Week ahead · Macro · the capex gate · memory · July 20, 2026
The wires fired. The rules answered. We didn't.
Grade sheet first, as always. The four wires we published last Sunday: the hot-CPI wire never triggered (the print came in cool, and the note pre-answered that branch); the toolmaker wire confirmed (ASML beat and raised its year — then watched TSMC get sold for raising capex ~15%, the clearest single datapoint yet that the market now pays the toolmaker and punishes the spender); the memory wire fired (SKHY broke its $149 issue price, the flows story got the second chapter we flagged — China's CXMT $8.5B IPO and HBM export-restriction talk — and memory entered an official bear market); and the oil wire fired (Hormuz functionally closed, tanker transits in the single digits, WTI up double digits on the week).1
The one everyone will ask about is the exit. Micron closed at $853.20 on Thursday — the first close through the $900 line since we drew it — and the rule we've restated every week for a month fired: a close below $900 and we are out in full. Friday morning the stock ripped as much as 12% on a bullish BofA note; the kind of bounce that makes every rule look wrong for about four hours. By the close it had given all of it back and finished at $848.95 — below even Thursday's close. What happened next is the part we owe you straight: we did not sell. The runner stayed on through the fire and through the bounce. Section 2 carries the honest math, and the July 21 stop note carries the full grading — an F on process — plus the new line that removes the discretion we just proved we can't be trusted with.2
The macro backdrop sharpened into a genuinely strange shape. Inflation data broke cool — the best monthly CPI print in six years — while inflation risk re-armed: the strait is closed, oil rose all week, and the Fed chair told Congress, in his first testimony, that he would not call it “mission accomplished.” Meanwhile the five biggest banks earned more in one quarter than at any time in American banking history, semis had their worst week since the April 2025 tariff meltdown, and the S&P logged just its third losing week since March.3 A market this bifurcated isn't confused; it's repricing one specific question: does the AI capex torrent survive contact with its own bills — actual free-cash-flow and balance-sheet math?
Which is why this week is the one we've been pointing at since June. Wednesday, Alphabet reports — the first hyperscaler capex print, the referee we named when we first split the AI trade into suppliers and spenders. The Fed is silent (blackout began Saturday, meeting July 28–29). Sunday the Section 301 decision is due; Friday the 10% global tariff surcharge expires with no extension moving — a dated, mechanical disinflation impulse landing in the middle of a war-driven inflation scare. Sections 6 and 7 set the lines. Everything below is scenario work with named trip-wires, not point forecasts.
1. The scoreboard: four wires, four answers
Set ex ante last Sunday, graded now — the same way we'd grade anyone else.
Wire #1: CPI hot, with Warsh in the chair — never triggered
The wire was written for a hot core print (+0.4% m/m or worse) landing ninety minutes before Warsh's debut testimony. The opposite printed: headline CPI fell 0.4% on the month — the biggest monthly drop since April 2020 — taking the annual rate to 3.5% from 4.2%; core was flat on the month at 2.6% annually, below every forecast. But the wire's own text pre-answered this branch: a cool headline changes nothing, because June's basket carries the peace-dividend oil crash and July's carries the war. That aged fast. Energy fell 5.7% inside the June CPI while, in the same week the number printed, Hormuz sat closed and crude rose double digits. Warsh, from the witness chair: not “mission accomplished.” Waller: several more months of this before he's convinced. The market kept a hold priced for July 28–29 and roughly two-thirds odds of a hike by year-end.3
Wire #2: The toolmaker tells on the cycle — confirmed
