Moose | Long-Term Investor retweeted
I invest in Archer.
A sector where Archer is the only representative.
That’s the thesis.
$ACHR
Moose | Long-Term Investor retweeted
Dopamine from information gathering is a dangerous drug.
It’s the dopamine you get from reading. From planning. From learning. But never doing.
Information gathering as sport. Jammed into neat little dopamine feedback loops that convince you that you're doing something productive and valuable.
You're just one piece of information away from the big breakthrough. You'll start when your business plan is perfected. You'll meet your partner when you've scrolled through another round of profiles. You'll get that dream job when you have one more degree in hand.
Your entire life will change when you realize the world isn't run by a bunch of geniuses. The world is being run by a bunch of normal people with abnormal bias for action.
The answer you seek is hidden in the action you avoid.
Dopamine from information gathering is a dangerous drug.
Get your dopamine from action.
Moose | Long-Term Investor retweeted
BREAKING $AMD crushes $ARM $INTC CPUs 🚨
@AMD new EPYC “Venice” processors are steamrolling rivals in the agentic AI era, delivering up to 2.8× more agents per watt, 1.8× more agents per dollar, and over 2× more agents per rack than Intel Xeon, ARM, and even NVIDIA Vera designs. With industry-leading core density (up to 256 cores/512 threads), PCIe Gen 6, and unmatched efficiency, AMD is rewriting the rules of data center compute proving CPUs, not just GPUs, are the real kings of the AI revolution.
Enterprise Agentic AI differs fundamentally from the earlier chatbot/generative AI era. A single user query can spawn many concurrent agents that plan, act, use tools/scripts, communicate, and iterate. This creates heavy demand for CPU resources beyond traditional GPU inference. AMD argues that the industry underestimated the CPU needs for this shift. Data centers now require a mix of server types rather than a one size fits all approach:
~General-purpose CPU servers: Efficient performance across broad workloads, good power/performance balance, and strong confidential compute/security features. These support longer server lifetimes and high consolidation ratios (example cited: roughly 12:1 replacement of older Xeon-class servers).
~GPU / AI host-node servers: CPUs optimized to keep GPUs fully utilized: strong per core performance, high frequency, high CPU-GPU bandwidth (more memory channels, PCIe Gen 6).
~New “agentic CPU server sandboxes”: High core/thread density servers dedicated to running concurrent agents, tool execution, scripts, inter-agent communication, and related orchestration. Agents can map roughly to threads, so the key metrics become agents per watt, per dollar, and per rack.
Source: Building Agentic AI with AMD EPYC CPUs
youtube.com/watch?v=GCA7gOxG…
$AMD| Q2 2026 10-Q Full đź§µ
TLDR: Data center operating income is now the biggest driver of EPS and is the fastest growing segment.
The Q2 2026 10-Q largely aligns with the August 4 press release on headline results but adds substantial detail on balance sheet/cash flow dynamics, discontinued operations (ZT Systems), large customer warrants, export control nuances, long-term commitments, liquidity facilities.
1. Q2 ER:
~Revenue: $11.536 billion (+50% YoY from $7.685 billion; +13% QoQ from $10.253 billion).
~GAAP: Gross margin 54% (vs. 40% YoY), operating income $1.990 billion (vs. loss of $134 million), net income $2.297 billion, diluted EPS $1.38.
~Non-GAAP: Gross margin 56%, operating income ~$3.1 billion, net income ~$2.76 billion, diluted EPS $1.66.
~Segments (Q2): Data Center $6.718 billion (+107% YoY), Client and Gaming $3.841 billion (+6% YoY; Client $3.062 billion +23%, Gaming $779 million –31%), Embedded $977 million (+19% YoY). Data Center was 58% of total revenue.
~Q3 guidance (press release): Revenue ~$13 billion ± $300 million (~41% YoY / ~13% sequential at midpoint); non-GAAP gross margin ~56%.
~Key drivers cited in both: Strong EPYC and Instinct (MI350 series) demand; favorable mix; absence of the prior-year inventory/related charges tied to U.S. export controls on MI308 products (press release notes ~$800 million impact in Q2 2025).
2. Other disclosures:
~Balance sheet (June 27, 2026 vs. Dec. 27, 2025): Cash + short-term investments $13.1 billion (up from ~$10.6 billion). Inventories $8.468 billion (up; heavy work-in-process ~$5.265 billion). Accounts receivable $7.281 billion. Total assets ~$84.5 billion. Long-term debt principal ~$3.25–3.3 billion (current portion $875 million on the 4.212% notes due 2026). Stockholders’ equity ~$67.2 billion.
