JP Invests

1.1K posts

JP Invests banner
JP Invests

JP Invests

@JP_Invests

Follow for stock analysis, set up & framework | Concentrated growth + options | Trade Ideas | 7-Figure Self-Made Investor | No hype. Just the work.

Markets · Macro · Money Katılım Nisan 2026
115 Takip Edilen181 Takipçiler
Sabitlenmiş Tweet
JP Invests
JP Invests@JP_Invests·
I've been investing for ~10 years. Five years of compounding through stocks and options, the portfolio crossed seven figures. How I got here: an actual framework — concentrated bets on high-conviction growth, options used to amplify and generate income, position sizing tuned for asymmetric upside. The framework wasn't free. The first 3 years cost me real money chasing the wrong setups, sizing positions wrong, and holding losers too long. The losses paid the tuition. What you'll get here: the work — what I'm researching and why, the framework — how the math actually works, the lessons — what the years that cost me taught me. If considered concentrated investing is your game, follow me.
English
1
0
2
1.3K
JP Invests
JP Invests@JP_Invests·
🟢 $AEHR is up ~30% after hours. The number that did it isn't the beat. Q4 revenue was $18.8M against $18.69M consensus — up 33% from $14.1M a year ago. EPS came in at +$0.11 versus an expected loss of a penny. Good print. Not a 30% print. The 30% is fiscal 2027 guidance: $130M–$150M. That's 160–200% growth off a $50M base. Bulls going into the day were arguing over whether $70M was achievable. Management doubled the number. The non-obvious read: this stopped being a guidance story the moment they attached a backlog to it. Effective backlog is $100.6M. Record Q4 bookings of $60.7M. The low end of FY27 guidance is roughly 77% covered by orders already on the books. Most companies that guide 200% growth are selling a pipeline. Aehr is selling a contract. Second thing worth noticing: cash went from $37.1M in February to $116.5M in May. They're funding the ramp without diluting into it. 🐻 The bear case, which is real: FY26 revenue actually declined — $50.0M versus $59.0M in FY25. Adjusted net income fell to $0.9M from $4.6M. The stock had already dropped ~33% in a month before this print, and it's up 380%+ over a year. Concentration risk is the structural one — "lead hyperscale customer" and "lead silicon photonics customer" carry this whole story. Lose one and the backlog math changes overnight. My read: the guide is credible because the backlog underwrites it. That's rare enough to take seriously. What I'm watching is conversion — backlog to recognized revenue in Q1, not more order press releases. Orders were never the question. Revenue was. $AEHR
English
0
0
0
21
JP Invests
JP Invests@JP_Invests·
$TSEM up ~19% pre-market (to ~$272). The catalyst: a $3B Japan expansion — with ~$1B in grants from Japan's government (METI) — to scale 300mm silicon photonics, SiGe & advanced packaging for AI optical interconnect. This is a sovereign-backed AI infra story now. 🧵 The number that moved it: Tower RAISED its 2028 targets to $3.6B revenue / $1.2B net profit — up from $2.8B / $750M. That's ~+29% revenue and ~+60% profit vs the prior plan. First phase: Arai fab converts to 300mm silicon photonics, full ops by Q4'27. Even after the pop it's ~40x forward earnings, and shares are already +89% YTD. The optical-interconnect TAM is real, but 2028 is far off and $3B of capex is a big swing. Great catalyst, rich price. Own the story, respect the valuation.
English
0
0
0
20
JP Invests
JP Invests@JP_Invests·
🟡 June CPI came in soft across the board. Here's what the print actually says. Headline CPI fell 0.4% month-over-month — the sharpest single-month drop since April 2020. Consensus was looking for -0.1%. Core, ex-food and energy, printed flat versus the +0.3% Street estimate. Year-over-year, headline landed at 3.5% (vs. 3.8% expected) and core at 2.6% (vs. 2.9%). Both decelerated from May. The asymmetric read: this is the kind of disinflation print that pulls forward rate-cut odds. A flat core month after May's +0.2% isn't noise — it's the sequential trend the Fed has been waiting to confirm before it moves. If the next print rhymes, the cut math changes. What this favors: long-duration growth. Concentrated positions in high-multiple compounders re-rate hardest when the discount rate falls. The names carrying the most rate sensitivity are the ones that catch the biggest bid on a dovish repricing. $SPY $QQQ
English
0
0
0
22
JP Invests
JP Invests@JP_Invests·
$MYRG : The Quiet Electrician Behind the Data-Center Boom Some of the best AI-infrastructure exposure isn't a chip stock. Before a single GPU gets racked, someone has to build the power. $MYRG is the listed contractor doing exactly that — wiring substations, transmission lines, and the data centers themselves. The business is firing. Q1 revenue crossed $1B for the first time (+20%), and earnings blew past estimates by ~44% — the fourth straight quarter of accelerating beats. Better still, margins are expanding, not just revenue. The backlog just hit a record $2.84B, giving real visibility into 2027–2028. Data centers are only ~6% of revenue today but set to grow 30%+ next year. The overlooked kicker: management is buying into the boom, not just riding it. The $328M Valley/Comet deal (closing ~July 1) plants $MYRG deeper into the Western markets — Pacific Northwest, Southern California — where demand runs hottest. Here's the catch. The stock has roughly doubled in a year and now trades ~33x forward earnings for a business that nets under 4 cents on the dollar. Four insiders — including the CFO — sold ~$10.5M into June's strength, with zero buys to offset. That's a lot of super-cycle already in the price. The verdict: elite business, durable tailwind, stretched quote. This is a hold-and-accumulate-on-weakness name, not a chase-it-here one.
English
0
0
0
36
JP Invests
JP Invests@JP_Invests·
