MARKET SNAPSHOT

A semiconductor-only week. The S&P (+0.55% WoW) and Nasdaq 100 (+2.37%) closed at fresh records on Friday’s Intel-detonated semi rip, while the Dow (-0.44%), NYSE Composite (-1.13%), and Healthcare (-3.51%) all finished red — the narrowest tape of 2026. Underneath, the macro backdrop deteriorated every session: Brent +15.98% above $100 on the Hormuz standoff, the Cleveland Fed CPI nowcast climbed to 3.7%, and Friday’s UMich sentiment crashed to a record-low 49.8 with 1-year inflation expectations un-anchoring to 4.7%. Yields rose +5.8 bps WoW and Polymarket cut-odds slipped 7.5 pp — records on a stagflation tape, with the index now levered entirely to the hyperscaler capex cluster Apr 28–30.

THIS WEEK AT A GLANCE

KEY THEMES

1. Hormuz Saga: US Tanker Seizure → Indefinite Ceasefire → Shoot-and-Kill Order → Pakistan Talks — Brent Closes the Week +16% Above $100

The core facts:The Hormuz standoff dominated every session. Monday: US Navy seized an Iranian cargo ship and oil surged 5–6% as the ceasefire teetered. Wednesday: Trump extended the ceasefire indefinitely while Iran seized two ships and the EIA reported a surprise 1.9 mb crude build — yet Brent still settled at $101.82. Thursday: Trump issued a Navy “shoot and kill” order against Iranian boats, the IEA declared the closure the “largest energy security threat” in history, and Brent spiked to $105 with JPMorgan estimating 13.7 mbd in April global supply disruptions. Friday: Brent closed at $106.20 (+15.98% WoW); Witkoff and Kushner travelled to Pakistan Saturday for direct US-Iran talks.

Why it matters:Brent above $100 every session of the week is the macro datum that locked in the stagflation frame. The Cleveland Fed CPI nowcast climbed to 3.7%, Q2 PCE forecasts jumped to 3.7% (Reuters poll of 103 economists), and Friday’s UMich 1-year inflation expectations un-anchored to 4.7%. Polymarket cut-odds slipped 7.5pp WoW (66.5% → 59%) as Fed Governor Miran openly rethought his cut path mid-week (see B1 Vol & Treasuries; D2 Polymarket WoW). Brent +16% with Henry Hub -5.76% is the cleanest US-vs-global energy split of 2026 (see B1 Energy; Energy sector +3.20% see B3).

What to watch:Brent’s Sunday-evening Asia open after the Pakistan talks: a gap below $95 = breakthrough, above $115 = breakdown. April PCE Apr 30 will reveal whether the energy pass-through is showing in the Fed’s preferred gauge. ISM Manufacturing May 1 confirms or refutes the Flash PMI input-price spike.

2. Intel’s Best Day Since 1987 Detonates a Semi Mega-Rally — SOXX Extends to 18 Straight Sessions, Nvidia Reclaims $5 Trillion

The core facts:The week’s biggest single catalyst arrived Thursday after-hours: Intel reported Q1 non-GAAP EPS of $0.29 vs $0.01 consensus, revenue $13.6B vs $12.36B, and Data Center & AI of $5.05B (+22% YoY) crushing the $4.41B consensus. Q2 guidance ($13.8–14.8B) printed well above Street. Friday delivered the read-through: INTC +23.6% (best day since 1987), AMD +13.91% on Stifel’s PT raise to $320 plus D.A. Davidson’s Neutral → Buy / $375 PT, KLAC +6.59%, and NVDA +4.32% reclaiming $5T market cap. The PHLX Semiconductor index extended its winning streak to 18 consecutive sessions. The S&P 500 (+0.80%) and Nasdaq 100 (+1.95%) closed at fresh records; AMD +24.94% and INTC +20.50% led the weekly mover board (see B4).

Why it matters:Intel’s Data Center +22% YoY validated the bullish AI-capex thesis at the CPU layer — the missing link that had kept the rally Nvidia-only. The semiconductor complex is now functioning as a single trade: beat one chipmaker, all chipmakers rally. But the rally was emphatically narrow — Tech (+3.39%) topped B3 only because of AMD and INTC; strip those two and Tech is roughly flat (see B3 single-name dominance). Dow (-0.44% WoW), NYSE Composite (-1.13%), and DJ Transports all closed red. The index is now almost entirely levered to whether MSFT/GOOGL/META/AMZN validate the consensus ~$520B 2026 AI-capex print next week.

