Today on The Tape Reader: NVDA delivered the beat, the guide, the buyback, and the dividend hike — and the stock fell. The week of coverage that flagged this outcome is now resolved; what matters next is what comes after the bellwether can't rally on a flawless print. Underneath: a hawkish Fed confirmed by its own minutes, software setting up its biggest mean-reversion trade in years against semis, and the AI-power thesis widening from utilities into miners and microreactors.
The beat-and-fade base case confirmed. Q1 FY27: revenue $81.61B vs $79.19B consensus, EPS $1.87 vs $1.77, data-center $75.25B (+92% YoY), networking +199% YoY, gross margin 75%. Q2 guide $91B ±2% crushed the $86.84B–$87.36B Street range. Board approved an incremental $80B buyback and raised the quarterly dividend from $0.01 to $0.25 — a 25x hike. Initial AH reaction: -0.7% to -1.6%, stock back to roughly flat by Thursday open. Asian AI proxies caught the move instead — SK Hynix +11%, KOSPI +8.3%. Jensen disclosed Vera CPU targets a $200B TAM and is expected to generate $20B in revenue by fiscal year-end, becoming NVDA's second-largest line. China remains zero ($4.6B YoY hole); supply commitments jumped to $119B vs $95.2B prior quarter. The $355B options-implied swing produced essentially nothing — May 235-strike calls expired worthless, with $114M in premium gone. Mott Capital pegs drift target at $195–200 if call premium unwinds; gamma flips dealer-short above $230.
Why it matters
Prior briefings flagged beat-and-fade as the base case given the options setup, exhaustion signals, and hedge-fund de-grossing. That thesis is now resolved — what matters from here is the sequencing of three structural drags that converge post-print: (1) gamma flip above $230 with dealer books net short, meaning any drift lower is self-reinforcing; (2) the Vera CPU disclosure is Jensen explicitly defending against Google/Amazon/Microsoft custom silicon in inference — competitive positioning, not pure demand expansion — which changes the bull leg's quality; (3) China isn't recovering: the RTX 5090D V2 loophole just closed (covered in story #12 today), making the $4.6B YoY hole structurally wider, not cyclically recoverable. Watch $232 as the intraday line; loss opens $212 quickly.
Intuit beat fiscal Q3 estimates and raised full-year guidance, then announced a 17% workforce reduction (~3,000 jobs) tied to an AI-driven operating model pivot. Stock dropped 11% after-hours. The pattern echoes ZoomInfo's -36% AH on a similar beat + 20% RIF + consumption-pricing pivot (5/18 briefing). HubSpot reported the opposite — gapped +4.3% on a Q1 beat with insider buying — but BofA cut its target from $300 to $180 citing AI-structural disruption.
Why it matters
There's a coherent setup forming in enterprise SaaS: the market is rewarding companies that show AI agent monetization and punishing those that pre-announce restructuring without showing the revenue offset. INTU beat-and-raised but the layoff disclosure signals management doesn't yet trust the AI cost-base to flex on its own — a tell. For positioning, the short candidates are SaaS names announcing RIFs alongside guidance changes (read: GTM, INTU); the long candidates are platforms where insider buys + agentic product traction is paired with the beat (HUBS being the live test). Watch CRM next week as the highest-stake read in the series.
Target Q1 EPS $1.71 vs $1.46, revenue $25.44B vs $24.64B, comps +5.6% (first positive in five quarters, vs +2.4% consensus), full-year sales guide doubled to ~4% — the guide-raise that was the headline in yesterday's briefing. New development: stock fell ~4% as management's cautious tone on consumer macro and tariff hedging dominated the call despite the clean beat. BofA simultaneously raised its target to $110 from $106 but kept Underperform, warning of demand rollover. ELF Beauty Q4 beat ($0.32 vs $0.29; revenue $449M vs $423M) but issued FY27 guidance below consensus and disclosed plans to test price cuts after a $4 reduction on Halo Glow drove a 40% unit lift — direct consumer-elasticity confirmation.
Why it matters
Yesterday's briefing highlighted TGT's doubled growth forecast as the lead signal; today's update is the price action: +beat +raise = -4% stock. The beat-and-fade pattern that confirmed in NVDA overnight is now visible in consumer as well. The ELF elasticity data (40% unit lift from a $4 price cut) is the most actionable new fact — it quantifies the pricing-power rollover and is directly relevant to any name in the consumer basket reliant on tariff pass-throughs. WMT this morning is the live read: if Walmart shows the same comp strength with the same cautious tone, it confirms the pattern and creates a tactical short window across consumer-discretionary names.
Goldman Sachs prime brokerage: hedge funds liquidated a record $4.6B in single stocks the week of May 12–16, with $3.01B concentrated in technology and semiconductors — the same week PHLX Semiconductor Index dropped 4% on May 16 (worst day since March 30) and another 2.5% on May 19. The selling preceded NVDA's beat-and-raise, confirming institutional positioning was already in distribution before the print that the consensus expected to reignite the rally.