~Cash flows (six months): Operating cash flow from continuing operations $5.321 billion (strong, driven by net income plus non-cash items, partially offset by working-capital uses including higher receivables and supply-agreement prepayments). Investing uses ~$5.2 billion (short-term investment purchases, ~$1.2 billion PPE, long-term investments, small acquisitions). Financing uses ~$365 million (stock repurchases $221 million + tax withholdings, partially offset by employee equity proceeds).
~Other income (expense), net: $598 million in Q2 (large unrealized gains from public listings of previously non-marketable equity securities).
~International sales: ~70% of Q2 revenue.
~Remaining performance obligations: $222 million.
~R&D $2.528 billion (+33% YoY, higher headcount for AI); MG&A $1.401 billion (+41%, go-to-market). Amortization of acquisition-related intangibles remained material (~$544 million in Q2 in the “All Other” category).
ZT Systems
Acquired March 2025 for ~$4.4 billion (cash + stock + contingent consideration) to support end-to-end AI infrastructure. Manufacturing business sold to Sanmina in October 2025 for $2.4 billion cash (subject to adjustments) + Sanmina stock + earn-out (up to ~$450 million through 2028).
Q2 2026 included a $243 million payment for post-closing net working capital adjustments. Earn-out receivable ~$348 million remains on the balance sheet (subject to impairment testing). Discontinued operations produced a small loss (~$8 million net of tax) in Q2 2026. A manufacturing services agreement continues with Sanmina.
Export China related details:
~MI308: New U.S. license requirements (April 2025) for shipments to China/D5 countries or related entities. Some licenses granted; limited shipping began late 2025. ~$440 million net inventory/related charges recorded in 2025 (press release highlighted the larger Q2 2025 figure). Potential (but not yet regulated) 15% revenue share to the U.S. government on licensed MI308 sales.
~MI325: Licenses granted February 2026 for certain China-based customers, but products must undergo U.S. inspection and face a 25% tariff. Unclear whether China will allow imports.
Capital allocation:
~Revolving credit facility expanded to $5.0 billion (May 2026, replacing prior $3 billion); commercial paper program to $5.5 billion. None drawn.
~Unconditional purchase commitments ~$30.3 billion ($17.4 billion remaining in 2026), mainly wafers, components, cloud capacity, and licenses.
~Data-center lease guarantees: maximum exposure $4.1 billion. Additional uncommenced leases ~$4.5 billion + subsequent ~$9.5 billion (commencing 2027–2028, terms up to 16 years).
~Subsequent equity/investment commitments up to $5.0 billion through fiscal 2028.
~Share repurchase authorization $14 billion with ~$9.2 billion remaining; $221 million executed in the first half. Equity incentive plan amended to add 65 million shares. Several executives adopted/modified Rule 10b5-1 plans (Dr. Lisa Su for 450,000 shares).
3. Verdict
AMD delivered a Triple Beat or clean beat and raise quarter with record revenue and sharp margin recovery. Data Center more than doubled, becoming the clear majority of the business, fueled by EPYC CPU strength (agentic AI driving higher CPU:GPU ratios) and Instinct GPU ramp. Gross margin expansion reflects both mix and the absence of the prior export-control inventory hit.
Operating leverage is visible despite elevated R&D and go to market spend for AI. Cash generation is robust, the balance sheet is liquid, and large multi-year GPU commitments (OpenAI, Meta, Anthropic, Microsoft...) plus Helios deployments provide multi-year visibility into the AI accelerator opportunity. Embedded returned to growth and Client held up well on Ryzen share gains despite a mixed PC backdrop. Management’s Q3 guide implies continued sequential acceleration, consistent with comments that Data Center growth will accelerate in the second half, and much more significant in Q4.
Current estimate for FY2026 Non-GAAP EPS:
$9.6 - $11.53 at $489/share would imply
42-51x FY2026 P/E
17-25x FY2027 P/E
Not Financial Advice! DYOR!
Source: ir.amd.com/financial-informa…
Moose | Long-Term Investor retweeted
$AMD CEO:
Excited to expand our partnership with @Azure to deploy @AMD Helios with our newest MI455X GPUs, Venice CPUs, and Pensando networking. Thanks @satyanadella @scottgu for the continued partnership. Proud of what our teams are building together!
Moose | Long-Term Investor retweeted
This post triggered many! Yes my bear case of $NOW for 2026 fiscal year end is $150 per share!
Frankly I don't care what the price of $NOW is today, next week or 6 months from now. These are my year-end valuation targets 2026–2029 (Bull/Base/Bear) based on my valuation models.