What to watch:Hyperscaler capex commentary the week of Apr 28 (GOOGL/MSFT/META/AMZN) is now the swing factor for Q2. A cumulative print below $500B risks a sharp semi unwind. SOXX breaking the 18-session streak would be the first concrete fatigue signal. AMD’s early-May earnings will validate or refute the bullish CPU thesis.

3. Warsh–Powell Saga: From “Stay in Its Lane” Hearing Mon, to Tillis Block Wed, to DOJ Drops Powell Probe Fri — Yields Round-Trip

The core facts:Monday: Kevin Warsh released his Senate hearing remarks — the Fed must “stay in its lane” on independence. Tuesday: Warsh testified before Senate Banking, vowing not to be Trump’s “sock puppet” while pledging “regime change” (trimmed-mean preferences, balance-sheet runoff, reduced forward guidance). Wednesday: Senator Tillis blocked the Senate floor vote until DOJ resolved the Powell criminal probe; earliest vote pushed to week of May 11. Friday: DOJ abandoned its Powell investigation, clearing Warsh’s confirmation path; the 10Y fell -2 bps to 4.305% and 2Y -4.2 bps to 3.78% on the relief. Net WoW: 10Y +5.8 bps, 2Y +7.9 bps — the relief partially offset, not reversed, the week’s hawkish repricing.

Why it matters:Personnel risk has been replaced by policy risk. Warsh remains a hawkish pick — his preferred balance-sheet-runoff regime would mechanically raise term premiums and tighten credit. Friday’s yield drop was a relief on institutional resolution, NOT dovish policy expectations; the bond market is now in transition. Polymarket ≥1-cut-2026 odds slipped 7.5pp WoW (see D2). Powell’s term as Chair expires May 15 — the runway from confirmation to handover is now narrow.

What to watch:Senate Banking Committee Warsh vote (expected next week). FOMC May 6–7 (likely Powell’s last as Chair) — any language shift on “patient” or “data-dependent.” If confirmed, watch the 10Y for a spike back above 4.50% within Warsh’s first 30 days as markets reprice balance-sheet runoff acceleration.

4. Records on the Narrowest Tape of 2026: S&P/Nasdaq Hit ATHs Wed, Reverse Sharply Thu, Fresh Records Fri — Dow and NYSE Composite Both Red on the Week

The core facts:The S&P 500 and Nasdaq notched simultaneous all-time records Wednesday on Trump’s indefinite ceasefire extension; reversed sharply Thursday off intraday ATHs as Trump’s shoot-and-kill order, ServiceNow -18%, and Microsoft buyout news triggered a software sell-off; then printed fresh records Friday on Intel-led semis. Net WoW: S&P +0.55%, Nasdaq 100 +2.37%, Dow -0.44%, NYSE Composite -1.13%, DJ Transports closed Friday 12.7% below its 7-session high while DJIA sat within 1% of peak (Dow Theory non-confirmation). Only 5 of 11 sectors finished green for the week.

Why it matters:The cleanest possible breadth-narrowness tell in modern records-on-records data — the index is a 4-5 mega-cap tech story while cyclicals, defense, healthcare, and small-caps refused to confirm. JPMorgan raised its S&P year-end target to 7,600 Tuesday on AI-earnings strength, but the bar for breadth confirmation is now extreme. Historically, sustained record-equity / record-low-sentiment divergences resolve within 1–2 quarters — either sentiment bottoms or equities mean-revert.

What to watch:Whether DJ Transports re-confirms the DJIA next week (Dow Theory). NYSE Composite re-pricing back above its prior week close = breadth restoration. Russell 2000 break above its prior all-time-high = small-cap participation confirmation.

5. Mega-Cap Tech Layoff Wave Funds AI-Capex Acceleration: Meta 8,000 Cuts + $135B Capex; Microsoft Voluntary Buyout for Up to 7% of US Workforce

The core facts:Wednesday/Thursday brought twin announcements that crystallized the Big Tech reset. Meta confirmed approximately 8,000 layoffs (~10% of workforce, effective May 20) plus 6,000 unfilled roles — explicitly to fund a doubling of 2026 AI spending to ~$135B (from $72B in 2025). Microsoft launched its first-ever voluntary buyout program targeting up to 7% of its US workforce. Friday’s formal Meta announcement re-confirmed the $135B figure. Both companies report earnings the week of Apr 28.

