Saturday, May 23, 2026

Your Saturday morning market coffee. The week NVIDIA reported the largest semiconductor corporate quarter in history — $81.6 billion — and its stock fell ~3.4%; SpaceX filed the most anticipated S-1 since Facebook with a $2 trillion valuation; Walmart CEO confirmed tariff price increases begin in late May, sending shares down 6.2%; the FOMC's April minutes signaled rate hikes are now on the table rather than cuts; oil fell 7% on Iran deal optimism that proved to be, in part, a "bogus report"; and the Dow Jones set a new all-time record of 50,579 — the first time it has ever closed above 50,500 — heading into Memorial Day.

Week in Review

The Numbers

The S&P 500 logged its eighth consecutive weekly gain — the longest winning streak since 2023. But the index that tells the real story is the Dow: +2.1% on the week to a fresh all-time record of 50,579.70, driven by defensive and industrial rotation rather than AI names. NVIDIA — the company that made the prior seven weeks — spent Thursday and Friday declining despite reporting the single largest semiconductor quarter in corporate history. The market's message was not subtle: perfection is the baseline, not the achievement.

The Story Arc

Monday May 18 opened into a bond market that had not fully processed the prior week's double inflation shock. The 10-year Treasury yield rose above 4.60% — its highest level in a year — and the S&P 500 tech sector fell more than 2% intraday before recovering to close -1.1%. The S&P 500 ended the day down 0.07% to 7,403.05; the Nasdaq fell 0.51% to 26,090.73; the Dow was the sole winner, rising 0.32% to 49,686.12 on defensive buying. Oil moved higher. The week looked like it would be a replay of May 11-16 — bond vigilantes versus AI believers — with the vigilantes winning early. KOSPI opened up 0.31% to 7,516, recovering off its prior-week crash. The Nikkei fell 0.97% to 60,816.

Tuesday May 19 brought the week's intraweek low: the S&P 500 dropped 0.67% to 7,353.61. Two earnings reports defined the session. Home Depot (BMO) reported Q1 FY2026: revenue of $41.8 billion versus the $41.5 billion estimate, EPS of $3.43 versus $3.41 — both technically beats, but gross margin compressed 75 basis points to 33% from the SRS/GMS acquisition integration, and comparable sales rose only 0.6% US (+0.4% comparable). Operating margin fell to 12.3% from 13.2% a year ago. The stock dipped on the margin compression story. CAVA Group (BMO) was the session's positive outlier: Q1 2026 EPS of $0.20 versus $0.17 estimate; revenue of $438.3 million versus $418.5 million estimate; same-restaurant sales guidance raised to 4.5-6.5%; full-year EBITDA guide raised. Shares surged 6% after hours and held a 3% gain through Wednesday. The Mediterranean fast-casual chain is benefiting from the trade-down effect as consumers substitute from full-service restaurants to premium fast casual.

Wednesday May 20 was the week's most structurally important day. At 2:00 p.m., the Federal Reserve released the minutes from the April 28-29 FOMC meeting — the last meeting chaired by Jerome Powell. The minutes were strikingly hawkish: "almost all participants" noted the risk that higher energy and input costs could persist; the "vast majority" judged that inflation could take longer than expected to return to 2%; many participants "indicated they would have preferred removing the language from the post-meeting statement that suggested an easing bias"; and a majority stated that "some policy firming would likely become appropriate if inflation were to continue to run persistently above 2 percent." The minutes explicitly shifted the policy discussion from "when to cut" to "whether to hike." Then, after the market close, two events reshaped the week's trajectory: SpaceX filed its S-1 with the SEC, targeting a Nasdaq listing under ticker SPCX with a valuation of $1.75–2 trillion, a planned fundraise of $40–80 billion, and an expected IPO date of approximately June 12. The filing disclosed 2025 revenue of $18.7 billion and adjusted EBITDA of $6.6 billion; Q1 2026 revenue of $4.7 billion. And NVIDIA reported Q1 FY2027: $81.6 billion in revenue (up 85% year-over-year, above the $79.2 billion estimate), data center revenue of $75.2 billion (up 92% YoY, above the ~$73 billion estimate), non-GAAP EPS of $1.87 versus $1.77 estimate, GAAP gross margin of 74.9%. The company guided Q2 FY2027 revenue of $91 billion — above the FactSet consensus of $87.2 billion. It raised its quarterly dividend 25 times (from $0.01 to $0.25 per share) and authorized $80 billion in additional buybacks.

