StockResearch Intelligence Weekly #17

The Chip Rip and the Oil Unwind: AMD’s Beat Detonates Semiconductors, Brent Snaps From $115 Toward $100, and the Six-Day Window Until Warsh | Week Ending May 8, 2026

1. Beneath the Chip Rip: The Cleanest Single-Week Bifurcation of the Year

The S&P 500 printed a fresh closing high. The Nasdaq took out 711 for the first time. The Russell 2000 added almost 2%. Every major index extended April’s record month with a constructive week — but underneath, the cleanest single-week separation between AI-capex beneficiaries and energy-equity longs of 2026 played out in plain sight. Semiconductors ripped almost 12% on AMD’s blowout. Energy stocks fell more than 5% even as Brent stayed above $100. The tape no longer pretends that “AI capex” and “macro inflation” are the same trade.

Three forces drove the week and they did not align. AMD’s Q1 print on Tuesday detonated the chip complex — $10.253 billion in revenue against a $9.89B consensus, $1.37 EPS against $1.29, data center revenue up 57% to $5.8 billion, and a Q2 guide of $11.2 billion that crushed the $10.52B Street number by nearly $700 million. The stock ripped 16% Wednesday, finished the week up roughly 26%, and dragged Nvidia +8%, Broadcom +2%, and the SOXX ETF to a fresh all-time high. Brent crude collapsed from $115 toward $100 as Iran reviewed a one-page, fourteen-point US-mediated proposal to reopen the Strait of Hormuz; weekly Brent and WTI both posted losses of more than 6%, and energy-sector ETF XLE finished the week down more than 5%. The macro stack delivered mixed signals into the data vacuum before Warsh — April nonfarm payrolls printed +115K against +62K consensus, jobless claims held at a 1969 floor, but ISM Services Prices Paid stayed pinned at 70.7, the highest reading since October 2022 and the largest single confirmation that the easy-disinflation phase is over.

Key themes this week:

• The AI-capex trade narrowed and deepened. AMD’s Q2 guide adds roughly $1 billion of incremental revenue versus consensus. The read-through hit Nvidia (pre-earnings May 20), Broadcom, and the entire HBM-memory supply chain. SOXX is now within 0.03% of its all-time high; the question is whether Nvidia confirms next week or pauses the rally.
• Energy stocks de-rated even as oil held above $100. The market is now openly pricing in an Iran deal that resolves Hormuz. With Brent off $115 highs toward $100 and the 60-day War Powers clock approaching, the equity market is positioned for crude to mean-revert toward $85-90 rather than persist at wartime levels.
• The labor market is cooling at the edges, not collapsing. April NFP +115K beat the +62K consensus comfortably, March revised up to +185K, and weekly claims stayed in 1969 territory. But the household survey showed unemployment held at 4.3%, AHE growth at 3.6% YoY, and the ISM Services Employment index at 48.0 — a second straight month in contraction.
• The handover begins Friday. Powell’s term as chair ends May 15. Warsh inherits a Prices Paid index at a four-year high, a fragmented FOMC, and six weeks of vacuum until the June 16-17 meeting. The market is pricing patience.

