BLUF — May 9, 2026
Cleanest Strong Add names: NOW, TSM, GOOGL, COHR — all combine accelerating fundamentals with valuation cushion.
MU and PLTR are held — do not add at current prices; wait for entry zones.
NVDA is Hold-into-May-20 print. NEE is defensive ballast, not alpha.
📊 Master Summary Table
▼
| Ticker | Layer | Price | Fwd P/E | Rev Growth | PEG | Val | Fund | Flow | Verdict | Entry Zone |
|---|---|---|---|---|---|---|---|---|---|---|
| NVDA | AI Compute | $215 | 24.5× | +73% | 0.65 | A | A+ | A | HOLD→print | $195–210 |
| TSM | AI Foundry | $400 | 26× | +41% | 0.85 | A+ | A+ | A | Strong Add | $360–395 |
| GOOGL | AI Software | $401 | 27× | +22% | 1.0 | A | A+ | A+ | Strong Add | $370–400 |
| COHR | AI Optics | $322 | 30–35× | +21% | 0.9 | A | A | B+ | Add pullback | $300–330 |
| NEE | AI Power | $93 | 23× | +10% EPS | 2.5 | B | B+ | A | Ballast | $85–92 |
| NOW | AI Workflow | $93 | 21–25× | +22% | 0.84 | A+ | A | B+ | Strong Add | $85–100 |
| PLTR ⚑ | AI Software | $144 | ~110× | +85%* | 1.5 | D | A+ | C | HOLD only | Reload $100–115 |
| MU ⚑ | HBM Memory | $757 | ~22× | +196% | <0.5 | A | A | A+ | HOLD/dip add | Add $600–650 |
⚑ = already held · * PLTR 85% is official 8-K total YoY; see PLTR tab for data conflict note
Staged Capital Deployment
Stage 1 — Immediate (50–60%)
NOW$85–100
Most compelling rerating. Fwd P/E 21-25× vs 95-110× historical. AI ACV accelerating. Iran derate is non-fundamental.
GOOGL$370–400
Long-term compounder. Cloud +63%. Cleanest AI monetisation. Accumulate gradually.
TSM$360–395
Binding AI constraint. Fwd P/E 26× for 30%+ growth. Wait for sub-$395 entry.
Stage 2 — Conditional (25–30%)
NVDA$195–225
Only after May 20 print. Need: rev beat ≥$2B, Q2 guide ≥$87B, Rubin H2 ramp confirmed.
COHR$300–330
Deploy on pullback to $300-315 OR Q4 FY26 print delivers ≥$2.0B revenue (high-end guide).
Stage 3 — Tactical (15–20%)
MU ⚑$600–650
Add to existing position only on 20%+ pullback. Do NOT chase $757.
NEE$85–92
Ballast only. Buy for portfolio defence vs AI capex digestion scenario.
PLTR ⚑$100–115
HOLD only at $144. Reload exclusively at $100-115 on 25-30% drawdown.
🚫 What NOT to Do
▼
- ✗Chase MU above $760. The +38% week reflects momentum/squeeze, not new fundamentals.
- ✗Add PLTR above $140. Burry still short, Thiel selling $580M, Jefferies $70 PT — asymmetry is wrong.
- ✗Add NVDA above $225 ahead of May 20. Goldman says bar is high. Wait 7 trading days.
- ✗Treat NEE as an AI growth story. It's a utility with AI optionality. Buy for ballast not alpha.
- ✗Add a second memory/storage name (SNDK, WDC, STX) on top of MU — duplicative exposure.
- ✗Deploy all capital at once. AI cohort hit record highs week of May 8. Stage over 6–8 weeks.
Investment Philosophy
Passive Core
40%
CSPX
Market beta · ballast · absorbs errors
Crypto
10%
IBIT
Asymmetric bet · uncorrelated upside
Active Picks
~50%
5–10 stocks + cash buffer
AI tech stack · target >42% CAGR
◆
Blended CAGR target >26% — requires the active book to compound at >42% on its allocated capital.
The honest baseline: a 10–15% long-run CAGR compounding over decades is genuinely sufficient. The active book is a disciplined attempt to exceed it — not a requirement. The passive core exists precisely so that errors in stock selection are absorbed by structure, not predicted away.