Why it matters:The cost-cut/capex-spike trade-off is now the explicit Big Tech template. Meta’s $135B is the highest in the sector and lifts the bar for ROI demonstrations. Markets initially reacted negatively (Thu software sell-off; XLK’s 16-day winning streak snapped) before the Friday Intel halo lifted everything. The bigger signal: even structurally-profitable mega-caps believe the AI-capex environment requires headcount cuts to fund.

What to watch:META and MSFT earnings calls week of Apr 28 — commentary on AI revenue contribution, Reality Labs operating loss trajectory (Meta), Azure capex (Microsoft). Any guide-up to $135B+ would be a red flag; specific monetization milestones would re-anchor the thesis.

6. Q1 Mega-Cap Earnings Tour Bifurcates: AI-Capex Names Crush (UNH, GEV, INTC, TXN, PG); “Beat-but-Not-Enough” Names Punished (IBM, RTX, GE, TSLA, NOW, AXP, TMO, BX)

The core facts:Q1 reporting accelerated dramatically. Winners: UNH +6.96% Tue (beat + raised FY guide to >$18.25), GE Vernova +13.75% Wed (Electrification booked $2.4B in Q1 data-center orders, raised 2026 outlook), Boeing +5.53% Wed (record $695B backlog), Intuitive Surgical +7% Wed, Comcast +7.73% Thu, Union Pacific +8.77% Thu, Texas Instruments Q1 beat Apr 22 (+20% next-day), Procter & Gamble +1.70% Fri (volume +2 pts — first growth in a year), Intel Q1 blowout. Losers despite earnings beats: RTX -4.40% Tue, GE Aerospace -5.56% Tue, IBM -8.25% Wed (maintained guide), Tesla AMC mixed (50K inventory, energy storage -38% QoQ), ServiceNow -18% Thu (margin guide cut triggered software sell-off), American Express -4.31% Thu, Thermo Fisher -9.20% Thu, Blackstone -5.70% Thu, Capital One -6% Wed (NIM compression).

Why it matters:The Q1 reporting universe so far is delivering 88% EPS beat / 84% revenue beat / +13.2% YoY blended growth. But four of the week’s top 5 decliners (RTX, IBM, GE, TSLA — see B4) DELIVERED earnings beats and were sold anyway. The market is rewarding only one specific pattern: AI-capex-aligned beat-and-raise. Everything else — even genuine beats — gets repriced for guidance disappointment, NIM compression, or 2027 risk.

What to watch:Hyperscaler week (Apr 28–30) brings GOOGL, MSFT, META, AMZN, AAPL, LLY, MA, CAT, MRK, COP. The reaction asymmetry — how cleanly are beat-and-raise prints rewarded vs beat-and-maintain prints punished? — is the cleanest read on whether the Q1 narrative is durable.

7. ServiceNow -18% Snaps XLK’s 16-Day Win Streak in Software Sector Sell-Off — First AI-Capex-ROI Crack of the Cycle

The core facts:Thursday Apr 23 brought the week’s sharpest single-name move: ServiceNow (NOW) plunged 18% after cutting its margin guidance, citing higher-than-expected AI infrastructure investment costs. The disclosure ricocheted across the software complex — XLK’s 16-day winning streak snapped, IBM -8.25% (maintained guidance only), and software peers sold broadly. The S&P and Nasdaq reversed sharply off intraday ATHs to close lower on the day — the only red day for indices on the week.

Why it matters:The first concrete crack in the AI-spending narrative. NOW disclosed that the AI-build-out is compressing software margins faster than top-line revenue can offset — the thesis-killer scenario for SaaS multiples. Notably the SOXX index continued higher Thursday on hardware/chip demand strength even as software cracked — the AI trade is bifurcating into “capex beneficiaries” (semis, power, data-center infra) and “capex bearers” (software). Friday’s Intel print masked but did not erase the warning.

What to watch:MSFT and GOOGL margin commentary the week of Apr 28 (do hyperscalers show the same NOW-style margin compression?); CRM, ORCL, ADBE forward guides over the next 2 weeks; XLK relative to SOXX as the bifurcation indicator.