2. Strengths: Chip Beats Drive the Cohort, Healthcare Holds, Cross-Asset Discipline

• AMD’s print reframed the AI capex story from narrative to income statement. Data center revenue of $5.8 billion (up 57% YoY) now represents 56% of total revenue, and Q2 guidance of $11.2 billion implies roughly 46% year-over-year growth at the midpoint, with non-GAAP gross margin guided to expand from 55% to 56%. Goldman Sachs and Bernstein upgraded to Buy on the print; the stock added more than $150 billion of market cap in a single week. The read-through is the binding evidence behind SOXX printing +11.7% in five trading days and within 0.03% of its all-time high. The May 6 PT raise wave — KeyBanc to $530, Bernstein to $525, Cantor/Barclays/Roth to $500, Truist to $478, Susquehanna to $450, Goldman to $450, Morgan Stanley to $410 — pushed the freshly-revised analyst mean near $466. AMD at $455 closed the week just inside that post-print band, not above it. The asymmetric trade is no longer “buy AMD here at $455 after a 26% week” — it is “buy what AMD told you about Nvidia, Broadcom, and the memory complex” before those second-derivatives have fully repriced.
• Cirrus Logic confirmed the iPhone supplier read-through with a clean beat. Q4 FY26 revenue of $448.5 million came in above the midpoint of management’s $410-470 million guide; non-GAAP EPS of $1.95 beat the $1.76 consensus by 11%. Full-year FY26 revenue hit a record $2.0 billion (up 5% YoY) and non-GAAP EPS of $9.26 was up 23% from FY25’s $7.54. Management ended the year with $1.2 billion in cash, no debt, and an explicit reference to early-content customer ramps driving Q1 FY27 guidance. The print validated last week’s call on Apple’s beat-and-raise and on the Cirrus Face ID design win. CRUS was up 4.5% on the week; the cohort cohort — TXN, ADI, MCHP — also caught a bid as memory-pricing concerns gave way to volume-confirmation signals.
• Healthcare quietly held its leadership rotation candidate status. XLV finished the week down 1.15%, modestly underperforming the index for the second straight week. But Eli Lilly stayed within 1.5% of last week’s close even as the broader healthcare cohort drifted — and the analyst flow has been a one-way street since the April 30 Q1 beat. Barclays raised its target to $1,400 from $1,350 on May 5; Morgan Stanley raised to $1,344 from $1,327 on May 1; Cantor Fitzgerald raised to $1,230 from $1,205 the same day. Consensus mean now sits at $1,261.11 across covering analysts — the highest aggregate target in the mega-cap healthcare universe. Honest caveat: LLY still sits roughly 16% below its 52-week high near $1,134, and the rotation thesis is patience-dependent. If AI doesn’t consolidate, the call doesn’t pay.

3. Weaknesses: Energy Disconnect Resolved Down, Software Bifurcation, Stagflation Print

• Energy equity went the wrong way on the right macro setup. XLE finished the week down 5.35%, USO down 6.45%, and the major integrated names — XOM, CVX, COP — all closed the week negative even as Brent held above $100 every trading day. The equity market is signaling disbelief in $100-plus crude as a sustainable level: the moment a credible diplomatic path opens, oil retraces toward $85-90 and the energy long is wrong. The Iran-US one-page fourteen-point MOU under review by Tehran is the single largest swing variable of the next four weeks. If Iran accepts, Brent breaks $90 in days; if Iran rejects and missiles resume across Hormuz, Brent can re-test $115. The energy-equity weakness through both regimes suggests the market is no longer trading the macro — it is trading the binary.
• The software bifurcation deepened with Palantir down 4% on the week. IGV finished the week up 5.22% — a respectable bounce — but the cohort dispersion is wider than the index reads. Palantir gave back 4.35% post-print on guidance concerns; Microsoft eked out a 0.16% gain off a low base; ServiceNow, Salesforce, and Workday remain in valuation purgatory ahead of late-May prints. The AI-disruption-of-SaaS narrative has not gone away — it has just been temporarily overshadowed by AMD’s hardware beat. The risk: when CRM, NOW, and ADBE print in the next 30 days, any sign of net-new subscription growth slowing puts the IGV bounce back at risk.
• The stagflation print got confirmed across both ISM surveys. ISM Services PMI for April came in at 53.6, holding 22 straight months of expansion, but Prices Paid stuck at 70.7 — the highest reading since October 2022. New Orders dropped 7.1 points to 53.5, the largest single-month decline in over a year, and Employment held at 48.0 in contraction. The manufacturing PMI told the same story from a different angle: 52.7 composite, with Prices surging at the fastest pace since April 2022. The dual signal — sticky prices, softening orders, contracting services employment — is the textbook stagflation setup. Warsh inherits this Tuesday. The market is now openly pricing rate cuts in 2026 closer to “if” than “when,” and the June 16-17 FOMC has become the most important set-piece of the next ninety days.

4. Weekly Market Recap — Week Ending May 8, 2026

The week’s signature: a fresh closing high on narrowing leadership, with semiconductors carrying nearly all of the upside. S&P 500 and Nasdaq each printed all-time-high closes; semis ripped on AMD’s blowout while energy collapsed on Iran de-escalation reads. The cross-asset signal de-risked in step — gold up on inflation persistence, oil down on supply normalization, the dollar flat, and the long bond mostly unchanged. Equal-weight underperformed cap-weight by hundreds of basis points for the third week running. The cap-weighted index is doing nearly all of the heavy lifting — which is a feature, not a bug, of an AI-capex-led tape.