AI valuations are stretched and difficult to anchor with certainty. The appropriate response is sizing, not avoidance — concentrate in the highest-conviction names, maintain dry powder for drawdown entries, and hold nothing in the active book that the passive core could substitute. When no position meets the entry criteria, cash is the default.
Anything beyond 15% is a bonus. The goal is not to maximise returns — it is to take considered, asymmetric bets on the AI infrastructure buildout while remaining whole if the thesis is wrong or the cycle turns faster than expected.
NVDA
NVIDIA Corporation
AI Compute Core
$215.20
May 8, 2026 · Mkt Cap $5.23T
52W: $83.89 – $214.18
⏸ HOLD INTO MAY 20 PRINT
Add zone: $195–210
AI infrastructure's primary supplier. FY26 $215.9B revenue (+65% YoY). Blackwell fully ramped; Rubin ramps H2 2026. Q1 FY27 ($78B guide) reports May 20 — the critical gate.
Price
$215
TTM P/E
43.6×
Sector avg 34.7×
Fwd P/E
24.5×
PEG 0.65
Rev Growth
+73%
Q4 FY26 YoY
Gross Margin
71.1%
GAAP TTM
FCF Margin
~45%
$96.6B TTM
Q1 FY27 Guide
$78B
±2%; May 20
Valuation vs Growth
A
Fwd P/E 24.5× on ~70% growth and 75% gross margin is the best setup in megacap AI. 3Y/5Y avg P/E 63×/67× — current is steep historical discount. PEG 0.65 — lowest in universe. Q4 FY26 $68.1B, +73% YoY. Data Center $62.3B, +75%.
Justified by growth. Not stretched; discount to own history.
Fundamentals vs Hype
A+
Gross margin 71% (75% guided Q1 FY27). FCF $96.6B. Estimate revisions UP across 36 sell-side analysts. Data Center = 90%+ of revenue. China DC revenue zeroed ($50B TAM impaired). AI demand confirmed — $500B Blackwell/Rubin customer commitments through 2026.
Grounded. Best verified fundamentals in the AI universe.
Institutional vs Retail
A
66.4% institutional; 5,605 13F filers. Hedge funds net +4.15% aggregate Q1 2026. Top holders: Vanguard, BlackRock, State Street. Short interest <1%. Jensen Huang $2.9B+ cumulative insider sales under 10b5-1 (programmatic). Retail "Extremely Bullish" Stocktwits.
Institutionally driven with retail tailwind.
Add Verdict
HOLD into May 20 earnings. Goldman expects $80B and $1.86 EPS (above consensus $78B/$1.74) but explicitly notes "bar for outperformance is relatively high." Wait for the print. On a clean beat with strong Q2 guide ≥$87B and Rubin H2 confirmation: deploy in $195–225 zone. On a miss or guide-down: $180-195 re-entry zone.
Target: $195–210 (pre-print) · $215–225 acceptable on confirmed beat
⚠ Key Risks
▼
- 01China DC revenue permanently impaired — $50B addressable TAM gone absent policy reversal
- 02Rubin H2 2026 ramp slip — any commentary suggesting delays would compress multiple 20%+
- 03CoWoS packaging bottleneck at TSMC — single point of failure; TSMC capex constrained
- 04Hyperscaler capex digestion — any single MSFT/META/GOOG/AMZN cut cascades to all AI names simultaneously
- 05Insider selling — Huang new 10b5-1 plan authorises 6M more shares; aggregate $3.3B+ sold since 2024
💬 Analyst Quotes (X)
▼
"I still believe Nvidia #NVDA will reach $20 trillion by 2030, but that does not mean it is the best AI trade in 2026."
@bethkindig (Beth Kindig / IO Fund) · LinkedIn, Apr 2026
"Nvidia will be leaving out Micron HBM4 and primarily using SK Hynix and Samsung HBM4 for first year Rubin builds."
@SemiAnalysis (Dylan Patel) · Jan 28, 2026 Institutional Report
@aleabitoreddit (Serenity) — No direct NVDA-specific public quote verifiable from open-access
TSM
Taiwan Semiconductor Manufacturing
AI Foundry Bottleneck
~$400
May 8, 2026 · Mkt Cap ~$1.95T
52W: $130 – $415
✦ STRONG ADD
Entry zone: $360–395
Every NVIDIA AI chip is fabricated by TSMC. CoWoS advanced packaging is the binding constraint for the entire AI value chain. Most attractive valuation/growth combination among megacaps.