8. Apple CEO Tim Cook to Step Down September 1 — John Ternus Named Successor at the World’s Largest US Company

The core facts:Tuesday Apr 21 brought the largest CEO transition in US-equity history: Tim Cook will step down as Apple CEO effective September 1, with John Ternus (currently SVP Hardware Engineering) named successor. The transition is being framed as planned succession rather than a sudden departure. Apple reports Q1 earnings Apr 30. BNP Paribas had upgraded AAPL to Outperform on Monday on the AI-monetization thesis — the upgrade positioning landed against very different leadership context one day later.

Why it matters:Apple is roughly 7% of S&P 500 weight; CEO transitions at this scale historically generate 6–12 months of multiple-volatility regardless of fundamentals. Ternus is a hardware specialist taking over at the moment Apple needs to credibly accelerate its AI software / Apple Intelligence narrative. The succession announcement timed three weeks before the Apr 30 earnings call ensures management commentary will be the next inflection.

What to watch:AAPL Q1 earnings Apr 30 — transition commentary, AI-product roadmap, capex commitment. Other mega-cap CEO transitions in the queue (succession activity tends to cluster). Sympathetic moves in MSFT, GOOGL, META leadership-stability narratives.

9. Amazon Commits Up to $25 Billion More to Anthropic — The Year’s Largest AI Infrastructure Deal

The core facts:Monday/Tuesday Amazon committed up to an additional $25B to Anthropic, deepening the AI-infrastructure partnership and locking in long-term AWS Bedrock workload commitments. The deal lands as the year’s largest single AI investment and signals AWS’s intent to match Microsoft-OpenAI commercial scale. Amazon reports Q1 earnings the week of Apr 28.

Why it matters:Reinforces the cumulative ~$520B 2026 hyperscaler-capex consensus that the index now requires next week to validate. The Amazon-Anthropic / Microsoft-OpenAI / Google-internal triumvirate is committing Big-Tech balance sheets to AI infrastructure at unprecedented scale — the question shifts from “will hyperscalers spend?” to “will the spending generate ROI faster than ServiceNow-style margin compression eats it?” (see story 7).

What to watch:AMZN Q1 earnings — AWS revenue growth, AI-revenue contribution, Anthropic commercial milestones. Whether MSFT/GOOGL announce comparable counter-commitments next week.

10. CUSMA / USMCA Trade Talks at Risk: USTR Greer Calls Resolution “Unlikely” by July 1; Carney Signals Tough Canadian Stance

The core facts:The trade-policy backdrop deteriorated through the week. Monday: the US weighed tougher auto-import content rules that could impose an effective 10% USMCA tariff. Tuesday: USTR signaled tariffs are “here to stay” in USMCA talks with Mexico. Wednesday: Greer testified to Congress that USMCA renegotiation is “unlikely” to resolve by July 1, saying US-Canada trade systems “don’t fit together very well.” Thursday: Canadian PM Mark Carney signaled a tough Canadian stance ahead of US trade renegotiation. Separately, Commerce finalized steel/aluminum tariff exemption procedures Thursday.

Why it matters:Three weeks ago the consensus was a clean USMCA renewal by July 1. The week reset that view: tariffs as the new baseline, with auto, steel, and aluminum sectors carrying multi-quarter tariff overhang. Combined with a CBP $166B IEEPA tariff-refund portal opening Monday for previously-collected unconstitutional duties, the trade-policy regime is increasingly fragmented — a tax on supply chains that compounds the Hormuz oil shock on Q1 GDP.

What to watch:Auto-sector earnings (GM, F, STLA, TSLA) commentary on tariff scenarios; Mexico/Canada negotiator statements; whether the July 1 deadline slips with a stand-still or escalates.

UMich Consumer Sentiment Final 49.8 — Lowest on Record Since 1952; 1-Year Inflation Expectations Spike to 4.7% (UMich, Fri Apr 24)

What they’re saying:Final April Consumer Sentiment 49.8 vs 48.0 expected, revised up from the 47.6 preliminary, but still the lowest reading in data back to 1952 — below both Great Recession and COVID lockdown troughs. 1-year inflation expectations jumped to 4.7% (from 3.8% March) — largest monthly increase since April 2025. 5-year expectations climbed to 3.5%, the highest since October 2025. Headline -6.6% MoM, -4.6% YoY; broad-based decline across all age, income, and political-affiliation groups.