5. Top 3 Stock Picks for Long-Term Investment

#1: NVIDIA Corporation (NVDA)

Price: $215.20 | Market Cap: $5.23T | 52W Range: $120.28 – $217.80 | Earnings: May 20, 2026 AMC

Thesis: Nvidia is now the cleanest pre-event setup in the megacap S&P universe. The May 20 Q1 FY27 print is the single largest binary of the next ten trading days, and the read-through from AMD’s Tuesday beat — Data Center revenue up 57%, Q2 guide of $11.2 billion crushing consensus by nearly $700 million — is unambiguous. AMD’s commentary on inferencing and agentic AI driving compounding demand for high-performance accelerators is exactly the demand backdrop Jensen Huang has been describing since GTC. Hyperscaler capex remains the binding force: Microsoft, Alphabet, Amazon, and Meta have collectively guided to roughly $670 billion of 2026 spend, and UBS now models the aggregate at $770 billion by year-end. Nvidia captures the majority share of that AI-accelerator wallet. Analyst consensus mean target sits at $275.74 across 37-42 covering analysts (a $400 high from Tigress, a $360 mid-range from multiple firms, a $260 low from Morgan Stanley), implying roughly 28% upside to the mean from Friday’s close.

Why now: NVDA closed Friday at $215.20, within 1.2% of its 52-week high of $217.80, with a 50-day moving average of $188.65 and 200-day MA of $184.73. The Q1 FY27 print on Wednesday, May 20 is the next discrete catalyst — consensus EPS $1.76 on revenue of $78.4 billion, implying roughly 77% year-over-year growth. The setup is favorable: AMD beat consensus by ~3.7% on revenue and gave a Q2 guide ~7% above Street; if Nvidia delivers similar magnitude on a much larger base, the print drives both NVDA and the broader SOXX complex through fresh highs. The Friday close at $215 plus 50-day MA at $188 leaves room to accumulate on any pullback before the print. With SOXX +11.7% this week and at all-time highs, mean reversion from extension is the primary near-term risk.

Key Levels: Support at $200 (psychological / pre-AMD-print level), $188.65 (50-day MA), and $184.73 (200-day MA, deeper support). Resistance at $217.80 (52-week high), then $250-260 (Wedbush/Wells Fargo target cluster). A weekly close above $220 on volume opens the path to $250 into the May 20 print; a break below $200 ahead of earnings means the trade is wrong on timing.

Risk Factors: Stretched setup at all-time highs ahead of binary print; consensus already expects 77% revenue growth — even a clean beat could disappoint if the magnitude is not large enough; China revenue headwinds (H20 restrictions cost an estimated $4-5B in 2026); custom-silicon competition from Broadcom and hyperscaler-internal designs (Google TPU, Meta MTIA, Amazon Trainium) is structurally real over 2027-28; the 37x forward P/E leaves no margin if Vera Rubin platform ramp slips.

Verdict: BUY with discretion on pullbacks to the $200-208 band. My 12-month target band: $275 base case (consensus alignment), $325 stretch on a clean May 20 guide-up plus Vera Rubin ramp commentary. Stop discipline below $185.

#2: Eli Lilly and Company (LLY)

Price: $948.45 | Market Cap: $893.2B | 52W Range: $623.78 – $1,133.95 | Earnings: Last reported April 30 (Q2 due August 5)

Thesis: Eli Lilly is the cleanest non-tech secular growth story in the S&P 500 and the highest-conviction answer to the concentration-risk question that defined April. Q1 2026 revenue grew 56% YoY to $19.8 billion, beating consensus by $2 billion. Adjusted EPS of $8.55 beat the $6.97 consensus by nearly 23%. Management raised full-year revenue guidance to a midpoint of $83.5 billion (28% growth versus 2025) and lifted adjusted EPS guidance to $35.50-37.00. The growth engine remains Mounjaro and Zepbound; Foundayo, the once-daily oral GLP-1 launched this spring, reportedly reached 20,000-plus patients in launch weeks per CEO Ricks.