TTM P/E
~30×
Fwd P/E
24–27×
PEG 0.85
Q1 Rev Growth
+41%
USD YoY
Gross Margin
66.2%
Expanding
Op Margin
58.1%
FY26 Guide
>30%
USD raised
Capex 2026
$52–56B
Valuation vs Growth
A+
Fwd P/E 24-27× for 30%+ revenue growth. 3Y/5Y avg P/E ~22× — modestly above history. Sector median 34.7× — TSM at 25% discount to US semis peers. Q1: $35.9B (+40.6% USD), beat ~$35.5B est. Q2 guide $39-40.2B.
Most attractive valuation/growth combo in megacap AI. Geopolitical Taiwan discount is the structural overhang.
Fundamentals vs Hype
A+
Gross margin expanding: 58.8% → 62.3% → 66.2% over 4Q. HPC platform = 61% of Q1 revenue. 3nm = 25% of wafer rev. CEO C.C. Wei: AI accelerator revenue "mid-to-high 50s CAGR through 2029." Capex raised to top of range ($56B). Estimate revisions UP.
Grounded — picks-and-shovels of all AI chip designers.
Institutional vs Retail
A
Institutional ownership ~17-20% of US ADR float (most held in Taiwan). Hedge funds slightly trimmed Q4, passive flows positive. Short interest ~0.53%. Superinvestor presence on Dataroma. Retail sentiment moderate, less frothy than NVDA.
Institutionally driven.
Add Verdict
Strong Add. Best risk/reward in megacap AI outside of MU. Forward P/E 25× with 30%+ growth, 66% gross margin, AI capacity sold into 2027, and 2nm volume production H2 2026. Taiwan geopolitical risk is structural — price in a permanent discount, not a binary risk.
Entry zone: $360–395 (current ~$400; wait for sub-$395 entry)
⚠ Key Risks
▼- 01Taiwan Strait geopolitics — binary tail risk; not priceable but should shrink position vs ideal sizing
- 02CoWoS supply execution — any yield/ramp slip cascades to NVDA/AMD/AVGO timelines
- 03Margin dilution 2–3pp from N2 ramp + Arizona/Japan overseas fabs (higher cost structures)
- 04NVIDIA customer concentration ~19% of FY25 revenue — NVDA capex digestion impacts TSM directly
- 05Middle East war: helium/specialty gas cost increases; management flagged profitability risk
💬 Analyst Quotes (X)
▼"SK Hynix and Samsung are both forecast to overtake TSMC in profits in 2026… [memory's] structural transformation from commodity to AI bottleneck is the defining story of 2026."
@bethkindig (Beth Kindig / IO Fund) · LinkedIn Apr 2026 (paraphrase from article)
@SemiAnalysis, @aleabitoreddit — No direct TSM-specific quotes verifiable from open-access search
GOOGL
Alphabet Inc.
AI Software Compounder
$400.80
May 8, 2026 (all-time high) · Mkt Cap ~$4.84T
52W: $145 – $398
✦ STRONG ADD
Entry zone: $370–400
Full-stack AI: TPU chips, Gemini models, Cloud infrastructure, Search/YouTube distribution to 4B+ users. Q1 2026: $109.9B revenue +22% YoY; Cloud +63% to $20B; EPS beat by 95%. Capex $180–190B 2026.
TTM P/E
30.6×
Fwd P/E
27×
PEG ~1.0
Rev Growth Q1
+22%
11th consec. DBL-digit
Cloud Growth
+63%
$20B Q1
Op Margin
36.1%
+200bps YoY
EPS Q1
$5.11
vs $2.62 est (+95%)
Net Income
$62.6B
+81% YoY Q1
Valuation vs Growth
A
Fwd P/E ~27× for 22% revenue growth and 36% op margin. PEG ~1.0. Q1 revenue $109.9B vs $107.2B est. Google Cloud backlog $462B (nearly doubled QoQ). 11 consecutive quarters of double-digit growth. 3Y/5Y avg P/E ~24×/26× — near historical average.