Thursday May 21 was the paradox made visible. NVIDIA — despite its most extraordinary quarter and a guidance raise that beat consensus by $3.8 billion — fell 1.77% to $219.51. The company had reported the largest single semiconductor corporate quarter ever recorded and the market sold it. Meanwhile, three other major earnings stories dominated: Walmart (BMO) reported Q1 FY2027 revenue of $177.8 billion (+7.3% YoY), above the $174.8 billion estimate, US comp sales +4.1%, global eCommerce +26%. EPS came in at $0.66, in-line with estimates. But CEO John Furner confirmed that tariff-driven price increases on a wide range of imported goods are imminent — beginning in late May and accelerating through June. Walmart stock fell 6.2% to $122.78. The consumer spending message was simultaneously reassuring (Walmart's highest transaction growth in 18 months) and alarming (that growth is consolidation into the cheapest possible destination, a classic distress indicator). Deere & Company (BMO) reported Q2 FY2026: EPS of $6.55 versus the $5.74 estimate (14% beat); revenue of $13.37 billion versus $11.5 billion estimate. But EPS fell 1.4% year-over-year from $6.64; the Production & Precision Agriculture segment — Deere's most-watched unit — generated $4.5 billion but fell 14% year-over-year. The agricultural equipment cycle remains in contraction even as financial results beat depressed estimates. Stock dipped. Iran noise defined the afternoon: a report circulated claiming the US and Iran were "near" a draft resolution — Russell 2000 and Dow rose sharply on the headline. The report was subsequently characterized by TheStreet as "bogus." Then Iran's Supreme Leader issued a directive that near-weapons-grade uranium could not be sent abroad — hardening Tehran's position on the single issue that blocks any nuclear deal. The IRGC issued a statement threatening that war would "spread far beyond the region" if US attacks resumed. WTI crude surged 2.31% to $100.50; Brent climbed 1.8% to $106.90 as the uranium directive dashed deal hopes.

Friday May 22 was the resolution — and the record. Oil retreated: WTI fell back to approximately $98.15 (down 7.5% for the week), Brent to approximately $105 (down 3.7%), as markets net-priced the Iran situation as "slow progress" rather than "deal collapse." The Dow rose 294 points to 50,579.70 — its first-ever close above 50,500, a new all-time record. S&P 500 gained 0.37% to 7,473.47, logging its eighth consecutive weekly gain. The Russell 2000 added 0.91%, bringing its weekly gain to 2.7% — small caps benefiting disproportionately from oil's weekly decline (lower input costs) and the fading rate-hike premium. NVIDIA fell another 1.69%, closing around $215.81 — down approximately 3.4% cumulatively since reporting the largest semiconductor corporate quarter in history Wednesday night. The 10-year Treasury yield ended the week at 4.56% (from 4.59%); the 30-year at 5.06% (from 5.13%) — both slightly lower as oil's decline eased inflation fears marginally but left the structural picture unchanged. Markets closed early for the Memorial Day long weekend.

The week's narrative: NVIDIA proved AI infrastructure demand is real and enormous. The FOMC proved rate hikes are now on the table. Walmart proved tariff costs are reaching consumers. And SpaceX proved that the era of mega-private-company IPOs is not over — it is just beginning. The market ended the week higher anyway, carried by the Dow's defensive quality and the Russell's small-cap oil relief. The AI trade is alive; its direct expression is just, for now, priced at a level where "historic" isn't enough.