Why now: The analyst flow since the April 30 print has been a one-way street. Barclays raised to $1,400 from $1,350 on May 5. Morgan Stanley raised to $1,344 from $1,327 on May 1. Cantor Fitzgerald raised to $1,230 from $1,205 on May 1. Jefferies sits at $1,300. Consensus mean across covering analysts now stands at $1,261.11 — a high of $1,400, a low of $1,070, a median of $1,280 — implying roughly 33% upside from Friday’s $948.45 close. With XLV down again this week and the broader healthcare cohort still discounting tariff and policy uncertainty, LLY is functioning as the one-stock barometer for what works when AI consolidates for breath. The stock sits within 1% of its 50-day moving average of $941.78 and roughly 3.5% above the 200-day at $916.32 — clean support structure for accumulation.

Key Levels: Support at $941.78 (50-day MA — the recent base), $916.32 (200-day MA), then $870 (mid-March pullback). Resistance at $1,000 (psychological), $1,133.95 (52-week high, January top), then the $1,261 analyst target band. A weekly close above $1,000 on volume opens the path back to the 52-week high.

Risk Factors: Pricing pressure remains real — management guided to a low-to-mid-teens price headwind for the full year. FDA proposed earlier this month to exclude tirzepatide active ingredients from compounding-eligible lists (net favorable but adds short-term policy noise). Execution risk on Foundayo manufacturing scale-up at the new Indianapolis and Lebanon facilities. Competitive dynamics with Novo Nordisk’s pipeline. M&A integration risk (Kelonia, Ajax, multiple smaller deals) into 2027.

Verdict: BUY. Consensus mean target $1,261.11 (+33% upside), median $1,280, high $1,400. My 12-month target band: $1,260 base case (consensus alignment), $1,400 stretch if Foundayo launch metrics keep accelerating and FY27 EPS approaches $44+. Stop discipline below $870.

#3: Microsoft Corporation (MSFT)

Price: $415.06 | Market Cap: $3.08T | 52W Range: $356.28 – $555.45 | Earnings: Last reported April 30 (Q4 due July)

Thesis: Microsoft is the beaten-down value entry inside the AI-capex trade — and the cleanest rotation candidate inside the Mag 7 cluster. The stock fell 3-4% post-Q3 print on capex absorption concerns, even as Azure printed close to 40% growth and the company guided 2026 capex to roughly $190 billion. Two weeks later, AMD’s blowout has reframed exactly that capex commitment: the spend is showing up as compute revenue at AMD, at Nvidia (May 20 print), and at the memory-and-substrate complex. The market punished Microsoft for being the spender. The bull case is that the spend was always going to be a feature, not a bug — and that the post-print weakness offers an entry to a name that compounds free cash flow at scale once the build-out matures. Consensus mean target sits at $556.88 across 28+ covering analysts (a $680 high, a $415 low, a $550 median), implying roughly 34% upside from Friday’s $415.06.

Why now: MSFT closed Friday at $415.06, sitting 25% below its 52-week high of $555.45 and roughly 11% below the 200-day moving average at $465.66. The Q3 print on April 30 already absorbed the capex-shock reaction — there is no immediate binary in the next four weeks. With IGV up 5.22% this week, the software cohort is showing the first real signs of rotation back to large-cap AI-integrated platforms. The asymmetric setup is to accumulate at-or-below the 200-day MA into the late-May/early-June software prints; CRM, NOW, and ADBE all report into the late-May/early-June window, and clean prints from any of them re-rate the broader cohort multiple upward, with Microsoft as the largest beneficiary.

Key Levels: Support at $400 (psychological / recent post-print low), $398.15 (50-day MA), then $356.28 (52-week low). Resistance at $465 (200-day MA), $500 (psychological), then the $556.88 consensus target band. A weekly close above $465 on volume signals the post-print capex narrative is being absorbed.

Risk Factors: Capex spending is now stretched relative to operating cash flow; if Azure growth decelerates below 35% on any subsequent print, the multiple compresses; OpenAI/Microsoft commercial relationship continues to draw scrutiny, with The Wall Street Journal reporting concerns last Friday about OpenAI revenue and growth that may flow back to Microsoft; software-cohort AI displacement (Copilot displacing point-solution SaaS) cuts both ways — Microsoft wins gross, but the burden of proof on subscription net-new keeps rising.