Justified by growth — most balanced AI monetisation in megacap software.
Fundamentals vs Hype
A+
Cloud op margin 32.9% (vs 17.8% prior year — doubled in 4 quarters). Gemini API 16B tokens/min (+60% QoQ). GenAI Cloud products +800% YoY. 350M paid AI subscriptions. AI Overviews not cannibalising Search — +19% YoY. Estimate revisions UP.
Grounded. Cleanest full-stack AI monetisation anywhere.
Institutional vs Retail
A+
~70% institutional. Net buying Q1 2026 (Fisher, other growth managers). Disciplined long-only ownership base. Short interest <1%. Superinvestor presence on Dataroma (value-tilted managers). Retail bullish but less frothy than NVDA/PLTR.
Institutionally driven — disciplined long-only base.
Add Verdict
Strong Add. Accumulate at $370–400. Long-term compounder — the only company that owns chips (TPU), models (Gemini 3), infrastructure (GCP), and global distribution (Search, YouTube, Android). Cloud growth +63% vs AWS ~17% represents a serious competitive gap being established. The stock hit all-time high $398.37 May 8 — new highs often beget new highs in genuine compounders.
Entry zone: $370–400 (accumulate; current is ATH territory)
⚠ Key Risks
▼- 01Capex $180–190B in 2026 weighs on FCF; depreciation ramp will compress margins 2027
- 02Search disruption — AI Overviews currently stable but monetisation per query unclear long-term
- 03EU $11B fine pending; US DOJ ad-tech monopoly remedies; breakup risk non-zero
- 04Cloud growth deceleration: Q2 will test whether +60%+ is sustainable or Q1 was pull-forward
- 05AI infrastructure overbuild — Pichai noted "compute constrained in near term"; $462B backlog requires delivery
💬 Analyst Quotes (X)
▼@aleabitoreddit, @bethkindig, @SemiAnalysis — No direct GOOGL-specific quotes from tracked analysts verifiable from open-access search for this period
COHR
Coherent Corp.
AI Optical Picks-and-Shovels
~$322
May 7, 2026 (fell 6.7% post-Q3 print) · Mkt Cap ~$66B
52W: $85 – $380
✦ STRONG ADD POST-PULLBACK
Entry zone: $300–330
800G and 1.6T optical transceiver supplier for hyperscaler AI networking. NVIDIA invested $2B (March 2026). Record bookings; backlog extends to calendar 2028. GAAP P/E misleading — use non-GAAP ~30-35× fwd.
Fwd P/E (non-GAAP)
30–35×
GAAP P/E
~316×
II-VI amortisation
Q3 FY26 Rev
$1.806B
+21% YoY
Q4 Guide
$1.91–2.05B
Non-GAAP EPS
$1.41
vs $1.43 est (miss)
Gross Margin
39.6%
Expanding
Analyst PT Range
$170–455
⚠ Wide dispersion
Valuation vs Growth
A
Non-GAAP fwd P/E ~30-35× on 30%+ EPS growth. PEG ~0.9. Q3 miss was $9.7M on $1.8B revenue (0.5%) — noise, not signal. Q4 guide midpoint $1.98B implies sequential acceleration. IO Fund entered $185 Dec 2025 (+72%).
Most compelling rerating in optical AI. Minor miss overreaction.
Fundamentals vs Hype
A
DC & Comms = 75% of revenue, +40%+ YoY. Gross margin expanding +105bps YoY. Record bookings. Backlog LTAs extending to end of decade. NVIDIA $2B equity investment (Mar 2026) validates strategic positioning. S&P 500 inclusion Mar 2026 = structural bid. 800G → 1.6T transition accelerating.
Grounded. Record backlog + NVIDIA partnership = real, not hype.
Institutional vs Retail
B+
84% institutional ownership. Top holders: FMR/Fidelity ~12%, Vanguard 7.45%, BlackRock 7.6%. Heavy insider selling (22 sales/0 buys 6mo) — but Bain Capital legacy distribution from II-VI conversion, not panic. S&P 500 passive bid. Not a retail darling.
Institutionally driven. Insider selling = technical, not signal.