The context:Inflation expectations becoming unanchored is the single development a Fed governor fears most — it forces tightening reflexively. The disconnect with markets is now extreme: S&P at all-time highs (see B1) WHILE the household sector signals distress reading. Yields rose +5.8 bps WoW and Polymarket cut-odds fell -7.5 pp (see D2 above) — the market read the print as inflation-firm, not consumer-broken. Field window captured both sides of the April 8 ceasefire and the reading still hit ATL.

What to watch:March PCE Apr 30 is now the critical data point — if YoY core PCE accelerates above 3%, the inflation-expectations spike becomes validated and the Fed loses its rate-hold optionality. Conference Board Consumer Confidence Apr 28 for cross-confirmation.

Cleveland Fed CPI Nowcast Climbs to 3.7%; Reuters Poll Lifts Q2 PCE Forecast to 3.7% (Cleveland Fed, Mon Apr 20 / Reuters, Wed Apr 22)

What they’re saying:The Cleveland Fed’s April CPI nowcast climbed from 3.28% (April 1) to 3.58% on Apr 20 and 3.7% later in the week — a 42-bp move in three weeks. On Wednesday a Reuters poll of 103 economists lifted the Q2 PCE forecast to 3.7% and pushed the first Fed cut to no sooner than September.

The context:Brent settled above $100 every session of the week (Brent +15.98% WoW — see B1 Energy), and the Cleveland nowcast and Reuters poll are mechanically pricing the energy pass-through into headline inflation. The 2Y rose +7.9 bps WoW and Polymarket cut-odds fell -7.5 pp (D2) — markets agree the Fed is being squeezed off its dovish path.

What to watch:Official BLS April CPI mid-May. April PCE Apr 30 (Fed’s preferred gauge). If Brent stays above $95/bbl through end-April, expect further nowcast upticks.

S&P Global Flash PMI: Manufacturing 47-Month High but Output Prices Spike to Sharpest Pace Since 2022 (S&P Global, Thu Apr 23)

What they’re saying:April Flash US Manufacturing PMI 54.0 (from 52.3) — a 47-month high. Manufacturing Output Index 55.7 — a 48-month high. Services PMI 51.3 (from 49.8). Composite output 52.0 (from 50.3). But output prices rose at the sharpest pace since mid-2022, explicitly linked to war-driven input costs.

The context:The headline strength is real but the composition is uncomfortable for the Fed: tariff front-running plus energy pass-through is exactly the pattern that made the Volcker squeeze necessary in 1979–81. Strong output AND accelerating prices is not a cut signal; it’s a hold-longer signal — consistent with the 2Y +7.9 bps WoW move (see B1). Initial jobless claims printed 214K Thursday (continuing 1,821K), confirming labor still firm.

What to watch:ISM Manufacturing May 1 — if both ISM and S&P PMI show the same pattern, the Fed’s “wait-and-see” extends well into H2. Dallas Fed (Mon Apr 27) and Richmond Fed (Tue Apr 28) regional reads provide the next cross-check.

March Retail Sales Surge 1.7% — Strongest in 3+ Years, but Record Gas Station Receipts (+15.5%) Mask Modest Core Spending (Census, Tue Apr 21)

What they’re saying:March retail sales +1.7% MoM — the strongest monthly gain since 2022. Headline boosted by a record 15.5% spike in gas-station receipts (oil-price pass-through). Core control group +0.7% vs +0.2% expected — modest beat but materially below the headline. ADP’s monthly employment improved for the fifth time. Pending home sales +1.5%, triple consensus.

The context:The headline is misleading — gas-station inflation does not equal real demand strength. The signal: nominal consumer-spending resilience has paradoxically become bearish for equity valuations because it gives the Fed zero cover to cut (see story 5 in C). The +0.7% control-group does feed GDPNow constructively, but Atlanta Fed’s tracking nonetheless slipped to 1.2% later in the week as investment and government revisions offset.

What to watch:BEA advance Q1 GDP Apr 30 — the gap to GDPNow 1.2% will recalibrate recession-probability models. Next retail-sales print mid-May for whether gas-station inflation persists.