Verdict: BUY for patient investors with a 12-month horizon. My 12-month target band: $500 base case (200-day MA + earnings recovery), $560 stretch on consensus alignment plus Azure re-acceleration. Stop discipline below $390.

6. Low-Cap Opportunities

Three setups, three different verdicts, three discrete catalysts. All three are true low-caps under the $15B ceiling.

Oklo (OKLO) — Nuclear-AI Energy, Pre-Print, Catalyst Tuesday

Price: $72.51 | Market Cap: $12.6B | 52W Range: $28.16 – $193.84 | Earnings: Tuesday, May 12, 2026 AMC

Oklo is the cleanest expression of the nuclear-AI energy thesis at sub-$15B market cap. The company’s Aurora Powerhouse is a 15-75 megawatt fast reactor designed for data center and industrial deployment; the customer pipeline now stands at roughly 14 gigawatts — driven by the 12 GW corporate power agreement with Switch signed in 2025, plus letters of intent with Equinix (500 MW with a $25M prepayment), Prometheus Hyperscale (100 MW), and Diamondback Energy (50 MW). The balance sheet ended Q3 2025 with approximately $1.2 billion in cash and marketable securities and no debt, following the $540 million Q3 equity raise and a $400 million shelf offering in December. Three catalysts have stacked in the past three weeks: the April 23 partnership with Nvidia and Los Alamos National Laboratory to advance nuclear fuel validation for AI-focused facilities; the NRC’s expedited approval of the Aurora reactor’s Principal Design Criteria last week (which triggered a 13% same-day move); and Tuesday’s Q1 financial results and business update. Tigress Financial initiated with a Buy and $130 price target on April 27; HSBC initiated with Buy and $96 on April 23; Texas Capital Securities maintained Buy with $120 after the NRC approval. Consensus mean across covering analysts is $114.50, with a high of $150 (Seaport Global) and a low of $82 (Barclays) — implying roughly 58% upside to the mean from Friday’s close.

Why now: Oklo reports Q1 financials and a business update after the close on Tuesday, May 12 — the company is pre-revenue, so EPS of -$0.20 (consensus) is not the signal. The signal is management commentary on customer pipeline progression, the deployment timeline for the first commercial Aurora at INL Idaho, fuel recycling progress, and any new agreements with hyperscalers or sovereign customers. With SMR, the publicly-traded nuclear peer, reporting a quarter earlier and printing softer-than-expected, expectations into Tuesday are reset. The asymmetry is reasonable: $58 stop versus an upside band that consensus has at $114.

Verdict: SPECULATIVE BUY at $70-75 with a $58 stop (below the 50-day MA $60.82 by a meaningful margin). My target band: $96 conservative (HSBC alignment), $115 base case (consensus mean), $150 stretch on a clean Tuesday print plus visible hyperscaler deal commentary. Smaller-than-normal position size given binary print risk.

Cirrus Logic (CRUS) — Post-Print, Held the Line, Near Consensus

Price: $171.67 | Market Cap: $8.8B | 52W Range: $92.02 – $179.00 | Earnings: Last reported May 6 (Q1 FY27 due August 4)

Cirrus Logic delivered exactly the print that last week’s call anticipated. Q4 FY26 revenue of $448.5 million came in above the midpoint of management’s $410-470 million guidance range. Non-GAAP EPS of $1.95 beat the $1.76 consensus by 11%. Full-year FY26 revenue hit a record $2.0 billion and non-GAAP EPS of $9.26 was up 23% year-over-year. Management ended the year with $1.2 billion in cash, no debt, added a $350 million credit line, and called out continued strong demand for the latest custom-boosted amplifier and 22-nanometer smart codec. The Face ID design win with Apple — a smart power IC for 3D sensing — was reiterated as the first product in an entirely new application space. The print was clean, the read-through to Apple’s iPhone supplier complex was confirmed, and CRUS finished the week up 4.5%.

The issue is the price has caught up to most of the thesis. CRUS closed Friday at $171.67, within 4% of its 52-week high of $179. The post-print analyst flow has been positive — KeyBanc raised to $200 from $175 on May 6, Stifel raised to $197 from $185 the same day — putting the freshly-revised mean target near $190 (the FMP aggregate at $169 hasn’t yet refreshed for these revisions). CRUS at $171 now sits ~10% below the post-print mean, with a high target of $200 implying ~17% upside and a low of $110 implying ~36% downside. With no near-term catalyst until Q1 FY27 prints in August, the asymmetry has narrowed but not closed.