• Valuation based on forward P/E multiples applied to consensus EPS (2026: $4.20 → 2027: $5.05 → 2028: $6.15 → 2029: $7.20).
• Bull case: 63x P/E is betting on accelerated AI-native adoption and Context Engine leadership premium
• Base case: 49x P/E is beating on steady high-teens growth and margin expansion
• Bear case: 36x P/E with potential headwinds from slower monetization or macro factors
$NOW ’s enterprise AI orchestrator role supports these valuation scenarios long-term. NFA. DYOR.
Moose | Long-Term Investor retweeted
The ultimate trifecta trade I planned for 18 months ago. $MU $AMD $NOW
This is what I do! I understand the product, the business, fundamentals, valuation and THEN (and only then) click BUY!
My portfolio is still up on the day despite the significant SaaS drawdown this week. If AI is a bubble, my positioning is straightforward:
• Bought $AMD and $MU when they were widely dismissed in 2024/early 2025; the hardware leaders building AI inference infrastructure and direct beneficiaries of massive AI capex spend.
• Hold one high-conviction SaaS stock that will benefit regardless of whether the bubble inflates or pops: $NOW
$NOW remains the essential orchestrator for enterprise AI agents, enforcing policy, regulation, and governance at scale. Only started building a position in Feb 2026.
P.S I sold out completely of $MU earlier in 2026. Reasons mentioned in previous posts.
Moose | Long-Term Investor retweeted
🛑No one will listen and then FOMO!
The real edge comes from truly understanding the business: that’s what tells you whether it’s genuinely differentiated SaaS or just riding the wave. $NOW is here to stay. $NOW is a $350 stock in 3 years.
Why ServiceNow $NOW is here to stay, and why I see it as a solid investment: It's all about the growing need for AI governance and control. In our world, full of regulations from governments and businesses alike, AI can't just go rogue. Software has to build in rules, compliance, and oversight right from the start. ServiceNow really gets this right.
In our heavily regulated world, AI has to follow the rules. Take things like the EU AI Act, ISO 42001, or NIST frameworks... the fines for slipping up can be huge. Businesses need ways to keep an eye on risks, stick to policies, and audit everything. That's where ServiceNow's AI Control Tower shines: It's a central spot for seeing all your AI models and agents, handling automated workflows, managing risks in real time, and weaving in compliance checks.
Even Palantir $PLTR understands this well. Their AIP platform weaves in ethics and governance from the beginning: Things like data governance, risk management for high-stakes AI, audit trails, ways to reduce bias, and georestrictions to stay compliant with different regions. It even supports EU AI Act requirements straight out of the box. If a company like Palantir, with its focus on defense, makes this a priority, it's clear this is essential for any enterprise AI to thrive.
What gives ServiceNow $NOW an edge? It's basically the go-to system for managing workflows, now boosted with strong governance features. No more hidden AI messes... it handles the full lifecycle from start to finish. Even Microsoft chose it to manage their own AI agents. As regulations keep getting stricter around the world, the demand for this is only going up. And yeah, the valuation makes sense for that kind of long-term strength.
AI without proper safeguards is just asking for trouble. ServiceNow turns compliance into something that actually helps you win. I'm in for the long haul.
Moose | Long-Term Investor retweeted
$AMD CEO: MI450 demand is exceeding AMD’s original plans
"As we approach production, demand for MI450 series GPUs continues to strengthen with lead customer forecasts now exceeding our initial plans and a growing number of new customers engaging on large-scale deployments, including additional multi-gigawatt opportunities. With this expanded visibility, we have strong and increasing confidence in our ability to deliver tens of billions of dollars in annual data center AI revenue in 2027, and to exceed our long-term growth target of greater than 80% in the coming years."
$AMD CEO: "Data Center is now the primary driver of our revenue and earnings growth. As AI adoption scales, demand is increasing not only for accelerators, but also for the high-performance CPUs that power and orchestrate those workloads...data center revenue increased 57% year over year to a record $5.8 billion, led by strong demand for our EPYC CPUs and Instinct GPUs...Gross margin was 55%, up 170 basis points versus a year ago, driven by a favorable product mix, including a higher data center revenue contribution."
Moose | Long-Term Investor retweeted
Oscar Health just reported our first quarter 2026 results, announcing significant year-over-year improvements across our core metrics.
Our innovative technology, affordable products, and superior experience continue to fuel strong growth and retention.
We’re reaffirming our full year 2026 outlook across all metrics and are on track to significantly expand margins and achieve meaningful profitability in 2026.