Recession-Narrative Inflection: Miran Pulls Back on Cuts; Moody’s 42–48.6%, Wilmington 45%, FRED 48% — Polymarket Cuts -7.5 pp WoW (Multiple, Apr 23–24)

What they’re saying:Wednesday’s Reuters poll lifted Q2 PCE forecasts to 3.7% and pushed the first cut to September. Thursday Fed Governor Miran told a Washington forum he is revisiting his prior six-cut path: “I might have three, I might have four. I haven’t made up my mind” — explicitly citing the Iran war. Friday Moody’s Zandi reset recession risk to “uncomfortably high” ~42% (internal ML 48.6%); Wilmington Trust 45%; FRED smoothed 48% (highest since 2020); Goldman 30%; JPMorgan 35–40%. Polymarket ≥1-cut-2026 odds: 66.5% → 59% (-7.5 pp WoW).

The context:Miran’s rethink matters because he anchored the SEP dot plot’s lower end — if his dots move from six to three, the median moves with him. Institutional house spread (Goldman 30 → Moody’s 48.6) averages ~40% vs Polymarket’s 26% — the gap between bottom-up models and prediction markets is now 14 pp. Historically when bank houses and prediction markets disagree by >15 pp, the bank side has been early rather than wrong.

What to watch:FOMC May 6–7 (Powell’s last as Chair) — any language shift on “patient” or “data-dependent” will move the curve materially. NY Fed Household Debt & Credit Report (early May) for delinquency-transition rates. Next FRED smoothed update Apr 30.

1. Intel (INTC): +20.5% wk | Q1 Blowout — Data Center & AI +22% YoY, Q2 Guide Above Street, Best Day Since 1987

The Numbers:Q1 non-GAAP EPS $0.29 vs $0.01 consensus (+2,800% surprise). Revenue $13.6B vs $12.36B est (+10% beat). Data Center & AI segment $5.05B (+22% YoY) crushed $4.41B consensus. Q2 revenue guide $13.8–14.8B (consensus $13.07B); Q2 non-GAAP EPS guide $0.20 (consensus $0.09). Sixth consecutive quarter beating financial targets. Reported AMC Apr 23.

The Problem/Win:Margins and EPS came in above the upper end of guidance — not a one-line surprise but broad-based execution. The Data Center +22% YoY decisively confirms enterprise CPU refresh cycles are accelerating to host AI workloads, not being cannibalized by GPU-only architectures. Mobileye goodwill impairment ($4.1B) is non-recurring noise.

The Ripple:Friday triggered an industry-wide repricing — AMD +13.91%, KLAC +6.59%, NVDA +4.32% (back through $5T cap), PHLX SOX 18 straight sessions. Stifel raised AMD PT $280 → $320; D.A. Davidson upgraded AMD Neutral → Buy with $375 PT, citing “the magnitude of Intel’s Q1 beat indicating a structural increase in CPU demand.”

What It Means:The CPU is no longer melting ice; it is a complement to the AI accelerator stack, participating in the same capex super-cycle. Intel is investable on fundamentals for the first time in years. The print landed at the moment hyperscaler capex commentary becomes the index swing factor next week.

What to watch:Q2 Data Center trajectory — if the segment maintains 20%+ YoY in the next print, the structural thesis is confirmed. AMD’s early-May earnings for cross-confirmation.

2. UnitedHealth (UNH): +9.33% wk | Q1 Beat with FY Guide Raised >$18.25; MCR 83.9%; Multiple Sell-Side Upgrades

The Numbers:Q1 EPS $7.23 (+10% beat). Revenue $111.7B (+1.9% YoY). Earnings from operations $9.0B. Medical Care Ratio improved to 83.9% (key turnaround metric). FY26 adjusted profit guide raised to >$18.25 (from $17.75). Reported BMO Apr 21; +6.96% on day.

The Problem/Win:The MCR improvement to 83.9% is the headline — a multi-quarter trend reversal proving the cost-management initiative is working ahead of schedule. UNH separately announced a $1.5B 2026 AI investment program (Avery chatbot, Optum Real platform).

The Ripple:Argus upgraded Hold → Buy citing MCR improvement; Goldman PT $435; TD Cowen $337; Piper Sandler $420. Healthcare sector still finished the week worst (-3.51%, see B3) — UNH’s rally was idiosyncratic, not sector-wide. Bouncing off a -16% 1Y / -27% 3Y drawdown (see B4 Year column).

What It Means:Counter-trend bounce off washed-out positioning. The MCR trajectory and AI cost-take-out story give UNH idiosyncratic upside even as the broader Healthcare sector remains the structural laggard 3M-deepening signal.