Verdict: HOLD at current $171; SPECULATIVE BUY only on pullback to the $155-162 band with a $148 stop. My target band: $190 base case (post-print consensus alignment), $200 stretch (KeyBanc high target). Don’t chase at $171 after a +4.5% print-week; the post-print rally has discounted most of the near-term setup, and the next discrete catalyst is three months out.

AeroVironment (AVAV) — No Catalyst, Below Entry, Wait

Price: $168.29 | Market Cap: $8.4B | 52W Range: $154.39 – $417.86 | Earnings: Q4 FY26 due June 23, 2026

AeroVironment broke below last week’s $180-195 SPEC BUY entry zone with a 9% decline, but never triggered the $155 stop. The thesis remains structurally intact — Switchblade family loitering munitions, the LOCUST X3 platform launch, BlueHalo acquisition expanding the franchise into space, cyber, and directed energy, and a $4.6 billion year-to-date backlog. Consensus mean price target sits at $343.60 across covering analysts, with a high of $450, a low of $236, and a median of $330 — implying more than 100% upside to the mean. But the dispersion is the real story: the post-Q3-miss rerating cut Jefferies, Baird, and UBS targets meaningfully even as ratings stayed at Buy/Outperform.

The discipline here is patience. The Q4 FY26 print is not until June 23, capturing the post-reset operational period. There is no near-term catalyst that justifies adding before then, and the $168 close sits below the $180 entry zone but above the $155 stop — a no-man’s-land for new positions. Existing holders should reduce or wait; new entrants should wait for either (a) a pullback to $155-165 with a tighter stop, or (b) the June 23 print to provide a fresh setup. With XLI flat on the week and defense names showing rotation softness, the broader cohort isn’t yet supporting a tactical re-add.

Verdict: NO TRADE / REDUCE existing positions. Wait for the June 23 print or a pullback below $160 for a tighter stop. The thesis is not broken — but the timing isn’t right for fresh capital. My target band on the eventual setup: $235 conservative (Baird/UBS floor), $310 base case (post-cuts consensus midpoint), $450 stretch only on Iran tail risk reasserting.

7. Strategic Recap & Forward Outlook

April closed the strongest broad-index month since 2020, and the first week of May extended the gains — but the narrowing of leadership accelerated. The Mag 7 is now visibly three trades: AI-capex spenders being repriced for what they’re getting (Microsoft, Alphabet), AI-capex beneficiaries being repriced for what they’re selling (Nvidia, AMD), and value-trap defensives (Apple, Meta, Tesla) that depend on a non-AI thesis to keep working. The chip rip plus the oil unwind plus the Friday Fed handover puts the next four weeks into a new regime: the next FOMC isn’t until June 16-17, the next mega-cap binary is Nvidia on May 20, and the macro calendar between now and then is quiet. Patience over conviction. Quality over beta.

The trade is to hold the names you’d own anyway, define risk explicitly with stops that respect each ticker’s volatility, and use any pullback into the Nvidia print as accumulation rather than chase. If May 20 confirms hyperscaler demand by delivering a guide-up against an already-elevated $78B revenue consensus, the AI-capex tape extends through June. If Nvidia disappoints or merely meets, the cohort rolls and rotation to healthcare, defensives, and beaten-down software accelerates fast. Either way, the next 30 days are a rotation tape, not a beta tape — and the data, not the narrative, will do the choosing.

Forward Outlook — Week of May 11–15, 2026

This is the first full week of the Powell-to-Warsh transition, with April CPI on Tuesday as the largest single set-piece between now and Nvidia’s May 20 print.