Biggest Movers

Winners

Losers

Market Scoreboard

Weekly Index Performance

KOSPI note: South Korea's benchmark continued its volatile recovery from last Friday's historic 8,000-intraday-then-crash session, gaining approximately 4.4% on the week to 7,848. Thursday saw an extraordinary +8.42% single-day surge — the largest one-day gain since early 2023 — on a combination of Iran deal optimism, oil decline, and continued AI supply chain earnings tailwinds for Samsung, SK Hynix, and TSMC (whose equipment suppliers' record quarters feed directly into Korean fab capex).

Nikkei note: Japan's benchmark gained ~3.1% for the week, recovering to 63,339. The recovery reflects the mirror image of last week's pain: modestly lower global bond yields and oil easing slightly, combined with a weak yen supporting export earnings. The BOJ's path remains unchanged; Nikkei is trading on global sentiment as much as domestic fundamentals.

Commodities & Rates

The defining chart of this week is WTI crude's 7.5% weekly decline — the largest weekly oil drop since the Pakistan-mediated US–Iran ceasefire of April 8. But the decline was not fundamental; it was Iran-deal-optimism driven, and that optimism rested partly on a "bogus report" of a near-resolution on Thursday. The structural issue — Iran's near-weapons-grade uranium enrichment position — hardened on Thursday before oil eased on Friday. The net -7.5% weekly move is not a policy-confirmed signal; it is a market pricing a lower probability of deal collapse while the actual probability of a deal remains uncertain. Gold's relative stability (down only 0.6% after last week's 3.8% decline) reflects the gold market finding a floor: real yields are no longer rising — the 30Y fell ~11–12 bps to 5.06% — giving gold room to stabilize. Bitcoin's flat weekly move continues a pattern of crypto underperformance in elevated-rate, elevated-inflation environments where risk-free assets yield 4.56% with near-zero default risk.

Earnings Recap

The Week's Defining Print: NVIDIA

NVIDIA's Q1 FY2027 is almost certainly the largest single corporate quarter ever reported in terms of revenue for a semiconductor company. The data center result of $75.2 billion (+92% year-over-year) is larger than the total quarterly revenue of all but a handful of S&P 500 companies combined. The annualized run rate above $360 billion — implied by the Q2 guide of $91 billion — exceeds the entire annual revenue of most Fortune 50 companies. And yet the stock fell for two consecutive days post-earnings.

The AI Infrastructure Thesis:

NVIDIA's $75.2B data center quarter is the hyperscaler capex commitment made visible. Last month, when Alphabet, Meta, Microsoft, and Amazon announced a combined ~$700–725 billion in calendar 2026 AI capex, the market asked: where does that money go? Wednesday night provided the answer — $75.2 billion of it, in one quarter, flows through NVIDIA's data center division. The ai_infrastructure_layer thesis is not speculation anymore; it is an earnings fact. And the $91 billion Q2 guide — issued without assuming any China data center revenue — means the hyperscaler spend continues into June regardless of US-China relations.

Retailers: The Tariff Pass-Through Arrives

Walmart's 6.2% decline is the tariff pass-through trade going live in the market. The revenue beat (and the 7.3% revenue growth) was irrelevant against one sentence from CEO John Furner: price increases on a wide range of imported goods begin in late May 2026 and accelerate through June. Walmart is the world's largest goods importer. When Walmart says prices are going up, they mean the tariff cost has exceeded what can be absorbed in margin — and consumers will pay. The global_consumer_staples thesis faces a new test: Walmart remains the trade-down destination (highest transaction growth in 18 months confirms flight to value), but the margin and stock impact of tariff-driven cost increases is real. Home Depot's margin compression story is also tariff-adjacent: building materials, power tools, and imported seasonal merchandise are all subject to the 30% US tariff. The retail_deep_value thesis faces a challenging second quarter.