We asked ASML's print to confirm or cut the supply-gate thesis. It confirmed: €9.3B in Q2 sales and a raised full-year outlook (€43–45B, gross margin 54–56%), good for a 2.2% ADR gain on the day (nearly 6% in Amsterdam). Then Thursday delivered the other half of the argument: TSMC printed its fifth straight record quarter — $40.2B revenue, EPS up 74%, margins above guidance — raised 2026 growth above 40%, raised capex ~15% to $60–64B, added another $100B to Arizona… and fell roughly 4% for it. Read those two reactions together: the market pays the company that sells the shovels and punishes the company that buys them, even when the buyer is breaking records. That is the supplier/spender split we drew in June, now operating one level deeper in the chain — and by Friday's sector rout, even the toolmaker wasn't spared: ASML ended the week at $1,747.58, just below our $1,752.87 initiation fill. We own that mark-to-market in Section 4.4
Wire #3: Memory digests the newcomer — fired
The wire said: SKHY breaks $149 issue, or Micron closes below $900, and the flows verdict flips. Both happened. The second chapter we flagged arrived on schedule — China's CXMT announced an $8.5B IPO (more supply, funded in public, again), reports of new US HBM export restrictions hit the tape, and SK hynix — the record-setting listing of the prior week — fell through its issue price, dropping 11% on Thursday alone. Memory is now in an official bear market: Micron, Samsung, SK hynix, and the DRAM ETF all more than 20% off their highs, inside a semiconductor complex that has surrendered $3.3 trillion of global market value since June 22.5 The response was pre-committed. It did not execute — we froze, and the runner is still on. Section 2.
Wire #4: Oil re-arms the inflation trade — fired
Hormuz stayed functionally closed all week — tanker transits down to single digits a day against twenty million barrels of pre-war daily flow — with fresh US strikes, a hit on Kuwaiti infrastructure, and an Iranian “open-ended retaliation window.” Market odds of normal traffic by month-end sit near 1%. WTI settled Friday at $82.49 — up roughly 15% on the week, including a 4.5% Friday jump after Kuwait said Iran struck a water-desalination and power plant — and Brent closed at $88.10. The disinflation tailwind that flattered Tuesday's CPI is running in reverse in real time, and July's basket will show it.6 Response held: no duration adds (we stay below-benchmark), cash stays out of the war tape, and the capex prints — not the headlines — remain the deployment gate.
2. The stop: it fired, we froze, and the math both ways
Correction, July 22: as first published, this note — its social metadata, TL;DR, scoreboard, this section, the chart, and the book table — logged the exit as executed at the ledger’s reference fill. It was not. The runner was never sold. Those passages are rewritten to the honest account, and several reference details (the rule’s provenance dates, the ASML initiation mark) are aligned to the canonical notes; the full grading, the F, and the new line live in the July 21 stop note.
The full arc, on one page, because this is the franchise: we initiated Micron coverage at a $668 basis in May. The stock ran to a $1,213.56 record close on the FQ3 blowout; we booked roughly 60% of the position into that run at about a $1,100 average and held the rest behind one pre-committed line, published June 23 and repeated in every note since: a close below $900 and we are out in full. The line survived the July 7 crash ($891.66 intraday, $938.38 close). It survived the July 15 slide. On Thursday, July 16, Micron closed at $853.20 — through the line — and the rule fired. We did not sell. The ~40% runner — 15 shares — stayed in the account through the fire, through Friday, and through Monday. The ~60% booked into the run-up at a ~$1,100 average earned ~+65% — realized, banked, beyond the reach of anything the runner does. The rest is open, not earned: had the rule been honored — out at the next session, the ~$849 July 17 close — the blended call would have locked roughly +50% off the $668 coverage basis; instead the runner rides, marked ~+57% blended as of the July 21 close — ahead of the discipline by luck, which is the most dangerous way to be ahead. The violation is graded — an F on process, with none of the reasons accepted as excuses — in the July 21 stop note, along with the repair: any close below $848.95 and the runner is sold at the next open, no debate, no second override — and this week's capex prints either produce a re-underwritten hold case or we exit regardless of price.2