Add Verdict
Strong Add on pullback. The 6.7% post-earnings drop on a 0.5% revenue miss vs $1.8B is a non-fundamental derate. Q4 guide acceleration to $1.91–2.05B (sequential +12%) is the more important signal. At $300-315 the risk/reward is asymmetric. PT range $170-455 reflects analyst uncertainty, not stock weakness.
Entry zone: $300–330 (current $322 is acceptable; $300-315 preferred)
⚠ Key Risks
▼- 01AI capex digestion: COHR is 75% hyperscaler-driven — any slowdown in MSFT/META/GOOG capex hits volumes directly
- 02Lumentum competition — NVIDIA also invested $2B in Lumentum (same day as COHR); margin pressure from duopoly dynamics
- 03Indium Phosphide ramp execution — capacity doubling by Q4 FY26; yield ramp risk
- 04PT range $170–455 is widest in cohort; represents analyst inability to model transition — high multiple compression risk if narrative cracks
- 05Industrial segment soft ($444M, down YoY) — mixed earnings quality; AI wins masking legacy weakness
💬 Analyst Quotes (X)
▼IO Fund entered COHR at $185 in December 2025 (confirmed in attached reference document) — up ~72% as of May 2026. IO Fund's thesis: record bookings + NVIDIA partnership + 800G/1.6T demand inflection = "best execution story in AI networking/optics complex."
@bethkindig (Beth Kindig / IO Fund) · via IO Fund portfolio note referenced in attached docs
@SemiAnalysis — Covers optical ecosystem (connector content for NVIDIA NVL72) but no direct COHR quote verifiable from open-access. @aleabitoreddit — No COHR coverage from tracked analysts on X verifiable.
NEE
NextEra Energy, Inc.
AI Power Infrastructure
$93.20
May 9, 2026 · Mkt Cap ~$194B
52W: $63 – $101
◆ HOLD / ADD FOR BALLAST
Entry zone: $85–92
Largest US clean energy utility (FPL regulated + NEER renewables). AI data center power demand creates multi-decade offtake thesis. Dividend +10% 2026, +6% through 2028. 8%+ adj EPS CAGR through 2032. Not alpha — portfolio ballast.
Fwd P/E
22–23×
Sector: 15×
PEG
~2.5
On 8% EPS growth
Q1 EPS Beat
$1.09
vs $1.02 est (+7%)
Q1 Rev Miss
$6.70B
vs $7.35B est (-9%)
Dividend Yield
~2.7%
+10% growth 2026
Net Debt
$104B
Rate-sensitive
EPS CAGR
8%+
Through 2032
Valuation vs Growth
B
Fwd P/E 22-23× on 8% EPS growth. Utility sector median 15× — NEE trades at 54% premium. Premium justified only by AI power optionality and renewables scarcity value. PEG 2.5 on pure utility basis. Dividend yield 2.7% limited cushion at this multiple.
Premium priced vs peers. Multiple only justified if AI power demand materialises at scale.
Fundamentals vs Hype
B+
Record clean energy backlog. 9.5 GW landmark gas projects with Japan. Duane Arnold nuclear restart Q1 2029. Google Cloud partnership (Rewire initiative). Q1 EPS +10% YoY. FCF negative (-$15.4B TTM) due to $27.7B capex. Debt $104B — rate-sensitive.
Grounded — real AI infrastructure thesis but slow. Not hype; not fast.
Institutional vs Retail
A
80%+ institutional (typical utility). Vanguard, BlackRock, State Street dominate. Quiet, stable flows. Short interest <2%. No ARK position. Retail is income-focused, not momentum. Boring — which is the point.
Institutionally driven (utility characteristic).
Add Verdict
Hold / Add for ballast only. NEE belongs in a concentrated AI portfolio as a hedge against AI capex digestion — when NVDA/MU/TSM get hit, utilities hold. The AI power narrative is real (hyperscalers need 24/7 reliable power away from grid delays) but NEE's share of that wallet is still small and unproven at scale. Don't overpay: the sector premium at 22× is pricing in perfect execution.