Why it matters
This gives the 'smart money de-risked INTO the print, not out of it' thesis a hard number. It layers on top of the BofA FMS data from yesterday (cash at 3.9% below the 4% sell trigger, semis crowding at 73% from 24% in a single month) and the Morgan Stanley 2.36M SOXX liquidation — three independent data sources all pointing the same direction. The actionable read is unchanged but sharpened: any tactical bounce in semis on NVDA sympathy should be treated as a fade candidate. SMH 50-DMA is the line; loss of it accelerates toward the 200-DMA gap the SOX hasn't closed in 18 months.
Deutsche Bank upgraded HUM from Hold to Buy and lifted PT to $441 from $235 — 87% upside and one of the most aggressive single-name re-rates of 2026 — valuing the name at 12x 2028 EPS of $36.74 inclusive of Star ratings recovery. Mizuho concurrently raised to $335 from $290 on reduced MLR downside. Stock has already rallied 51% in one month; October Stars results are the binary catalyst.
Why it matters
An EP setup with a defined event calendar: 2026 as the trough year, Star ratings October, and a 12x multiple basis that's defensible if the recovery thesis holds. The 51% one-month run is the warning — entry risk is real and this is a 'right thesis, wrong timing' candidate if there's any tactical drawdown into Q3. For positioning, the cleaner expression is patience — wait for a pullback to the $290 zone before adding, with the October Stars print as the conviction trigger. Note the read-across to other MA names (UNH, CVS, ELV) where the Stars normalization thesis is universal but the trough timing differs.
FDA approved LLY's Foundayo (orforglipron), the first oral GLP-1 pill cleared without food, water, or timing restrictions for obesity and overweight with weight-related conditions. ATTAIN-1: 11.2% weight loss at 36mg vs 2.1% placebo. Type-2 diabetes filing planned for later in 2026. Removes needle-aversion barrier, opens primary care prescribing, and leverages small-molecule manufacturing scale — a direct competitive asymmetry vs. Novo Nordisk's injectable peptide franchise (Rybelsus is the closest oral comp but with eating-window constraints).
Why it matters
This is a clean pair-trade setup: long LLY / short NVO with a defined catalyst calendar. Foundayo's efficacy is slightly below tirzepatide but the convenience advantage in primary care is real, and the manufacturing-scale economics on small molecules vs. peptides give LLY a structural margin edge. The T2D label filing later in 2026 is the next binary catalyst. For positioning, the entry is mechanical — wait for any sympathy spike in NVO to fade, scale into LLY on the pullback after the initial approval gap. Sympathy reads: oral GLP-1 development-stage names (TERN, ALT) and the 'GLP-1 is overhang' basket (HSY, DENN, PEP) get incremental pressure.
Medtronic announced intent to acquire SPR Therapeutics for ~$650M cash, adding the FDA-cleared SPRINT PNS System (temporary peripheral nerve stimulation) to MDT's neuromodulation portfolio. Real-world data on 6,100 patients showed 71% pain relief. Close expected in 1H of MDT's FY27 (starting April 25, 2026); minimally dilutive to FY27 EPS, neutral-to-accretive thereafter.
Why it matters
Smaller than the headline M&A but tactically interesting: it validates the temporary PNS segment, which had been a fragmented private market, and signals MDT is buying growth into neuromodulation against an aging core. For pair candidates, NUVL and other non-opioid pain-platform names get a comp print on category valuation. The real signal is that the non-opioid alternative thesis is now strategic-buyer-validated at scale ($650M for a private company is meaningful), supporting the broader specialty-pain biotech basket on read-across.
China stopped granting import permits for NVDA's RTX 5090D V2 gaming card on May 15 — a Blackwell-architecture workaround Chinese AI buyers had been using to circumvent H200 procurement restrictions. The ban coincided with Jensen Huang's attendance at Trump's state-visit delegation and reflects Beijing's directive to source domestically from Huawei Ascend and Cambricon Siyuan. NVDA confirmed zero China revenue in Q1 vs. $4.6B YoY.
Why it matters
This is the missing piece in the NVDA fade narrative: the China hole isn't just a near-term revenue miss, it's structurally widening as Beijing closes loopholes faster than NVDA can route around them. The V2 was low-single-digit % of quarterly revenue, but the directionality matters — every NVDA Q&A from here will have a China headwind that scales. Sympathy beneficiaries: Huawei suppliers and Cambricon's listed comps (688256-CN). For NVDA positioning specifically, this layers on top of the gamma-flip and fund-flow setup — three structural drags converging into the next leg.