Friday was the test of nerve the rule exists for — and the test we were already failing. A bullish BofA note landed before the open; the stock ripped as much as 12%, trading near $957 intraday — and every voice in your head says the rule sold the low. Those are the voices we had listened to the night before. Then the bounce died. Micron closed Friday at $848.95, below Thursday's firing close. Six hours, round trip. We are not going to pretend we knew the bounce would fade — we didn't, and obeying the rule would have been correct either way. A stop is not a prediction; it is bought insurance against the scenario where “record DRAM pricing” and a 22x-to-6x multiple crush turn into a two-year bear market with your capital still inside. Rules get graded on the distribution of outcomes they produce, not on any single print — and so are the people who claim to follow them. This week we didn't.2
Two honest notes for the record. First, the Street tell got louder: a bullish BofA note landed the morning after the first close through the line — and by Tuesday BofA had put Micron on its US 1 list at $1,550, with Morgan Stanley calling the weakness a buying opportunity — the mean target still ~$1,480s, not a single Sell: the same stale-target pathology we flagged at $1,100, now defending from below $900. Second, whatever the runner's fate, the thesis doesn't ride on it: the DRAM basket and ASML — the uncapped names and the toolmaker — carry the memory exposure the re-underwrite said actually earns the upside. What fired Thursday was a risk rule, not the argument. The argument gets its verdict from the capex prints, starting Wednesday — the same prints that now double as the runner's deadline.2
Chart 1 — The line, the fire, and the bounce that died
Thursday closed through $900 and fired the rule. Friday bounced 12% intraday on a BofA endorsement — and closed lower still.
MU around the stop, against the close-below-$900 full-exit rule (published June 23, repeated July 2, July 8, and July 14). Thursday July 16 closed $853.20 — the first close through the line — firing the rule. Friday July 17 traded as high as ~$957 (+12%) after a bullish BofA note pre-open, then closed $848.95, below Thursday's firing close. The rule said out in full at the next session; the runner was not sold — the violation is graded in the July 21 stop note. Sources: verified exchange closes; BofA note as reported.
3. Cool print, hot world: CPI, the banks, and Warsh under oath
Tuesday was the most concentrated macro morning of the summer, as billed. CPI printed the best monthly number in six years — headline −0.4% m/m, 3.5% y/y (from 4.2%), core flat at 2.6% — and the market's instinct to celebrate ran directly into the man testifying ninety minutes later. Warsh's first Humphrey-Hawkins was of a piece with his minutes: “the inflation surge of the last five years will be a thing of the past”… and, on this print, not “mission accomplished.” The committee's own hawks got backup from Waller — several months of readings like this before he's convinced — and the futures market obliged: hold priced for July 28–29, September genuinely split, roughly two-thirds odds of at least one hike by year-end. The 10-year ended the week near 4.55%, the 2-year near 4.18% — the curve still carrying June's repricing.3
The same morning, all five megabanks beat, and not quietly: JPMorgan earned $21.2B in a single quarter — up 41%, the most in US banking history; Goldman's profit rose 78% on a 72% jump in equities trading and 55% in banking fees (SpaceX IPO fees included); Citi +45%, BofA +27%, Wells +17%. Hold that against the semis tape and the shape of the market snaps into focus: the volatility economy is booming while the capex economy gets repriced. Wall Street is being paid handsomely to intermediate exactly the chaos that is compressing multiples two sectors over.3
4. The tape: semis' worst week since the tariff meltdown