Industrial Bellwether: Deere's Split Screen

Deere beat a depressed bar, not a healthy one. The EPS beat of 14% and revenue beat of 16% look impressive until you notice that EPS declined year-over-year from $6.64 to $6.55 and that the Production & Precision Agriculture segment — Deere's core — fell 14% YoY. The agricultural equipment demand cycle remains in contraction; farmers are not buying new equipment as commodity prices and farm income face pressure. The agriculture_food thesis requires patience: Deere is performing adequately at a trough, not accelerating from strength.

The Consumer's Bright Spot

CAVA is the consumer trade-down thesis in real-time. As full-service restaurants see traffic pressure from $109-per-fill tanks and tariff-driven grocery price increases, Mediterranean fast casual at $9–15 per entrée offers a premium experience at a discount-to-full-service price. The k_shape_economy strategy captures this dynamic: high-income consumers are trading down from fine dining to fast casual; moderate-income consumers are trading down from fast casual to QSR. CAVA sits at exactly the sweet spot where both segments are currently landing.

The Week's Event: SpaceX IPO Filing

SpaceX filed its S-1 with the SEC on May 20, 2026. Planned Nasdaq ticker: SPCX. Expected IPO date: approximately June 12, 2026. Valuation range: $1.75–2 trillion. Fundraise target: $40–80 billion. The filing discloses 2025 revenue of $18.7 billion and adjusted EBITDA of $6.6 billion; Q1 2026 revenue of $4.7 billion and adjusted EBITDA of $1.1 billion. SpaceX's stated total addressable market: $28.5 trillion — encompassing $370 billion in space, $1.6 trillion in connectivity, and $26.5 trillion in AI. The $2 trillion IPO valuation, if achieved, would make SpaceX one of the five most valuable companies in the world at listing. The pre_ipo_innovation_funds and ai_infra_picks_shovels theses both have direct exposure to the SpaceX listing thesis.

Geopolitical Update

Iran/Hormuz: The Draft Resolution That Wasn't

• Mon May 18: Oil rose on Iran war uncertainty. Markets absorbed the prior week's deal-collapse narrative. Both parties continued exchanging proposals through Pakistan back-channel.
• Wed May 20: FOMC minutes flagged "almost all participants" cited persistent energy/input cost risk from Middle East conflict. The Fed has formally integrated the Hormuz blockade into its inflation modeling.
• Thu May 21 (morning): A report circulated — later characterized as "bogus" by TheStreet — that the US and Iran were "near" a draft resolution. Russell 2000 and the Dow rose on the headline. Oil held initially.
• Thu May 21 (afternoon): Iran's Supreme Leader issued a directive that Iran's near-weapons-grade uranium enrichment cannot be sent abroad — the single issue that most directly blocks any US nuclear deal framework. The IRGC simultaneously issued a statement threatening that war would "spread far beyond the region" if US attacks resumed. WTI surged 2.31% to $100.50; Brent climbed to $106.90.
• Fri May 22: Oil retreated sharply; WTI fell to approximately $98.15 (weekly low). Markets interpreted the week's Iran signals as "no imminent deal but no imminent escalation either." The Polymarket prediction market for a US-Iran permanent peace deal continues to trade below 30%.

Net assessment: The week's Iran noise produced a 7.5% weekly decline in WTI — but driven by oscillating hope rather than structural resolution. The uranium enrichment issue is not a negotiating tactic; it is a red line the Supreme Leader has drawn publicly. Any deal requires Iran to agree to halt all enrichment (US demand) or the US to agree to allow some enrichment (Iran demand). These positions are irreconcilable under current leadership rhetoric. Pakistan remains the mediating back-channel; no formal session has been announced. Oil at $98 WTI / $105 Brent still implies persistent Hormuz premium versus pre-crisis levels of ~$68–72 WTI.