What to watch:Q2 MCR trajectory; whether peers (HUM, CI, CVS) post similar cost-management progress to validate the sector inflection.

3. GE Vernova (GEV): +14.6% wk | Electrification Books $2.4B in Q1 Data-Center Orders Alone — More Than All of 2025; FY Outlook Raised

The Numbers:Q1 revenue $9.34B; net income $4.75B; basic EPS $17.65. Backlog grew to $163B (+$13B sequential). FY26 outlook raised to $44.5–45.5B (post Prolec GE acquisition). $200B backlog target pulled forward to 2027 (from prior). Reported Apr 22; +13.75% on day.

The Problem/Win:The single number that mattered: Electrification booked $2.4B in Q1 data-center orders, more than the segment booked in all of 2025. AI-data-center power demand has migrated from announcement-stage to balance-sheet-stage commitment.

The Ripple:Validates the AI-power-grid thesis that motivated Friday’s NVDA-Oklo-Los Alamos partnership and HSBC’s Oklo Buy/$96 PT initiation. Sets up positive read-through to Eaton (ETN), Vistra (VST), Constellation Energy (CEG), Quanta Services (PWR) as power-grid buildout pure-plays. GEV is now +75.83% YTD per B4.

What It Means:Power constraints, not chip supply, are now the binding constraint on AI capex. GEV (and adjacent industrials) becomes the second-derivative trade if hyperscaler capex commentary next week confirms the spending arc.

What to watch:Hyperscaler capex calls Apr 28–30 for explicit nuclear/grid line-items. ETN, PWR earnings in coming weeks for sector confirmation.

4. IBM (IBM): -8.48% wk | Solid Beat on Top and Bottom, but FY Guidance Maintained Rather Than Raised — Software-Sector Contagion

The Numbers:Revenue $15.92B (+9% YoY) vs $15.62B consensus. Non-GAAP EPS $1.91 vs $1.81. FY revenue growth maintained at >5% constant currency; FY free cash flow guide kept at +$1B YoY. No upward revision despite the beat. Reported AMC Apr 22; -8.25% Apr 23.

The Problem/Win:The print itself was clean — double-beat with constructive segment commentary. Management framed the unchanged guide as “prudent” given macro / geopolitical headwinds. The market read it as either confidence-deficit signal or front-of-house caution about Q2 / H2.

The Ripple:IBM’s -8% reaction landed the same day as ServiceNow -18% on margin-guide cut, snapping XLK’s 16-day winning streak (see C story 7). Investors generalized: “beat-and-maintain” in software is a sell. IBM is now -21.68% YTD per B4. Notably the SOXX index continued higher Thursday on the chip side — the AI bifurcation between hardware and software hardened.

What It Means:The market is no longer crediting beat-and-maintain; it requires beat-and-raise to compensate for AI-build-out margin compression risk. Mega-cap software (CRM, ORCL, ADBE) faces an elevated bar through next earnings cycle.

What to watch:MSFT and ORCL forward guides; whether IBM management offers a Q2 raise on the next print.

5. Tesla (TSLA): -6.07% wk | EPS Beat Masks 7,600-Unit Delivery Miss, 50K Inventory Buildup, Energy Storage -38% QoQ, Capex Raised to $25B

The Numbers:Q1 adjusted EPS $0.41 vs $0.37 consensus (+11% beat). Revenue $22.39B vs $22.64B miss. Deliveries 358,023 vs 365,645 consensus (miss by 7,600 units, but +6.3% YoY). Production exceeded deliveries by >50,000 units — significant inventory build. Energy storage 8.8 GWh vs 12–14 GWh consensus — 38% drop from Q4 2025’s record 14.2 GWh; Energy segment revenue $2.41B (-12% YoY). 2026 capex raised to $25B (from $20B prior). Reported AMC Apr 22; -2% Apr 23.

The Problem/Win:Three structural concerns surfaced in one print: (1) the 50K inventory build signals demand-not-logistics issue; (2) energy-storage halving breaks the “Tesla-as-AI-power” narrative just as GEV demonstrates the data-center-power thesis; (3) capex raise of $5B mid-year is unusual and stresses cash conversion.