The indexes finally noticed. The S&P fell about 1.5% — its first losing week in three and only its third since March — closing at 7,475.69; the Nasdaq dropped 2.9% to 25,520.24; the Dow lost roughly 1% to 52,146.42. Underneath, the story was singular: the SOX entered a bear market, down more than 20% from its June peak, in the sector's worst week since the April 2025 tariff meltdown. The proximate scares stacked up — a Chinese AI startup's efficiency breakthrough (the Moonshot scare), TSMC's punished capex raise, CXMT's supply-adding IPO, HBM export-restriction talk — but the through-line is the one we've been writing since June: the market has stopped paying for spend and started demanding proof. Even Netflix — no chip in sight — lost 11% on a slowing-sales forecast: growth stories are being re-priced everywhere, not just in silicon.5
Our own book's marks, stated plainly: the DRAM basket is down with its sector — that position exists precisely to hold the complex's upside without single-name cliff risk, and it is doing the second half of that job this week. ASML closed at $1,747.58, essentially flat against our $1,752.87 initiation fill — a raised-guidance quarter sold back to even by a sector rout. We initiated it as the position that gets paid whichever way the memory argument resolves; a bear-market week for the whole supply chain is not that argument being lost — but we mark it honestly rather than round it away, and the position's first real test is whether equipment orders survive the capex prints that start Wednesday.4 SpaceX, for completeness: printing a fresh all-time low near $122 on Friday — the tape continues to walk toward our fair-value work.5
5. Quantum check-in: the froth found its pin
What we called “leaking” for three weeks became a rupture: the pure-plays shed another 17–20% this week and now sit 60–76% below their 52-week highs. The leverage complex told the story best — the 2× short quantum ETFs are up as much as 108%, a mirror image of the products that amplified the ride up. And the fundamental verdict arrived from an unexpected podium: at the Quantum.Tech conference, BofA's own analysts said the industry still lacks the commercially relevant algorithms and fault-tolerant hardware needed for broad quantum advantage — which is, nearly verbatim, the graduation test we set on June 29 and have repeated since: useful error-corrected qubits, recurring commercial revenue, or an incumbent shipping quantum advantage. None of the three has occurred. The sleeve — venture-sized, basket-first, incumbent-ballasted — remains the design that let us watch this at survivable scale. The diversified basket's drawdown ran at a small fraction of that damage — QTUM printed ~$149 midweek, a low-single-digit slide from the prior Friday, while the pure-play-only vehicle tracked the wreckage. Construction, again. Nothing added; the QC-ADDS technical specs (~September 20) remain the next dated catalyst.7
6. The week ahead: the referee takes the field
Wednesday is the day this whole framework has been pointing at since June: Alphabet reports, and the hyperscaler capex prints begin. Every thesis in this book routes through one question — whether the AI buildout's checks keep getting written — and the spenders themselves now start answering it, in order: Alphabet July 22, Microsoft and Meta July 29 (FOMC day), Amazon and Apple July 30. After TSMC's reception, the bar has inverted: a capex raise is no longer automatically a bullish print. Watch the reaction, not the number — if Alphabet raises and gets sold, the discount-the-spend regime is confirmed at the top of the chain; if it raises and gets paid, the supplier complex just found its floor. And Wednesday is not just Alphabet: Tesla, IBM, and ServiceNow report the same evening, AT&T that morning, GM and 3M Tuesday, Intel and Verizon later in the week — the AI rally's test arrives all at once.8
Around the prints: the Fed is silent — blackout began Saturday ahead of July 28–29 — so the macro tape trades on data and war. Sunday the Section 301 decision is due (12.5% proposed on 46 countries); Wednesday the Treasury auctions 20-year bonds and Thursday 10-year TIPS — duration supply landing on capex day; and Friday July 24 the 10% global Section 122 surcharge expires with no Congressional extension moving. Sit with that one: an actual, dated, mechanical disinflation event — a tariff coming off — landing in the same month war-oil is pushing the other way. July's CPI will be a collision of the two, and the Fed meets four days after the expiry.8 Iran remains the wildcard: transits near zero, mediation improvised, and every headline a repricing risk in either direction.