FOMC Minutes: Rate Hikes Formally On the Table

The April 28-29 FOMC minutes — released May 20, 2026, and representing the last meeting Powell chaired — represent a formal regime change in Fed communication. Key passages:

• Almost all participants noted risk of persistent energy and input costs from Middle East conflict
• Vast majority judged inflation could take "longer than previously expected" to reach 2%
• Many participants wanted to drop the post-meeting easing bias language entirely
• Majority stated "some policy firming would likely become appropriate if inflation were to continue to run persistently above 2 percent"

The current Fed funds rate sits at 3.50–3.75%. With CPI at 3.8%, PPI running hot, and the 30-year Treasury above 5%, the minutes effectively confirmed that the April 2026 meeting was the last before the Committee pivoted to explicitly discussing hikes. Warsh's June 16-17 FOMC meeting — his first as Chair — now arrives with the prior Committee already on record as considering tightening.

The Dow's 50,000 Milestone and What It Means

The Dow Jones Industrial Average's close at 50,579.70 on Friday — the first close above 50,500 in history — is both symbolic and structurally meaningful. The milestone is driven by the Dow's defensive-and-industrial composition: United Health, Goldman Sachs, Johnson & Johnson, Caterpillar, and Boeing carry more weight than pure-play AI names. In a week when NVIDIA fell and Walmart fell but the Dow rose to a record, the market is rotating toward quality earnings with domestic revenue exposure and away from globally-integrated supply chains facing tariff and geopolitical uncertainty. The boring_compounder and quality_factor theses are quietly outperforming.

Strategy Scorecard

Winners

Mixed

Losers

MVP of the Week

ai_infrastructure_layer — for the fourth consecutive week. In a week when NVIDIA's stock fell despite a record quarter, the underlying thesis — that hyperscaler AI infrastructure capex is macro-inelastic and compound-accelerating — has never been more empirically supported. Data center revenue of $75.2 billion in a single quarter, growing 92% year-over-year, issued without any China revenue assumption, and guiding $91 billion for the following quarter — this is not a cycle; it is a sustained capital allocation commitment from the five hyperscalers that no tariff, no rate hike, and no geopolitical event has slowed. The stock price is a separate question from the thesis. The thesis is confirmed. Honorable mention: k_shape_economy — CAVA's 9% gain on trade-down consumer thesis, alongside Walmart's tariff warning (boosting CAVA relative to full-service dining) and Home Depot's margin pressure story, is the clearest single-week expression of the bifurcated US consumer in 2026.

Next Week Preview: May 26 – May 30, 2026

Note: Monday May 25 is Memorial Day — US markets are closed. Four-day trading week.

Economic Calendar

Earnings Reports (Key)

Political / Central Bank

Geopolitical Watchlist

• Iran/Hormuz: The uranium enrichment red line is now explicit and public. A deal requires either side to move off their stated position. Oil at $98-105 currently embeds a "no deal, no escalation" scenario. Any new military incident — drone attack, ship seizure — would push Brent above $115 immediately. A genuine back-channel signal (Pakistan, Oman) could drop oil $8-10 to the high $80s.
• PCE + FOMC Minutes synergy: If April PCE confirms 3.8% CPI, Warsh faces a June FOMC with the prior Committee explicitly on record discussing hikes and fresh data supporting that discussion. He cannot cut. He might hike. The market has not fully priced a 2026 hike — doing so would reprice every rate-sensitive sector.
• SpaceX IPO mechanics: A $40-80 billion fundraise in June is the largest since Saudi Aramco. Index inclusion of a $2 trillion company in Nasdaq-100 would require every ETF holding QQQ to buy SPCX at inclusion weight. The forced-buy dynamic could dominate early June equity market flows.