As Oscar Health CEO Mark Bertolini notes:
“Consumers expect to shop for healthcare like everyday products – on choice, price, and value. Oscar’s exceptional technology, lifestyle products, and member experience deliver exactly that. The workforce is shifting, the individual market is resilient, and Oscar is leading the transition to a consumer-driven health economy.”
Moose | Long-Term Investor retweeted
When $AMD shareholders understood Dr. Su secured at least 30-40% TSMC 2nm allocation
$AMD has the supply to hit $12-$15 EPS đź§µ
@AMD has meaningful TSMC's 2nm allocation đź§µ
Lots of mixed information between 2nm and 3nm out there.
At $15-$18 Non-GAAP EPS from Q3 2026 to Q2 2027, AMD is trading at 19x-24x Forward P/E at $359/share.
1. Apple is reported to have >50% of 2nm capacity for Iphone 18 lineup, M6-series Mac chips. Iphone 18 total revenue is expected to be flat at best or fall modestly by 2% vs Iphone 17, so the 50% capacity is unlikely to increase much.
2. AMD EPYC Venice and MI455x are on TSMC 2nm with at least share above 30%
3. $NVDA Rubins are on 3nm
4. $AVGO $QCOM MediaTek are on 3nm as well
TSMC’s 2nm (N2) expansion is accelerating dramatically in 2026 . 2 primary 2nm fabs and scaling production to 60,000-65,000 WPM by end of 2026 or early 2027.
Read more on 5 more fabs below.
If Apple have a flat YoY on Iphone 18, then most of 5 2nm Fabs expansion are for $AMD.
Meaning, $AMD can secure significant % of EPYC Venice (CPUs) and MI455X(GPUs).
Not Financial Advice!
Moose | Long-Term Investor retweeted
One of the largest valuation disconnects I see is in $CELH, down -18% (vs NDX -3%) over the past week on news that $COST is rolling out a Kirkland Signature Sparkling Energy Drink as a private label product. The packaging looks a lot like Celsius (white can, fruit pictured on front) and the product is zero sugar, 200 mg of caffeine, and contains “7 essential vitamins” (same as Celsius). The -$1.8 billion hit to CELH market cap over the past week seems way overdone as that essentially writes down CELH’s entire Costco business, which makes up 11% CELH sales.
On a category level, private label accounts for just 6% of carbonated beverage unit sales and an even lower % for energy drinks, given consumer views that private label carbonated beverages simply don’t measure up in quality or taste to branded products. According to JP Morgan, Kirkland Sparkling Energy is priced at a significant discount in Costco ($16.99 for a 24-pack vs. $27.23 for a Celsius 18-pack). While the initial Costco rollout could get ample trial and impact next few quarter sales, we highly doubt the Kirkland Signature product will meaningfully impact long-term CELH valuation.
At a forward P/E of 21x 2026 EPS, CELH is no longer a long-duration growth stock and a near-term CELH revenue miss caused by the Costco launch is unlikely to do much damage beyond what’s already been discounted. Assuming company average growth and margin, which is likely too aggressive as club is typically lower margin, Costco could account for $85-$90M EBITDA, which at a 13-14 EV/EBITDA multiple pre-Kirkland Energy news = $1.2B equity value vs. ~$1.8B market cap reduction over the past week (Source: JP Morgan).
We would be a buyer of $CELH here at $34/share (-26% YTD vs NDX -8%).
AMD should be able to easily sell every single MI455X they can get produced. The final mosaic piece to roundhouse the Nvidia stranglehold is to get early access samples to SGLang, vLLM and other oss asap under NDA. Ampere took almost 2 years to get fully optimized. Hopper 9 monthish, Blackwell is not yet fully optimized even today for many workloads. Scaling up the slilicon to sdk to actual sw stack is where AMD can undercut NVIDIA hugely.
Moose | Long-Term Investor retweeted
When I first invested in $OSCR, I spelled it out clearly: “…through individual and ACA plans.” Not only ACA. Either way, the outcome of an Obamacare overhaul or enhancement is a win-win for $OSCR. I am adding to my position.
Just started a position in Oscar Health $OSCR at $12/share.
Oscar is a tech-driven health insurance company focused on making healthcare more accessible and affordable, especially through individual and ACA plans.
The stock looks undervalued with a Street price target of $20+, and my DCF model puts fair value closer to $37. Strong growth in membership, improving margins, and upcoming Q1 earnings could be catalysts.
Risk lies in regulatory changes + healthcare utilization trends, but I like the upside here.