Entry zone: $85–92 (current $93 — wait for sub-$92 entry)
⚠ Key Risks
▼- 01Interest rate sensitivity — every 50bps rate shock ≈ 10% P/E compression for regulated utilities
- 02IRA modifications under Trump — subsidy stack erosion would directly cut renewables margin
- 03FCF negative -$15.4B TTM; debt $104B — execution risk if financing costs rise materially
- 04Regulatory risk at FPL (Florida PSC) — rate case outcomes unpredictable
- 05AI offtake agreements with hyperscalers are not yet signed/disclosed at scale — narrative risk if they don't materialise
💬 Analyst Quotes (X)
▼@aleabitoreddit, @bethkindig, @SemiAnalysis — No coverage from tracked analysts on X for NEE. Utility is outside all three analysts' domains.
NOW
ServiceNow, Inc.
AI Workflow Software
~$93
Post-5:1 split (Dec 18, 2025) · Mkt Cap ~$192B
Down ~40% from split-day; down ~30% YTD
✦ STRONG ADD — MOST COMPELLING RERATING
Entry zone: $85–100
Forward P/E compressed from 95-110× historical to 21-25× — same 22% revenue growth, accelerating AI ACV. Iran war FX headwind caused non-fundamental 14-17% post-print selloff. Institutional PT consensus $182 (105% upside from $93).
Fwd P/E
21–25×
vs 95-110× hist avg
PEG
0.84
Q1 Rev Growth
+22%
Sub $3.67B, +22%
FCF Margin
44%
Q1 2026
Rule of 40
57
22% + 35% FCF
cRPO
$12.64B
+22.5% YoY
Consensus PT
$182
105% upside
Valuation vs Growth
A+
Fwd P/E 21-25× vs 3Y/5Y avg 111×/180×. Deepest discount to own history in NOW's 14-year public life. Same 22% growth rate, same quality business, same FCF generation — stock just fell 40%. PEG 0.84. Only 18% above peer software median. 2030 subscription target $30-32B (from $15.7B 2026).
Most compelling rerating opportunity in the universe.
Fundamentals vs Hype
A
Rule of 40 = 57 (22% growth + 35% FCF). Now Assist ACV target raised $1B → $1.5B mid-year. Customers spending $1M+ on Now Assist grew 130%+ YoY. RPO $27.7B (+25%). Beat Q1 revenue and sub guide. Stock fell on Iran FX headwind (75bps growth drag) — non-fundamental. Armis acquisition adds margin pressure near-term.
Grounded — non-fundamental derate creating opportunity.
Institutional vs Retail
B+
~88% institutional ownership. Mixed Q1 13F — some trimming after sell-off. Short interest 3.78% (elevated for large-cap software — bears are positioned). Consensus PT $182 vs spot $93. Wide PT dispersion ($85 KeyBanc → $236 Bernstein). Retail demoralized post-selloff.
Mixed flow — institutional accumulation thesis vs near-term derate.
Add Verdict
Strong Add. This is the most mathematically compelling setup in the universe right now: a 14-year compounder that has never had a bad year trading at its lowest valuation relative to own history, while its AI revenue (Now Assist) is accelerating and management raised full-year guidance on the same print that crashed the stock. The Iran war FX headwind (75bps) is the mechanism, not a structural deterioration. The business is exactly what it was when it traded at 95-110× forward earnings.
Entry zone: $85–100 (current ~$93 is within range — start building)
⚠ Key Risks
▼- 01Iran war FX headwind persistence — 75bps Q1 drag; every additional 25bps = $50M ARR impact into Q2/Q3
- 02$7.75B Armis acquisition — margin dilution and integration risk; gross margin already down from 79% to 75%
- 03DOGE federal budget pressure — government revenue segment facing headwinds; unclear magnitude
- 04AI agent disintermediation (UBS Neutral thesis) — if AI agents bypass workflow platforms, NOW's $30B 2030 target is at risk
- 05cRPO deceleration below 20%: if Q2 subscription growth prints below 19% cc, sentiment turns negative again
💬 Analyst Quotes (X)
▼@aleabitoreddit, @bethkindig, @SemiAnalysis — No coverage from tracked analysts on X for NOW. ServiceNow is outside the hardware/semiconductor focus of all three analysts.