The Ripple:TSLA -16.33% YTD per B4 vs S&P +0.55% WoW. The energy-storage miss is the most damaging cross-read — competitors GEV (+13.75% Wed) booked $2.4B in data-center orders the same week TSLA disclosed energy storage halving. Consumer Cyclical sector finished -1.31% on the week (see B3).

What It Means:Tesla’s growth-stock multiple compresses if both auto demand AND energy-narrative falter simultaneously. The “optionality on AI/energy/robotics” thesis that supported the multiple is being tested.

What to watch:Q2 deliveries (early July); whether inventory clears or grows further; energy-storage rebound in Q2 or sustained decline.

Q1 2026 earnings season accelerates dramatically next week, with the four largest hyperscalers all reporting and providing the critical AI-capex validation event for the broader market.

Verizon Communications (VZ) — BMO Monday April 27 — telecom bellwether ($196B cap); watch postpaid net adds, broadband subscriber trends, and 2026 capex/free-cash-flow guidance.

Cadence Design Systems (CDNS) — AMC Monday April 27 — chip-design EDA bellwether ($92B cap, just below mega-cap line but a key AI-adjacent read); watch design-IP backlog as a leading indicator of semi-cycle durability.

Alphabet (GOOGL) — week of April 28 — Google Cloud growth and AI-infrastructure capex; Cloud margin trajectory; Search ad-monetization commentary in light of generative-AI competition.

Microsoft (MSFT) — week of April 28 — Azure capacity/capex; impact of voluntary buyout on cost structure; AI-revenue contribution to Intelligent Cloud segment.

Meta Platforms (META) — week of April 28 — Updated 2026 AI capex guidance vs the announced $135B figure; justification for the 8,000-job cut; Reality Labs operating-loss trajectory.

Amazon (AMZN) — week of April 28 — AWS revenue growth; Anthropic commercial milestones; cloud-margin trajectory; retail-margin pass-through of any tariff or input-cost pressures.

Apple (AAPL), Eli Lilly (LLY), Mastercard (MA), Caterpillar (CAT), Merck (MRK), and ConocoPhillips (COP) are also scheduled for April 30 — Q1 2026 earnings season is now in full swing.

1. Does the cumulative hyperscaler capex print clear $520B? The S&P/Nasdaq records are now levered to GOOGL/MSFT/META/AMZN AI-capex commentary Apr 28–30. Consensus is ~$520B for 2026; below $500B risks a sharp semi-rally unwind given how narrow Friday’s leadership was. Above $540B and the AI-power thesis (GEV, OKLO, ETN) gets a second leg.

2. Does April PCE Apr 30 validate the Cleveland nowcast and UMich expectations spike, or fade them? Core PCE above 3% YoY paired with the 4.7% 1-year UMich expectations un-anchors the Fed’s hold optionality. Sub-3% with cool services keeps Polymarket cut-odds from sliding further from this week’s 59% level.

3. Will the Senate Banking Warsh confirmation vote hand the bond market its next inflection? Warsh confirmed within the week opens the door to balance-sheet-runoff regime change repricing — expect 10Y back above 4.50% within 30 days if confirmation lands cleanly. A second Tillis-style block extends the personnel-risk window into the May FOMC.

4. Did Saturday’s Witkoff-Kushner Pakistan talks produce a tradable Hormuz outcome? Brent’s Sunday-evening Asia open is the cleanest single signal: a gap below $95 = breakthrough (collapses the Cleveland nowcast story); above $115 = breakdown (locks the stagflation tape into the PCE print).

5. Does the “beat-but-not-enough” reaction pattern persist, or do hyperscaler beat-and-raise prints reset it? Four of the week’s top 5 decliners delivered earnings beats and were sold anyway (RTX, IBM, GE, TSLA — see B4). Whether GOOGL/MSFT/META/AMZN can both raise guidance AND defend margins through AI capex compression determines whether the Q1 reaction asymmetry was a tell or noise.

Chart of the Week: Lowry’s breadth study captures the week’s defining tension — the S&P 500 closed at fresh records yet the recent 10% pullback never produced a single 90% down-volume day, and the rebound relied on only an 80/80 buy-the-dip signal (the weakest variety in the Lowry framework). Pair this with the Dow / NYSE Composite both finishing the week red, DJ Transports declining to confirm the DJIA, and breadth narrowing into Friday’s semi-only rip, and the chart underwrites the question that defines next week: are records on this thin a foundation durable through the hyperscaler capex test?