipline you set before the pressure arrived, and the humility to follow it when every instinct says push on.

The discipline that saves a climber is the one written down before the summit was in reach, and it is the same discipline that saves a portfolio.

Each week the WMP scores the central economic argument about artificial intelligence as a contest between two forces. The Displacer is the case that AI substitutes for human labour, concentrating the gains in capital and eventually destroying the spending power the economy runs on. The Augmenter is the case that AI raises human productivity, expands output, and spreads the gains broadly. Both sides agree AI is powerful; they disagree about whether it is a demand shock or a supply shock.

• Pillar 1, adoption speed, the Augmenter. The earnings axis told in The Week That Was is the evidence: the market rewarded metered artificial intelligence and punished spending without a meter. That is monetisation discipline, adoption where it pays for itself, rather than a runaway substitution.
• Pillar 2, labour, the Augmenter. No named white-collar layoff wave citing AI as the cause surfaced this week, and the better reading of the evidence is that AI is compressing wages rather than cutting headcount, a complement with a price effect, not a replacement.
• Pillar 3, type of shock, the Displacer. The Displacer’s clearest point in weeks: the Alphabet cash-flow turn reported in The Speed of Now, set against an industry that needs trillions of revenue it does not yet have against roughly 150 billion dollars today. Capital is racing ahead of the demand that is meant to justify it, which is the demand-destruction risk in embryo.
• Pillar 4, compute cost, the Augmenter. That same cash burn, Amazon lifting its 2026 spending plan to 220 billion dollars, and power emerging as the binding input all say the ceiling is cost. When even the Fed chair is watching chip prices, the price of building AI, not the will to build it, is the constraint.

What would flip the score: a named AI-attributed white-collar layoff wave would hand the Displacer the labour pillar; a collapse in memory-chip prices that dropped the cost of compute below the cost of a worker would hand it the compute pillar; and a negative July retail-sales print would hand it the economic-shock pillar outright.

The ERDR (Equity Return for Debt Risk) framework tracks twelve income strategies that aim to earn an equity-like return for taking on debt-like risk. The twelve-strategy Standing Dashboard runs every week, each strategy refreshed with its current yield, its spread over investment-grade credit, and any change in thesis; a Deep Dive on a single strategy joins it from time to time.

The one macro cross-current worth naming this week is the same rise in long-term interest rates that is hurting long bonds: it is lifting the yields available across most of the income complex, from senior secured loans to private credit to preferred shares. The strategy to watch is private credit direct lending, where the AI-financing story in the Bubble scan is starting to matter: as data-centre operators borrow against fast-depreciating chips, the quality of the collateral in parts of the direct-lending market deserves a closer look than the headline yield suggests.

The same rise in long-term rates that makes long bonds the worst line on the Scoreboard is quietly lifting the yield on almost every income strategy below.

Each strategy is explained in full when it is the week’s Deep Dive.

The read this week: the oil shock left income markets remarkably alone, high-yield spreads at 284 basis points, the long end firm, so the dashboard’s moves are notes rather than signals. The two standing cautions are unchanged. The leveraged-loan and private-credit complex (strategies 6 and 11) still prices more dispersion in the market than the rating agencies admit. And the energy-linked royalty strategies gave a little back as crude cooled this week; the year’s windfall should still be read as volatility, not run-rate. No thesis changed this week.

What I am watching and why, not a recommendation to buy or sell. The scoring and ledger material lives here, in its own block, and does not lead the edition. No company cleared the strict new-catalyst test this week, so nothing new is added; the section does its other job, holding the open calls to account. One call reaches its thesis horizon and is scored in public.

No new call this week, by the rule, not by accident. Our test for putting a company in this section is strict: a named, company-specific event in the past seven days that has not already been used. This week there was none that cleared it, so nothing new is added, and the section does the other job it is built for, holding the open calls to account.

Venture Global (NYSE: VG): scored at its thesis horizon, and it came out flat

Venture Global (NYSE: VG) · twelve months · entry $13.08 (Wk 16) · close $13.38 (31 Jul) · thesis-horizon date 2 August.

We flagged Venture Global in Week 16 on a simple thesis: as Europe re-sources gas, the scarce thing is the American terminal that can ship it, and Venture Global has the most spot exposure among US exporters. This weekend it reaches its declared horizon at 13.38 dollars, above the 13.08 we flagged, so the condition that would have proved us wrong, a fall below entry with Asian gas failing to support margins, did not fire, and the structural read held. But honesty is the point of this section: over the same window the energy sector it belongs to, measured by the XLE fund, returned about the same, so the call earned essentially no edge over simply owning energy. Thesis right, no alpha. Rubric verdict: a flat result, and we log it as one, at zero on the five-point scale.