Monday Setup

Scenario A: PCE in-line, SpaceX IPO process catalyzes AI re-rating (~35% probability)

April PCE comes in at 3.6-3.8%, consistent with CPI — no shock but no relief. The SpaceX IPO process enters its roadshow phase, generating a wave of AI-ecosystem and space-economy analysis that re-rates NVIDIA's supply chain names. Marvell and Dell report strong data center guidance. Oil holds below $100. Markets extend the 8-week winning streak; S&P tests 7,500-7,550. Action: hold ai_infrastructure_layer and picks_and_shovels_ai at full weight; add pre_ipo_innovation_funds ahead of SpaceX IPO pricing; maintain warflation_hedge at half-weight while Brent below $100.

Scenario B: PCE hot — rate-hike pricing crosses 50% (~40% probability)

April PCE surprises above 3.9%. CME FedWatch probability of a 2026 rate hike crosses 50% for the first time. 30-year Treasury retests 5.13%; 10-year approaches 4.65%. Salesforce and Snowflake guidance softens as enterprises pause AI software spending to assess cost structures under rising rates. NVIDIA's post-earnings hangover continues through Memorial Day return. The 9-week winning streak ends. Action: add quality_factor and defensive_rotation; reduce AI tactical names to half-weight; extend warflation_hedge at full weight; raise cash to 18-20%; hold bond_duration_trade at zero exposure.

Scenario C: Iran back-channel produces framework announcement (~25% probability)

Pakistan's back-channel mediation produces a preliminary framework — a joint statement that both parties will halt offensive operations and begin formal negotiations within 60 days. Brent drops $8-12 immediately. 30-year Treasury falls to 4.85% on "inflation resolved" narrative. The Dow and S&P gap higher. Every rate-sensitive sector (REITs, utilities, homebuilders) surges. Action: add crisis_rotation; significantly reduce warflation_hedge to a 1% defensive residual; extend ai_mega_ecosystem; add dual_momentum to capture the rotation; buy utility_infra_income for rate-relief positioning.

Position Sizing

• Core (SPY, QQQ): hold through PCE print — this is macro data, not binary earnings
• AI infrastructure: maintain at full weight; NVIDIA stock decline is a valuation story, not a thesis failure
• Energy: hold warflation_hedge at full weight while Brent >$95; partial reduce if Brent falls below $95 on Iran framework
• Retail: no new adds to retail names; tariff pass-through makes margins unpredictable through June
• Cash: 15-18% heading into PCE — the rate-hike scenario is now the 40% probability base case
• SpaceX: begin building pre_ipo_innovation_funds exposure this week ahead of June 12 IPO

---

The Sell-the-News Paradox: Why NVIDIA's Greatest Quarter Ever Moved the Stock Nowhere

A meditation on what happens when a company becomes too good to beat, the mathematics of expectation premium, and why the world's most valuable company must now grow faster than the economy of France to surprise its own shareholders.

On Wednesday night, May 20, 2026, NVIDIA Corporation reported results for its fiscal first quarter of 2027. The numbers were extraordinary by any standard in the history of publicly traded companies:

• Revenue: $81.6 billion. Up 85% year-over-year. The largest single corporate quarter in semiconductor history.
• Data center revenue: $75.2 billion. Up 92% year-over-year.
• Non-GAAP EPS: $1.87, versus the $1.77 estimate.
• Q2 guidance: $91.0 billion — above the FactSet consensus of $87.2 billion by $3.8 billion.
• Dividend: raised 25 times, from $0.01 to $0.25 per share per quarter.
• Buyback: $80 billion authorization added.

The stock fell 1.77% on Thursday. Then another 1.69% on Friday. The company that reported the largest quarter in semiconductor corporate history was worth less at the end of the week than before it reported.

This is not the first time. NVIDIA's stock has now declined following all four of its last four earnings reports — four consecutive quarters in which it delivered some of the most extraordinary revenue growth ever recorded by a technology company. The pattern is so consistent that it deserves a name: the Expectation Premium Trap.

What the Expectation Premium Trap is.