PLTR
Palantir Technologies
AI Software Platform
~$144
May 4, 2026 close · Mkt Cap ~$350B
52W: $40 – $150
✗ HOLD ONLY — DO NOT ADD
Reload only at $100–115
📌 Current holding
Rule of 40 = 145% (71% growth + 57% FCF margin) — best in software universe. AIP platform creating government and commercial moat. But forward P/E 100-110× is the most stretched in the S&P 500 at current price.
Fwd P/E
~110×
Most expensive S&P 500
Rule of 40
145%
Best in software
Q1 Rev Growth
+85%*
*see data conflict flag
US Commercial
+133%
FCF Margin
57%
FY26 Guide
$7.65B
+71% YoY raised
Insider Selling
$433M
90 days; Thiel $580M
Valuation vs Growth
D
Fwd P/E ~110× is 6× the software sector median (18×). Growth is real (85% total rev, 133% US commercial) but even at Rule of 40 = 145, paying 110× forward earnings requires 25%+ EPS CAGR for a decade to justify. PEG ~1.5 on generous growth assumptions. Most expensive name in the S&P 500 software complex.
Priced to perfection. Zero margin of safety at current valuation.
Fundamentals vs Hype
A+
Rule of 40 = 145 is genuinely the best in software. $8B cash, no debt, 53% GAAP net margin (extraordinary). AIP boot camps converting real enterprise deals (615 US commercial customers +42% QoQ). US government revenue +84%. GAAP profitable 4+ consecutive quarters. FCF $925M in Q1 alone.
Grounded fundamentally — but valuation embeds 25%+ CAGR for a decade.
Institutional vs Retail
C
Only 56.22% institutional — unusually low for S&P 500; reflects active manager skepticism at this valuation. JPMorgan reduced 32%, T. Rowe Price reduced 24% Q1 2026. Short interest ~2.5% + Burry put position. Thiel sold $580M in 90 days. Retail "Extremely Bullish" Stocktwits. Michael Burry still short.
Retail momentum-driven. Smart money is selling into strength.
Add Verdict (Portfolio note: already held)
HOLD only — DO NOT ADD at $144. The combination of: (1) Burry still short with June 2027 puts and Dec 2026 $100 puts; (2) Peter Thiel selling $580M in 90 days; (3) Jefferies "Underweight" with $70 PT; (4) 110× forward P/E with institutional ownership at 56% (low) — means the asymmetry is against adding. The business is extraordinary. The valuation assumes perfection. Hold what you have. If it corrects to $100-115, that is the reload zone.
Hold zone: current position intact · Reload only at $100–115 on 25-30% drawdown
⚠ Key Risks
▼- 01Burry thesis materialises — Scion holds June 2027 puts + Dec 2026 $100 puts: "I am shorting the entire premise upon which the company rests" (Burry Substack, Apr 10, 2026)
- 02Peter Thiel insider sales $580M in 90 days — one of the most informed holders in technology signalling at scale
- 03Jefferies "Underweight" $70 PT (Apr 30, 2026) — implies 51% downside; not a fringe view
- 04DOGE / federal budget volatility — government revenue +84%, but budget freeze or contract cancellation would remove largest single growth driver
- 05US Commercial deceleration risk — 133% cannot compound; when the deceleration shows (likely Q3/Q4 2026), multiple contracts sharply
💬 Analyst Quotes (X)
▼Michael Burry (Scion Capital), Apr 10, 2026 Substack: "I am shorting the business model. I am shorting the entire premise upon which the company rests. I am shorting the CEO." Holds June 2027 $50 puts and December 2026 $100 puts on PLTR.
Michael Burry / Scion Capital · Substack Apr 10, 2026
@aleabitoreddit, @bethkindig, @SemiAnalysis — No direct PLTR-specific quotes from tracked analysts verifiable from open-access search.
MU
Micron Technology, Inc.
HBM Memory — AI Critical Path
$757.35
May 8, 2026 · Mkt Cap $729B · +38% on week (best week since 2008)
52W: $84.68 – $759.50
⏸ HOLD — ADD ONLY $600–650
Add only on 20%+ pullback
📌 Current holding
Only US-based DRAM/NAND manufacturer. FQ2 FY26: $23.86B revenue +196% YoY; 74.9% gross margin (record). HBM sold out at locked prices through 2026. FQ3 guide $33.5B at ~81% gross margin. S&P 100 member March 2026. Stock +800% in 12 months.