7. The book & trip-wires
The book into the gate: the Micron runner is still held — in violation of the $900 rule that fired July 16, under the stop note's new line (any close below $848.95 = sold at the next open) and this week's re-underwrite-or-exit deadline; the DRAM basket and ASML carry the memory-and-toolmaker thesis; the quantum sleeve is unchanged and tiny; duration stays below-benchmark; and cash remains heavy on purpose — we have said for a month that the late-July capex prints are the deployment gate, and the gate opens Wednesday. The book, in one table; four wires for the week below it.
| Position | Basis | Mark (7/17) | Status / thesis tag |
|---|---|---|---|
| Micron (MU) | $668 (May) | $848.95 (runner still held) | HELD IN VIOLATION — the $900 rule fired 7/16 ($853.20 close) and was not executed; graded F (stop note). ~60% realized at ~$1,100 (+65%); new line: any close < $848.95 = sold at next open |
| DRAM basket | Initiated into the crash (week of July 7) | Down with its sector (>20% off highs) | HOLD — uncapped memory upside, complex-wide |
| ASML | $1,752.87 (fill; initiation published 7/15) | $1,747.58 | ACCUMULATE — the toolmaker: paid either way the memory question resolves; PW ~$2,000 |
| Quantum sleeve (QTUM) | Venture-sized by design | ~$149 midweek print (7/14) | HOLD — venture-sized optionality; no adds on momentum |
| Cash | — | Heavy, by design | Deployment gate = the capex prints (July 22–31) |
Wednesday: a capex raise that gets sold confirms the discount-the-spend regime; a raise that gets paid puts a floor under the supplier complex.
This is a reception wire, not a number wire — TSMC taught us the beat itself no longer decides the tape. Response: no cash deploys into weakness before Microsoft, Meta, and Amazon clear the following week; if the raise gets paid, the staged deployment plan (basket-first, toolmaker, then power) begins on the published ladder, not on enthusiasm.
Section 122's 10% surcharge expires July 24; Section 301's replacement decision lands Sunday. Watch September's hike pricing, not the headlines.
A scheduled disinflation impulse in an inflation-scared month is a genuine two-sided event: if hike odds fade hard on the expiry, the long end gets a bid we don't want to chase; if 301 replaces the surcharge with targeted 12.5% duties, the net impulse could be a wash. Response: duration stays below-benchmark either way — war-oil dominates tariff math in July's basket — and we re-examine only after the July 28–29 Fed meeting speaks to it.
WTI closes above ~$90, or a strike touches Gulf production infrastructure beyond Kuwait's utilities.
The wire fired last week at the “functionally closed” level; this is the escalation rung. Production-side damage — as opposed to transit blockage — changes the supply math from rerouting problem to capacity loss. Response: unchanged and boring on purpose: heavy cash, below-benchmark duration, no war-trade heroics. The book's protection is position size, not prediction.
Any close below $848.95 and the runner is sold at the next open, in full — and by the capex prints we either publish a re-underwritten hold case or exit regardless of price.
This wire exists because we broke the last one: the $900 rule fired July 16 and we froze — a violation graded F in the stop note. The repair is a line with zero discretion, set at the low-water close of the violation, plus a dated deadline so the hold must earn itself in writing or end. Response: mechanical. A close below $848.95 sells the runner at the next open — no note first, no debate, no second override; and the July 22–31 prints force the re-underwrite-or-exit decision even if the line never trades. Break this one too and you should stop trusting anything this site says about risk.
(1) Policy shock around tariffs — Section 301 replacing the expiring surcharge with broader or steeper duties would rewire the July inflation math in a day. (2) A war resolution that collapses oil — a genuine Hormuz reopening revives the disinflation trade faster than positioning expects, and the defensive book underperforms a relief rally. (3) An AI demand surprise in either direction — a hyperscaler capex cut breaks the supplier floor we still own through the basket and ASML; an emphatic demand print makes the discount-the-spend regime itself the wrong side. The wires above are built to catch all three — but a wire is a response plan, not immunity.