A stock price is not a record of what a company did. It is a claim on what a company will do next, discounted to today. When NVIDIA's valuation implies an annualized run rate above $360 billion by Q2 FY2027, the stock has already priced in approximately this level of performance. "Beating" estimates does not trigger a price increase unless the beat itself was unexpected — and when the consensus has been repeatedly revised upward in the weeks before the report, beating the number means beating a bar that was already set very high.

In NVIDIA's case, the pattern in Q1 FY2027 was textbook:
1. NVIDIA guides $78 billion.
2. Analysts raise estimates to $79.2 billion based on hyperscaler capex signals.
3. Institutional investors build private "whisper" models at $83-85 billion.
4. NVIDIA reports $81.6 billion — between consensus and whisper.
5. Q2 guidance of $91 billion beats consensus of $87.2 billion — but institutional whispers were at $89–90 billion.
6. The result: every number beat the official consensus. No number beat the real expectation.
7. Stock falls.

The trap is structural, not idiosyncratic. It is what happens when a company becomes so important to so many portfolios that every investor who wants exposure already has it, and every analyst who follows it has already raised their target to fully reflect the upside scenario.

The mathematics of needing to surprise at scale.

NVIDIA guided $91 billion for Q2. To "surprise" the market at Q2 — to produce the upside reaction investors keep expecting — the company would need to report somewhere above $96-98 billion in actual revenue and guide $103-105 billion for Q3. At $100 billion per quarter, NVIDIA would be generating over $400 billion annually. For context: Apple's entire fiscal year 2025 revenue was approximately $416 billion. Amazon's 2024 revenue was approximately $638 billion, but Amazon includes AWS, Prime, retail, and advertising. Microsoft's FY2025 was approximately $281.7 billion.

NVIDIA's data center division alone — at $75.2 billion quarterly — is now generating more revenue per quarter than the entire annual revenue of companies like Goldman Sachs, American Express, or Adobe. To grow fast enough to "surprise" the market from here requires growing at a rate that would make NVIDIA, within five years, larger by revenue than any company that has ever existed.

This is not a criticism of NVIDIA. The company is legitimately extraordinary. It is an observation about what "priced to perfection" means when perfection itself is a moving target.

The historical parallel: Apple in 2012.

The last time this dynamic played out at comparable scale was Apple from late 2012 through 2013. Apple was then reporting the best quarters in consumer electronics history — selling more iPhones than analysts expected, generating record profits. The stock peaked in September 2012 at approximately $25 (fully split-adjusted, accounting for the 2014 7-for-1 and 2020 4-for-1 splits); equivalently ~$100 after only the 2014 split. It fell approximately 37% over the following six months (to March 2013); the full trough decline of ~43–44% was reached in April 2013, seven months after the peak, despite three consecutive quarters of beating estimates. The reason: the iPhone market was approaching saturation in early adopter demographics, and the rate of growth — while still enormous — was decelerating. The stock had priced in acceleration; deceleration (even while still growing fast) caused a repricing.

NVIDIA's situation differs in one critical respect: the hyperscaler AI capex cycle is not saturating; the $91 billion Q2 guide, issued without any China revenue, confirms continued acceleration. But the stock-price dynamic is identical: investors who bought NVIDIA expecting to profit from earnings beats have already received the earnings beats. The mechanism for further upside requires a new catalyst beyond sequential growth — either a market that hasn't fully priced the China revenue optionality, a new Blackwell-successor product cycle that expands the addressable market, or a SpaceX IPO-style event that forces re-evaluation of AI semiconductor TAM.

What it means for the thesis.

The ai_infrastructure_layer and nvidia_supply_chain theses remain unambiguously valid. The hyperscaler AI capex commitment is real, growing, and denominated in NVIDIA GPUs. But the direct-NVDA trade — buying NVIDIA stock expecting to profit from earnings beats — has graduated from "fundamental catalyst" territory to "priced to perfection plus more" territory. The highest-return expression of the AI infrastructure thesis may now be one layer removed: Marvell's custom silicon, Cisco's Ethernet fabric, Applied Materials' deposition equipment, and TSMC's advanced packaging — all companies whose own prices have not yet been bid to the level where perfection is baseline.