The IGV/SMH ratio's 14-week RSI hit a record low of 15. Software trades at 21.4x forward P/E vs. 23x for the S&P 500 — the first time on record software has been at a discount to the index — even as software earnings are growing 19%. IGV has formed a head-and-shoulders bottom; a clean break above $101 on volume would force dealer covering. Aschenbrenner's Situational Awareness Fund Q1 13F revealed a pivot from $5.5B long semis to $8.5B short semis while maintaining the long-AI infrastructure thesis.
Why it matters
This is a new data point that quantifies the rotation extreme the prior briefings have been framing directionally (software allocation collapsing from 12% to 2%, Morgan Stanley liquidating SOXX, the five-semis-drive-51.6%-of-S&P-gains distortion). The RSI 15 on IGV/SMH is the statistical anchor that turns a qualitative rotation thesis into a measurable setup. The Aschenbrenner 13F is the conviction signal: prime-broker positioning is already rotating in the direction the mean-reversion thesis requires. Risk to watch: if NVDA's fade triggers broader tech de-risking, IGV gets dragged in the first leg before the relative trade asserts — the cleaner expression is the spread (long IGV / short SMH), not the absolute long.
Bernstein research: listed Bitcoin miners (CORZ, IREN, HUT, CIFR, WULF, RIOT, MARA) have signed over $90B in AI infrastructure agreements covering 3.7GW of contracted capacity, and control roughly 27GW of planned U.S. power capacity. With grid interconnection queues at 50+ months for traditional data center developers, the 2021-2022-vintage interconnect permits these miners secured are now a more valuable asset than hashrate. Miner equity prices are visibly decoupling from BTC spot as contracted AI hosting revenue starts to dominate the P&L.
Why it matters
The AI-power trade has matured from utilities (NEE-Dominion at $66.8B) to a tradeable second-derivative basket where the assets are scarce permits, not chips. Miners now trade more like industrial REITs with contracted cash flows than like crypto beta. For sector rotation: the basket (CORZ, IREN, WULF) offers a different risk profile than CEG or VST — lower defensive characteristics but higher growth optionality and a real catalyst calendar (contract announcements, capacity online dates). Watch CORZ specifically — it's the largest pure-play conversion story. The structural read-across also strengthens nuclear (BWXT, NNE, X-Energy) and grid equipment (GEV, ETN, AES) as the same demand thesis expressed differently.
RRG sector rotation framework as of May 20: XLI (Industrials) and XLC (Communication Services) both rotated structurally into the Improving quadrant. XLE (Energy) failed its breakout attempt with a violent reversal as crude collapsed (WTI -6.82%, Brent -5.63% on Iran negotiation progress). Defensive sectors (XLV, XLP) unwound as capital rotated higher-beta. This pairs with the cyclical breakout reading from prior briefings on GE Vernova ($1,328 PT from Goldman) and Howmet ($320 Deutsche Bank).
Why it matters
The RRG rotation gives institutional confirmation to the cyclical breakout that's been visible in single names — and the failure of XLE on the same session is the cleaner short setup. The Energy fade is critical because it removes the 'inflation hedge' bid; if oil is rolling on real Iran de-escalation, the entire stagflation hedge basket (XLE, gold, TIPS) compresses while Industrials and Comm Services run. For positioning, the trade is long XLI / XLC (with GEV, ETN, AES as the AI-power expression and META/GOOGL as Comm Services expression) and short or fade XLE on any sympathy spike from inevitable Iran headline reversals. Watch the 10Y as the regime variable — if yields hold above 4.6% on growth concerns, the cyclical thesis runs further.
ES broke through overnight resistance at 7390/7415, tagged 7448 target by D-period yesterday with VIX above 17.14, formed a P-shaped profile (short covering). Key intraday pivot 7452: upside targets 7479/7509, downside 7424/7390. VIX 18.36 = strength fail; 16.54 = breadth confirmation. NQ in V-shaped recovery into 30k psychological — June futures at 29,268 with open interest 290k contracts. Pre-market Thursday: ES -0.18% to 7,419, NQ -0.48% to 29,158 as NVDA digestion drags. Brent below $105 on Iran ceasefire optimism; 10Y at 4.687%.
Why it matters
Concrete intraday levels for the open. The 7,424 ES floor is the line — hold it and the cyclical rotation runs into PMIs; lose it and the NVDA-fade narrative pulls the index lower with 7,390 as the next shelf. NQ at 30k is the cleaner asymmetric trade — proximity to psychological resistance + post-NVDA AH weakness + Aschenbrenner-style short positioning all argue for a reversal trade with tight risk above 30k. Watch claims (210k forecast) and PMIs (manufacturing 53.8, services 51.1) at 9:45/10am; manufacturing weakness paired with services strength keeps the higher-for-longer trade intact.