A correct thesis that matches the sector it belongs to is not a win; it is a reminder that being right and being paid are different things.

The running record, with denominators. With Venture Global closed, four On the Radar calls have now reached their thesis horizon: Bloom Energy, Freeport-McMoRan, Talen Energy and Venture Global. The analysis read the world broadly right on all four, but every one trailed or merely matched the benchmark it was measured against, by roughly one, thirteen and six points behind on the first three and level on Venture Global. That is the honest shape of the record so far, four theses that read the world correctly and none that beat the simpler thing you could have held instead. Zero double-misses. These are the calls logged in our call record since Week 13. From this edition the four-week tactical marks are tracked internally and no longer published week to week, because this is an investor’s letter, not a trader’s post-mortem; the full ledger, both horizons and every denominator, is published in the Quarterly Reckoning.

Two losers reach their horizon soon, pre-flagged. MP Materials and USA Rare Earth both reach their thesis dates in the next two-to-three weeks, on 9 and 16 August, and both are deep losers in absolute terms, down 38 and 29 percent from where we flagged them. The honest frame, and the one the Freeport lesson taught us, is to measure them against the sector: the rare-earth basket fell about as hard, so USA Rare Earth is actually a hair ahead of its benchmark and MP a hair behind. The miss, so far, is the sector, not the selection. We score them in full when they close, with the same prominence whether they land well or badly. They and the other open calls, Arista, YPF and Mitsubishi UFJ, are tracked in Portfolio Watch.

Spare a thought this week for the enduring optimism of the leveraged bull, cheering Thursday’s violent bounce in the AI-compute names from somewhere well below where he bought them. It is the financial equivalent of celebrating one warm afternoon in the middle of a long winter, and then telling your friends the season has turned. Charlie Munger, whose line opens this edition, had a name for the third of his three ways to go broke, and it was not the weather.

Three things to watch next week. First, the July jobs report next Friday, the first hard read on whether the labour market is soft enough to force a cut past the oil. Second, the two-year Treasury at Friday’s close, which settles the fence tell set out in the Takeaway. Third, the price of memory chips, the one dial on our new gauge that would tell you the AI shortage is genuinely ending rather than merely bouncing. If all three move the same way, weak jobs, a falling two-year, memory prices rolling over, the story flips from higher-for-longer to a cut-and-a-cooling in a single week. If they diverge, the standoff holds and we sit on the fence a while longer. We will be here to score it.

Until next week. Stay curious and stay hedged.
Anthony Rosenthal

Crude and the Baltic Dry are not just the year’s best performers, they are the wildest, swinging two to three times as hard as the S&P; the gap between best and worst is 82 points, and one shock, oil, is still doing much of the widening even as it cools.

26 assets ranked by year-to-date return · baselines locked 1 January 2026 · close of Friday 31 July 2026. The basket is a fixed set, chosen on 1 January and unchangeable during the year: twelve equity indices, four bond and credit funds, six commodities, two currencies and two cryptocurrencies. Nothing else is added, dropped or substituted mid-year, which is the only way a year-to-date table means anything. The single-company calls in Portfolio Watch are tracked separately. Annualised volatility shows how much each line typically swings in a year, judged from its last eight weeks: higher means bumpier, not worse, and a dash means we do not yet have enough weeks to measure it honestly.

Every close here is drawn from the same price record the table itself reads from, so the numbers on this page and the numbers in our record cannot drift apart. All closes are Friday 31 July 2026. MSCI EM is derived from the EEM ETF close (64.09) multiplied by the locked index ratio (28.367); Baltic Dry is the Baltic Exchange BDI as published by Hellenic Shipping News (31 Jul, 2,732). Every year-to-date figure is recomputed from the locked 1 January baselines rather than carried forward, so an error cannot compound week to week.

Every company that has appeared in On the Radar is tracked here until its formal score date. A company moves from On the Radar to this appendix when there is no fresh catalyst that week, the analytical call is intact, but there is nothing new to add. Every close below is the verified close we recorded on the date shown beside it. One name has left the table this week: Venture Global closed its thesis window and received its final verdict in On the Radar.

This week: five active calls tracked here, with Venture Global closed at its thesis horizon and scored in On the Radar. Three of the five here are in the money (Arista, MUFG and YPF) and two are behind, the losers named as loudly as the winners. MP Materials, down 38 percent, is the book’s biggest loss, and USA Rare Earth sits on the same repriced rare-earth thesis, though both are roughly in line with the sector that fell with them. The marks come quickly now: MP on 9 August, USA Rare Earth on 16 August and Arista on 28 August.

You do not need to read it. It is here so that you can.