The $81.6 billion quarter is not disappointing. It is just no longer surprising. And in markets, what surprises is what moves.

---

Sources:
- Stock Market Today (May 18, 2026): Nasdaq, S&P 500 sink — TheStreet
- Stock Market Today (May 21, 2026): Russell 2000, Dow rise on bogus Iran report — TheStreet
- Stock Market Today (May 22, 2026): Dow rises 294 points to record — TheStreet
- Stock Market Today, May 22: S&P 500 Posts Eighth Straight Week of Gains — Yahoo Finance
- Nvidia earnings takeaways: Data center revenue nearly doubles, stock slides — CNBC
- Nvidia Q1 earnings: Chipmaker beats earnings and boosts dividend, but forecasts disappoint — Fortune
- NVIDIA Q1 revenue hits $81.6B, ups dividend, buyback — StockTitan
- NVIDIA CORP Form 8-K FY2026 Q1 FY27 Press Release — SEC
- Nvidia Stock After Earnings: Nvidia Reports $81.6B Revenue, Raises Dividend 25x — IndMoney
- Walmart issues worse-than-expected outlook as high gas prices hit shoppers, shares drop 7% — CNBC
- Walmart Stock Plunges 6.20% to $122.78 After Q1 Earnings and Cautious Outlook — IBT
- Walmart Q1 FY27 Earnings: Reading the US Consumer Pulse — HeyGoTrade
- Home Depot Q1 2026: Revenue Hits $41.8B as Margin Compression Continues — TIKR
- Home Depot Q1 2026 Earnings: Mixed Results Amid Slowdown — FMP
- Deere Releases Q2 2026 Financial Results — Alphastreet
- Deere & Company Q2 2026 beats expectations, stock dips — Investing.com
- CAVA stock rises on earnings beat and raised guidance — Investing.com
- CAVA GROUP Releases Q1 2026 Earnings, Stock Rises — Quiver Quantitative
- SpaceX's historic IPO plans: Billions in losses and Musk's massive ownership — CNBC
- SpaceX IPO targets $28.5 trillion total addressable market — Fortune
- SpaceX files for IPO — SpaceNews
- Federal Reserve Board — Minutes of the FOMC, April 28-29, 2026
- FOMC Minutes May 2026: Three Fed Signals — The Alpha Pulse
- Treasury Yields Snapshot: May 22, 2026 — ETF Trends
- Current price of gold: May 22, 2026 — Fortune
- Bitcoin and ethereum prices today, Friday, May 22, 2026 — Yahoo Finance
- Stocks Halt Slide as Oil Declines on US-Iran Hopes — Bloomberg
- Asia-Pacific markets: Brent, Nikkei 225, Kospi, Hang Seng — CNBC
- 2025–2026 Iran–United States negotiations — Wikipedia
- The Week Ahead: Retail Earnings, PCE Inflation Data — Schaeffer's Research

---

Disclaimer

The information, strategies, research reports, pre-market analyses, trade recommendations, and all other content contained in this repository are provided for educational and research purposes only and do not constitute financial advice, investment advice, trading advice, or any other form of professional advice. Nothing herein should be construed as a recommendation, solicitation, or offer to buy or sell any securities or financial instruments.

All content in this repository — including but not limited to pre-market research reports, strategy analyses, trade recommendations, leaderboard rankings, and research findings — has been generated in whole or in part by large language models (LLMs) and automated systems. AI-generated content may contain errors, inaccuracies, hallucinated facts, incorrect dates, fabricated statistics, or misleading information. Users must independently verify all information, data points, dates, financial figures, earnings results, economic indicators, and factual claims before relying on them for any purpose.

Past performance, whether actual or backtested, is not indicative of future results. Trading and investing involve substantial risk of loss. Consult a qualified, licensed financial advisor before making any investment decisions.