The April 28–29 minutes released at 2pm ET Wednesday — three hours before NVDA's print, as flagged — revealed four dissents (most since October 1992), with a majority of officials warning rate hikes 'could become appropriate' if inflation persists above 2%. Three of four dissented against retaining easing-bias language. Backdrop: April CPI 3.8% YoY, PPI +6.0% (largest since 2022), PCE 3.5%. CME FedWatch now prices >50% probability of a December 2026 hike (up from ~1% six weeks ago and the ~50–60% range flagged in yesterday's briefing, now confirmed by the minutes). Kevin Warsh was sworn in as Chair Friday; his first meeting is June 16–17.
Why it matters
The minutes confirm — rather than introduce — the inversion from cuts to hikes that prior briefings have tracked from ~1% to >50% December odds in six weeks. What the minutes add is institutional texture: it's not just market pricing, it's a fractured committee at 8–4 with explicit on-the-record hike language. Warsh inherits a divided room, not a compliant one. For the equity duration trade, this closes the last ambiguity: the 10Y at 4.687% and 30Y above 5.17% (both in prior coverage) are now policy-confirmed, not just market-implied. Watch the May 28 PCE print as the last major inflation read before Warsh's June 16–17 debut.
Beat-and-fade is now the base case for crowded AI mega-caps NVDA delivered $81.6B revenue (+85% YoY), $91B Q2 guide above the $87B Street, $80B buyback, and a 25x dividend hike — and the stock fell after-hours. The muted reaction follows record $4.6B in hedge fund single-stock liquidations the week of May 12-16 ($3.01B concentrated in tech/semis). When the bellwether can't rally on a flawless print, the marginal buyer is exhausted.
AI capex narrative widens from chips to power, water, and contracted grid capacity Bitcoin miners have signed $90B in AI hosting deals controlling 27GW of permitted power (Bernstein); NextEra-Dominion at $66.8B is being reframed as the largest validation of 'power, not chips' as the AI bottleneck; Nano Nuclear's KRONOS microreactor advances at NRC; X-Energy gets bullish Wall Street coverage. The second-derivative trade has institutional sponsorship.
Software vs. semis mean-reversion sets up as the most extreme pair trade in years IGV/SMH 14-week RSI hit record low of 15 with software trading at a forward P/E discount to the S&P 500 for the first time on record despite 19% earnings growth. Aschenbrenner's Situational Awareness Fund flipped from $5.5B long semis to $8.5B short. Head-and-shoulders bottom in IGV sets up for a squeeze through $101 if dealer books get caught.
Fed reaction function has fully inverted from cuts to hikes April FOMC minutes (released 5/20) revealed an 8-4 hawkish dissent — most since October 1992. CME FedWatch now prices >50% probability of a December 2026 hike, up from ~1% six weeks ago. Kevin Warsh inherits a divided committee with PCE at 3.5%, oil up 50% since the Iran conflict, and the 10Y at 4.687%. Duration-sensitive multiples are repricing in real time.
The exhaustion stack is forming under a record-high tape Hedge fund SOXX exposure at 19% of global portfolio (record); BofA FMS cash at 3.9% (below the 4% sell trigger); SOX 62% above its 200-DMA (only Mississippi 1720 and March 2000 match); Bull & Bear Indicator at 7.8 (contrarian sell at 8.0). Citadel's Rubner flipped to 'flow fragility,' breadth in the 1st percentile, $203B in levered ETF assets concentrated in mega-cap tech.
What to Expect
2026-05-21—Earnings docket: WMT (consumer real-time read), NIO Q1 (8.4% implied move, second consecutive profit quarter expected), ROST, ZM, NTES. US S&P Global PMIs, Philly Fed, initial claims test whether USD strength has growth support.
2026-05-28—April PCE + Q1 GDP revision — last major inflation/growth print before Warsh's June 16-17 inaugural FOMC. BRP (DOO-T) reports with $500M S232 tariff overhang in focus.
2026-06-04—LULU earnings — $14.3B cap, RSI 24, Elliott $1B stake. Washed-out leader setup into the print; last comparable oversold reading delivered a 40% rally.
2026-06-12—Reported SpaceX IPO target date at $2T valuation — potential capital reallocation event that could drain liquidity from AI infrastructure mega-caps.
2026-06-16/17—Warsh's first FOMC. Market pricing >50% December hike probability; minutes from this meeting will set the 2H tone.
How We Built This Briefing
Every story, researched.
Every story verified across multiple sources before publication.
🔍
Scanned
Across multiple search engines and news databases
1257
📖
Read in full
Every article opened, read, and evaluated
212
⭐
Published today
Ranked by importance and verified across sources
13
— The Tape Reader
🎙 Listen as a podcast
Subscribe in your favorite podcast app to get each new briefing delivered automatically as audio.
Apple Podcasts
Library tab → ••• menu → Follow a Show by URL → paste