$PENG — the cheapest forward P/E in AI infrastructure, attached to the loudest red flag. Just opened a position at $41.36. This is either a setup the market hasn't priced or a value trap dressed in AI clothing. Both views are defensible. The math says the asymmetry favors the first. 📊 The setup at $41-44: - Forward P/E: 16x vs semi industry average 23x - Just doubled FY26 sales growth guidance: 6% → 12% - Just raised FY26 EPS guide: $0.85 → $1.30 (+53%) - OriginAI memory appliances for Nvidia B300 GPUs — real product, not a slide - Real AI factory deployments through Penguin Computing + Stratus brands 🚨 The honest bear case: - Trailing P/E of 63,185x — TTM earnings are essentially zero - Trading 38% above average analyst PT of $27.88 - Q2 revenues were DOWN year-over-year — the guide raise is forward-looking, not confirmed - The previous name (SMART Global Holdings) was a low-quality memory distributor - Memory-heavy revenue mix is structurally volatile The asymmetric math: If FY26 EPS hits the new $1.30 guide, this stock trades at 34x current earnings on a name growing 12% topline with real AI infrastructure exposure. Reasonable. If guidance disappoints, the stock probably gives back 30-40%. What the bulls see: forward P/E of 16x is the cheapest AI infrastructure multiple on the board. Cheaper than $ICHR (forward 32x), cheaper than $AMKR (forward 35x). If the new guide holds, the rerating is mechanical. What the bears see: a company that just doubled its own guidance two quarters into the fiscal year. That's either a real inflection or aggressive marketing. The next two quarters decide which. Long $PENG at $41.36. The cleanest forward multiple on this basket. The least proof of execution. Sized accordingly. $PENG
English
2
1
7
1.1K
Kevin Xu
Kevin Xu@kevinxu·
can someone smarter than me pls explain why $PENG is at breakeven after a double earnings beat and guidance raise?
Kevin Xu tweet media
English
105
1
133
65K
Kevin Xu
Kevin Xu@kevinxu·
NEW ALL IN*: $PENG @ $68.22. EARNINGS PLAY FOR TOMORROW. 60% CHANCE OF BEAT IMO. HIGH RISK - 26.3% IMPLIED MOVE. They help companies build their own "AI factory" which is just a fancy way of saying your own ~neocloud. - Validated with fancy partner awards from Nvidia, Dell, etc. and price upgrades ($75 PT from Rosenblatt today). - Exploded after last earnings in April and continually making higher lows. - Memory business grew 63% last quarter. $109.2M of cash used for inventory in Q2. They've been smartly stockpiling memory to meet demand. Now earnings are always a coin flip but the macro feels good and tailwinds on our side right now. They also hinted at guiding towards the higher end back on June 1. > "expects both net sales and diluted EPS for FY2026 to be at the high end of its previously issued outlook ranges" Source: ir.penguinsolutions.com/news/news-deta… Now think about all the developments in AI since June 1st. Given the insane demand for AI infra (specifically the local model meta) coming in H2, $PENG seems like a winner. --- *As a reminder: this is my challenge account where I restarted with $35k to go all-in swing trading 1 stock at a time to $10M again. Realtime alerts to subscribers. Summaries after market close to non-subscribers. As always, please do your own research with your own independent thinking and risk tolerance and decide your own buys and sells. I may trade on a whims notice.
Kevin Xu tweet media
English
108
16
289
380K
CK Capital
CK Capital@CKCapitalxx·
WOW $PENG. Record revenue of $479 million, up 48% year over year. The street was looking for around $405 million. They cleared it by roughly $70 million. Non-GAAP EPS of $0.84, up 79%. GAAP operating income up 417%. And they raised full-year guidance on both sales and EPS. Again. Here’s the part that still isn’t in the numbers. This is the AI Factory platform company. Every enterprise, neocloud, and sovereign AI buildout is leaning on their integrated memory and infrastructure, and that demand is only speeding up. They didn’t just beat. They raised the floor. Still one of our highest conviction name.
CK Capital tweet media
English
40
19
355
51.1K
Sam Badawi
Sam Badawi@Sam_Badawi·
Penguin Solutions - Q3 Earnings $PENG 70.61 [+12.21% AH] ✅ Revenue: $478.7M (Est: $400.7M) ✅ Adjusted EPS: $0.84 (Est: $0.54)
Sam Badawi tweet media
English
19
4
49
46.4K
Shay Boloor
Shay Boloor@StockSavvyShay·
$PENG Q3 EARNINGS • Revenue $479M vs Est. $421M • EPS $0.84 vs Est. $0.56 • Integrated Memory: $275M (+111% YoY) FY27 Guidance • Revenue 22% vs Est. 16% • EPS $2.60 vs Est. $2.28 Became an $NVDA AI Factory specialized partner.
Shay Boloor tweet media
English
57
27
348
70.7K
Wall St Engine
Wall St Engine@wallstengine·
$PENG | Penguin Solutions Q3 FY26 Earnings Highlights 🔷 Revenue: $478.7M vs $421.4M Est. 🟢 🔷 Adj EPS: $0.84 vs $0.56 Est. 🟢 🔷 Record quarterly net sales, up 48% YoY. FY26 Outlook Raised: 🔷 Non-GAAP EPS: $2.60 ±$0.05 vs $2.28 Est. 🟢 🔷 Net sales growth now seen at 22% ±2%, up from 12% ±5% 🔷 Non-GAAP gross margin: 28.5% ±0.5% Segment Revenue: 🔷 Advanced Computing: $137.6M, up 4% YoY 🔷 Integrated Memory: $275.1M, up 111% YoY 🔷 Optimized LED: $66.1M, up 7% YoY Business Highlights: 🔷 Integrated Memory net sales more than doubled YoY 🔷 AI Infrastructure added 4 new customer logos in Q3 🔷 Became an NVIDIA AI Factory Specialized Partner 🔷 Expanded ClusterWareAI with an AI Factory Operations Agent
Wall St Engine tweet media
English
16
15
198
56.5K
JP Invests
JP Invests@JP_Invests·
🟢 $PENG confirmed the thesis I opened on at $41.36. Stock's at $79 — a fresh 52-week high, up ~90% from my entry in less than 2 months. Q3 print: $0.84 adj EPS vs $0.56 expected. Record revenue $479M, up 48% YoY vs $421M expected. FY26 EPS guide raised to $2.60 from $2.15. AI-driven revenue now 74% of sales, up 104% YoY. What changes now: the 16x forward multiple that made this asymmetric is gone. The re-rating I was waiting on happened. Analyst PTs walked up to $75-85. The edge was the entry, not the price here. My read: the thesis played out and I'm holding — the preliminary FY27 guide keeps the growth intact. Found the setup when it was cheap and hated. That was the work. Follow for more value trades. $PENG $AI
JP Invests@JP_Invests