• Monday, May 11: Light docket. Treasury 3-yr auction.
• Tuesday, May 12: April CPI (8:30 ET) — consensus +0.6% MoM headline / +3.7% YoY, +0.3% MoM core / +2.7% YoY. The first inflation print that fully captures the Iran-war energy passthrough; the key sub-line to watch is core services ex-shelter. Oklo (OKLO) Q1 results after the close.
• Wednesday, May 13: April PPI (8:30 ET) — the producer-side read on whether Prices Paid spikes in both ISM surveys are showing up in input costs. Treasury 10-yr auction.
• Thursday, May 14: Initial Jobless Claims (8:30 ET) — prior 200K, near a 1969 low. April Retail Sales (8:30 ET) — the demand-side read into Walmart, Target prints later in the month. Nano Nuclear Energy (NNE) Q1 results after the close.
• Friday, May 15: Powell’s term as Fed Chair officially ends; Kevin Warsh sworn in. Powell stays as a governor. April Industrial Production / Capacity Utilization (9:15 ET). University of Michigan May Preliminary (10:00 ET).

S&P 500 Key Levels: The index closed Friday at 7,398.93 — just 0.03% off the 52-week intraday high of 7,401.50 and within 0.04% of its all-time closing high. Support at 7,300 (psychological / pre-NFP base), 7,250 (this week’s low), then 7,165 (last week’s breakout level turned support). Resistance at 7,401 (the print-day intraday high), 7,500 (next round-number extension), 7,750 (RBC raised year-end target to 7,900 from 7,750 on Friday, putting 7,750 on the path). A clean weekly close above 7,401 signals the AI-capex tape extends through the Warsh handover; a break below 7,250 on a hot CPI or weak Nvidia guide opens the path back to 7,165.

Sector Outlook: Semiconductors (SOXX) carry the historic May momentum but at +11.7% W/W and within 0.03% of all-time highs are mechanically extended; Nvidia’s May 20 print is the next test for the cohort. Healthcare (XLV) and Consumer Staples (XLP) lagged again and remain the cleanest rotation candidates if AI consolidates. Energy (XLE) closed the week negative despite Brent above $100 — the equity market is positioned for the Iran deal. Software (IGV) up 5.22% on bouncing off the post-Mag-7-print lows; CRM, NOW, and ADBE prints in the next 30 days will set the cohort tone.

Model Portfolio Verdicts Restated

• Nvidia (NVDA) — Verdict: BUY with discretion on pullbacks to the $200-208 band. My 12-month target band: $275 base case, $325 stretch on a clean May 20 guide-up plus Vera Rubin ramp commentary. Stop discipline below $185.
• LLY $948.45: BUY | Stop $870 | Target band $1,260-$1,400
• MSFT $415.06: BUY | Stop $390 | Target band $500-$560
• OKLO $72.51: SPEC. BUY | Stop $58 | Target band $96-$150
• CRUS $171.67: HOLD (add on pullback to $155-162) | Stop $148 | Target band $190-$200
• AVAV $168.29: NO TRADE / REDUCE | Re-engage below $160 or post-June 23 print

The Bottom Line

This is the week the AI-capex trade stopped pretending to be the whole market. AMD detonated the cohort. Energy de-rated through good news (Iran de-escalation) and bad news (Brent persistence) with equal indifference — that’s the signal. The labor market is cooling at the edges but still printing claims at 1969 levels. Services inflation is pinned at a four-year high. The Fed’s most divided meeting in 34 years just happened, and the chair seat changes hands Friday. There is no clean cyclical setup here — only a rotation tape with specific binaries (Nvidia May 20, CPI May 12, Warsh first FOMC June 17) and specific risks (Iran rejection of the MOU, OpenAI/Microsoft revenue questions, hyperscaler-internal silicon competition catching up faster than expected).

What could be true on Friday afternoon: Nvidia trades within 3% of $215 going into the May 20 print, CPI core comes in at-or-near consensus (+0.3% MoM), Brent holds $95-105 on continued Iran negotiations, and SOXX consolidates the +11.7% gain without breaking. What would change the thesis: a CPI core surprise above 0.4% MoM, an Iran rejection of the MOU framework, or a Nvidia pre-announce. What discipline the reader should hold to: position-size to volatility, respect the stops set in this section, and resist chasing the 7,401 ATH on extension before Tuesday’s CPI confirms the disinflation runway.

The data will choose. The narrative is a recap.

Subscribe

Disclaimer

This analysis is for informational purposes only and should not be construed as investment advice. Equity markets carry substantial risk, including the potential loss of principal. Past performance does not guarantee future results. Price targets are based on analytical frameworks and current market data but involve significant uncertainty. Always conduct your own due diligence and consult a qualified financial advisor before making investment decisions.