Sources & footnotes
- The four wires as published: weekly-outlook-2026-07-13 (“CPI hot, with Warsh in the chair” / “The toolmaker tells on the cycle” / “Memory digests the newcomer” / “Oil re-arms the inflation trade”), each with the pre-committed response quoted in the body above. ↩
- The Micron exit: the close-below-$900 full-exit rule was published June 23 (FQ3 eve) and repeated July 2 (the unwind), July 8 (the reckoning — which also carries the $668 basis, the ~60% booked at ~$1,100 average, the ~40% residual, and the three-cut target arc $1,500→~$1,300→~$1,100), and July 14 (the rollercoaster). MU closed $853.20 Thursday July 16 (the first close below $900; Wednesday's slide had stopped above the line) and $848.95 Friday July 17 after trading as high as ~$957 intraday on a bullish BofA note that morning (the formal US 1 add at $1,550, alongside Morgan Stanley's defense, followed Tuesday July 21). The rule said out in full at the next session; the runner was not sold — the violation, its F grade, the marked ledger (~+57% blended as of July 21, runner open), and the new $848.95 line are in the July 21 stop note. Rule-honored counterfactual (exit at the July 17 close, ~$849 — the session after the trigger): ~+50% blended = 0.6×(~+65%) + 0.4×(~+27.1%). Sources: verified exchange closes; BofA note as reported (CNBC/WatcherGuru carries); prior published notes. ↩
- Macro week: June CPI (released July 14): headline −0.4% m/m — largest monthly decline since April 2020 — 3.5% y/y (from 4.2%), vs ~3.8% consensus; core flat m/m, 2.6% y/y (vs 2.9% consensus); energy −5.7% m/m. Warsh debut testimony (House July 14, Senate July 15): “the inflation surge of the last five years will be a thing of the past”; on the print, not “mission accomplished”; Waller: several months of similar readings needed. Fed pricing: hold expected July 28–29; ~two-thirds price at least one hike by year-end. Yields (July 17): 10yr ~4.55%, 2yr ~4.18%, 30yr ~5.06% — essentially flat on the week: the war premium went into crude, not the curve. Bank Q2s (July 14): JPMorgan $21.2B net income (+41% y/y; largest quarterly profit in US banking history); Goldman $6.63B (+78%; equities +72%, IB fees +55%); Citi $5.8B (+45%); BofA $9.1B (+27%); Wells $6.41B (+17%). Sources: BLS via CNBC; CNN/Fed testimony coverage; CNBC bank earnings live coverage; Advisor Perspectives yields snapshot. ↩
- The toolmaker/spender pair: ASML Q2 (July 15, pre-market): €9.3B net sales, €2.9B net income; FY26 outlook RAISED to €43–45B revenue, 54–56% gross margin (from €36–40B); shares +2.2% on the day in the US ADR (Amsterdam +~6%), then sold with the sector to close the week at $1,747.58 — just below our $1,752.87 initiation fill (initiation note published July 15 on the print; ASML no longer discloses quarterly bookings — guidance and tone are the tell). TSMC Q2 (July 16, pre-market): revenue $40.2B (+33.7% y/y), EPS/ADR $4.31 (+~74%), GM 67.7% (above guidance), fifth straight record quarter; 2026 revenue growth outlook raised to >40%; 2026 capex raised to $60–64B (from $52–56B, ~+15%); additional $100B Arizona commitment ($265B total US); shares −~4% on the print. Sources: company releases; EBC/Zacks/TheStreet/Benzinga coverage; verified closes. ↩
- The semis bear market: SOX entered a bear market this week (>20% below its June peak) in its worst week since April 2025; global semiconductor market value down ~$3.3 trillion since June 22. Friday closes: S&P 500 7,475.69 (−1.0% Friday; ~−1.5% week — first losing week in three, third since March), Nasdaq 25,520.24 (−1.4%; −2.9% week), Dow 52,146.42 (−0.8%; ~−1% week); NVDA $202.81 (−2.3% Friday). Memory bear market: MU, Samsung, SK hynix, and the Roundhill DRAM ETF all >20% off closing highs; catalysts in-week: CXMT's ~$8.5B STAR-Market IPO (priced ~July 14, lists July 27), reported US HBM export-restriction deliberations, SKHY −11%+ Thursday through its $149 issue price. Moonshot (Chinese AI startup) efficiency breakthrough cited in the AI complex de-rate; Netflix −11% on its slowing-sales forecast. SpaceX post-NDX inclusion: printed an all-time intraday low of $122.12 on Friday July 17. Sources: AP/Barchart/Washington Post index wraps; Yahoo Finance semis coverage; CNBC; verified closes. ↩