$PENG — the cheapest forward P/E in AI infrastructure, attached to the loudest red flag. Just opened a position at $41.36. This is either a setup the market hasn't priced or a value trap dressed in AI clothing. Both views are defensible. The math says the asymmetry favors the first. 📊 The setup at $41-44: - Forward P/E: 16x vs semi industry average 23x - Just doubled FY26 sales growth guidance: 6% → 12% - Just raised FY26 EPS guide: $0.85 → $1.30 (+53%) - OriginAI memory appliances for Nvidia B300 GPUs — real product, not a slide - Real AI factory deployments through Penguin Computing + Stratus brands 🚨 The honest bear case: - Trailing P/E of 63,185x — TTM earnings are essentially zero - Trading 38% above average analyst PT of $27.88 - Q2 revenues were DOWN year-over-year — the guide raise is forward-looking, not confirmed - The previous name (SMART Global Holdings) was a low-quality memory distributor - Memory-heavy revenue mix is structurally volatile The asymmetric math: If FY26 EPS hits the new $1.30 guide, this stock trades at 34x current earnings on a name growing 12% topline with real AI infrastructure exposure. Reasonable. If guidance disappoints, the stock probably gives back 30-40%. What the bulls see: forward P/E of 16x is the cheapest AI infrastructure multiple on the board. Cheaper than $ICHR (forward 32x), cheaper than $AMKR (forward 35x). If the new guide holds, the rerating is mechanical. What the bears see: a company that just doubled its own guidance two quarters into the fiscal year. That's either a real inflection or aggressive marketing. The next two quarters decide which. Long $PENG at $41.36. The cleanest forward multiple on this basket. The least proof of execution. Sized accordingly. $PENG

English
0
0
0
210
JP Invests
JP Invests@JP_Invests·
Verdict: own all five, size to the risk. Highest conviction $MU . The fear that made these cheap — a supply glut — is the debate you're paid to take the other side of. If hyperscaler capex holds near the ~$667B guided, this dip reads as an entry, not a top.
English
0
0
0
15
JP Invests
JP Invests@JP_Invests·
The risk, stated plainly: a memory supply glut. Samsung signaling more NAND capacity is exactly what threatens $SNDK 's pricing and the fat margins that make these cheap. That's the real debate — size the cyclical names smaller and own it with eyes open.
English
1
0
0
18
JP Invests
JP Invests@JP_Invests·
the AI-infra trade cracked. Micron −6.5%, Sandisk −9%, Credo −9.3%, Vicor −12.4%. Memory entered a bear market, >20% off highs. BofA calls it "a bear trap" ahead of a $1.5T buildout. Five names went on sale together — here's which to own ↓
English
1
0
1
91