Fwd P/E (post-rally)
~22×
⚠ was 7.5× at $400
PEG
<0.5
Still cheap on growth
FQ2 Rev Growth
+196%
YoY
FQ2 Gross Margin
74.9%
Record
FQ3 Guide
$33.5B
~81% GM
FQ2 EPS Beat
$12.20
vs $9.19 (+33%)
Rule of 40 proxy
225+
Valuation vs Growth
A
Fwd P/E ~22× at $757 (was 7.5× at $400 — valuation has normalised rapidly as price ran). PEG still <0.5. Morningstar independent FV $933. Revenue +196% YoY; FQ3 guided +200%+. EPS beat consensus by 33%. 4 consecutive beats. S&P 100 inclusion March 2026 drives passive bid.
Optically cheap but cycle-extended. Best valuation/growth in memory; stock has caught up.
Fundamentals vs Hype
A
Gross margin 30% → 56% → 74.9% in 4 quarters — fastest GM expansion in industry history. FCF $6.9B Q1. HBM locked pricing through 2026. Capex raised to >$25B. Beth Kindig: "#1 AI chip pick for 2026." Counter: SemiAnalysis reports MU excluded from Vera Rubin Year-1 HBM4 builds — Samsung/SK Hynix dominate.
Grounded but cycle-extended. HBM4 Vera Rubin position = the swing factor for 2027.
Institutional vs Retail
A+
80%+ institutional. S&P 100 inclusion (March 2026) = forced passive buying. 2,950 13F filers. Top holders: Vanguard, BlackRock, Capital World. Short interest 3.22% (meaningful for context). Retail "Extremely Bullish" with parabolic momentum. D.A. Davidson initiated Buy late April 2026.
Institutionally driven with retail momentum overlay (parabolic week of May 8).
Add Verdict (Portfolio note: already held)
HOLD — do not add at $757. The thesis is intact: HBM supercycle, locked pricing, 74.9% gross margin, FQ3 guide at $33.5B. But the stock is up 800% in 12 months and just had its best week since 2008. Adding here is chasing a parabola. The risk/reward asymmetry is poor: upside to $900-1000 if cycle holds; downside to $450-550 if HBM4 Vera Rubin design win fails or cycle turns. Better to add the thesis at $600-650 where the margin of safety returns.
Hold current position · Add only on dip to $600–650 (currently $757 — no add)
⚠ Key Risks
▼- 01HBM4 Vera Rubin Year-1 design loss — SemiAnalysis (Jan 28): Samsung/SK Hynix took the slots; Micron may wait for HBM4E or 2027 refresh cycle
- 02Memory cyclicality — FQ2 gross margin 74.9% is extraordinary and unlikely to be the new normal; Samsung/SK Hynix capacity additions = pricing pressure H2 2027
- 03Parabolic technical setup — +800% in 12 months invites 30%+ drawdown on a single catalyst miss; no price target support at $757
- 04Capex >$25B FY26 — execution risk on HBM4 redesign and yield ramp if Vera Rubin design win delayed
- 05CXMT (China domestic DRAM) secular pressure — long-term market share erosion in legacy DRAM as China builds domestic capacity
💬 Analyst Quotes (X)
▼"MU stock is up 180% YTD – or 4X more YTD than AI heavyweight Nvidia. Recently, the high bandwidth memory, or HBM, content that Micron supplies has increased 3.5X." Listed MU as #1 AI chip pick for 2026.
@bethkindig (Beth Kindig / IO Fund) · Seeking Alpha / Medium, Nov–Dec 2025
"Micron is in active discussions with customers for HBM4 volumes and expects to sell out of capacity for 2026 over the next few months."
@bethkindig (Beth Kindig / IO Fund) · Medium article on MU, Dec 2025
"Nvidia will be leaving out Micron HBM4 and primarily using SK Hynix and Samsung HBM4 for first year Rubin builds… latest HBM4 engineering samples from memory suppliers remain below target specifications in terms of pin speed."
@SemiAnalysis (Dylan Patel) · Jan 28, 2026 Institutional Report (open-access summary)
@aleabitoreddit — No direct MU quote verifiable from open-access search for this period