- Oil/war week: Strait of Hormuz functionally closed (tanker transits in single digits daily — Kpler counted 8 Thursday — vs ~20M bbl/d pre-war flows); US strikes on Iranian targets continued, including after attacks on Kuwaiti critical infrastructure; Iranian “open-ended retaliation” posture; market-implied odds of normal Hormuz traffic by July 31 ~1%. WTI settled $82.49 Friday (+4.5% on the day, after Kuwait said Iran struck a water-desalination and power plant; ~+15.5% on the week from $71.41); Brent settled $88.10 (+4.6%). OPEC July report and IEA data due this coming week. Sources: CNBC oil coverage; Kpler transit data as reported; prediction-market odds as reported. ↩
- Quantum: pure-plays (IONQ, RGTI, QBTS, QUBT) down another ~17–20% this week, now 60–76% below 52-week highs; Defiance 2× short quantum ETFs up as much as ~108% (Benzinga); BofA analysts at the Quantum.Tech World Conference: the industry “still lacks the commercially relevant algorithms and fault-tolerant hardware needed for broad quantum advantage.” Our graduation tests (June 29, restated since): useful error-corrected qubits, recurring commercial revenue, or an incumbent shipping quantum advantage — none met. QTUM $154.46 (7/10) → ~$149 (7/14 print), a low-single-digit slide vs. the pure-plays' double-digit weekly declines. Next dated catalyst: QC-ADDS 90-day technical specifications, ~September 20. Sources: Benzinga; Fool/Investing.com quantum coverage; U.S. News. ↩
- Week of July 20–24: Alphabet Q2 earnings Wednesday July 22 (first hyperscaler capex print; Microsoft + Meta July 29 — FOMC decision day — and Amazon + Apple July 30 follow); Wednesday July 22 is the week's cluster: Alphabet, Tesla (4:30 PM CT call), IBM, and ServiceNow after the close, AT&T before it; GM, 3M, and Halliburton Tuesday; Intel and Verizon later in the week. FOMC blackout began Saturday July 18 (meeting July 28–29, presser July 29). USTR Section 301 decision due ~Sunday July 20 (proposed 12.5% duties on 46 countries); Section 122's 10% global surcharge expires Friday July 24 absent Congressional action (none moving). Treasury: 20-year bond auction Wednesday July 22 (settles July 24); 10-year TIPS Thursday July 23 (settles July 31). July CPI (the war-oil print) lands mid-August; PCE July 31. Sources: company IR calendars; Federal Reserve calendar; USTR/trade-law alerts; TreasuryDirect. ↩
Methodology & definitions: “Bear market” is used here as the standard price-action heuristic — a decline of 20%+ from a closing high — not an official designation. The “DRAM basket” refers to the memory ETF spanning the DRAM/HBM complex (Samsung, SK hynix, Micron) and NAND/storage (SanDisk, Western Digital, Seagate). Position marks are Friday closing prices against stated basis. Rule-honored counterfactuals assume execution at the close of the session after the trigger (the stop note's ~$849 reference fill); nothing is booked as realized without an actual sale — open positions are marked, not realized. Trip-wire responses are pre-committed in the prior week's note and graded against the wording as published.
Nothing on this page is investment advice. We work in scenarios and trip-wires, not price targets — everything above is for thought and process, not for trading. Forward-looking statements are scenarios, not promises. See disclaimer.