Moose | Long-Term Investor retweeted
$OSCR I’m invested not just because they’re set to disrupt US health insurance, but because of this CEO, he clearly knows his stuff!
Moose | Long-Term Investor retweeted
Replying to @eddie_beach3697
Margin collapse applies to undifferentiated SaaS. $NOW runs mission-critical workflows with high switching costs. AI gets monetized inside the platform, not against it. That’s pricing power, not erosion.
Moose | Long-Term Investor retweeted
How life really feels when Amazon $AMZN stock is up 2 days in a row.
Moose | Long-Term Investor retweeted
Why I am bullish on $AMZN for the next 5 years? Two main reasons:
1. We are seeing a massive Capex wave among hyperscalers; they’re pouring billions into AI infrastructure like data centers and GPUs. I believe depreciation is about to slam their EPS hard. These assets depreciate over 3-5 years; turning that upfront spend into ongoing expenses that erode profits.
For companies like $GOOGL, $META, and $MSFT; who’ve never been asset-heavy; this will unravel fast. Their high margins (think 30%+) have been built on software/services; not physical assets. Suddenly they’ll face ballooning depreciation charges; diluting EPS. We’ve seen hints in recent quarters; but the full hit is coming as Capex ramps to $100B+ annually across the board.
$AMZN on the other hand; has always been asset-intensive with warehouses; logistics; and fulfillment centers. Their margins hovered in the low single digits for a decade; so they’re conditioned to heavy depreciation. AWS aside; $AMZN’s core ops are built for this; they’ll weather the storm easier without the same margin shock.
2. I’m super bullish on $AMZN precisely because I believe they’ll be the first to demonstrate how robotics and AI slash costs and boost the bottom line. Picture this; manufacturer to warehouse to house with zero human touch. Automated picking; drone/robot delivery; AI-optimized supply chains. As peers struggle with depreciation drag; Amazon turns it into efficiency gains.
Long $AMZN the real AI winner in ops transformation. I have been buying heavy this week.
Moose | Long-Term Investor retweeted
VOLUME NEVER CHANGE:
Define your edge and verify positive expectancy with data. Backtest and track.
Plan the trade. Trade the plan. No ad‑hoc changes once you’re in.
Cut losses quickly. Small, frequent losses are tuition.
Let winners run.
Moose | Long-Term Investor retweeted
$OSCR - This will be a $100B company. We are witnessing the next $PLTR and $HOOD right before our eyes. Oscar's TAM could very like 3-4x over the next two years with the ICHRA push. There was a 171% growth of employees offered ICHRAs between 2022-2023. Just wait until it goes mainstream. It's better for the employer & the employee - it's a no brainer better wait to do healthcare. Now, legislation is very slow moving and these things take time but the path we are going on is clear. With 20% of US GDP spent on healthcare and a clear problem with Debt/GDP a solution isn't just ideal it's essential for the solvency of US debt.
Lets break a few key things down:
ICHRA = the payment mechanism (employer allowance)
ACA = the actual insurance marketplace providing the plans
Employees use the ICHRA funds to go shop for ACA marketplace plans.
Old model (traditional group insurance):
Employer → Buys group plan → Employees stuck with that plan.
New model
Employer → Gives $$ to employee → Employee → Buys an ACA individual plan
So the employee ends up in the ACA market, which is Oscar’s stronghold.
If ICHRA adoption grows, millions of employees move off group plans into ACA plans, increasing: ACA enrollment, Quality risk pools, Oscar’s revenue + margins.
Oscar tends to be: Price-competitive, Lower premium than legacy carriers, Attractive UI + member support, & Solid provider access.
----------------------------------------------------
In the ACA market, Risk Adjustment (RA) = the difference between making money or losing money.
Oscar’s entire stack is engineered around RA:
-predicting risk
-identifying missing codes
-guiding care
-stratifying members
-closing gaps automatically
When the massive shift happens when employers get off of group plans (Think Pensions to 401ks) - Anthem, United, Cigna, & Others will need to compete. (They will be nowhere near as successful). Oscar is a Tesla and the rest of the market is a 95 Honda Civic.
If RA becomes increasingly tech-driven (it is), Oscar has a long-term moat.
Other insurers will NEED Oscar's tech stack - but they won't build it. Why? Building a new core system = multi billion dollar death march. You can't just buy "Oscar's stack" - Oscar build in house. This is why +Oscar is a major future revenue stream that could turn this into an AI Healthcare Software Company. They can sell claims automation, risk adjustment engines, member engagement tools, care navigation models. Essentially becoming the shopify of insurance ops.
bUt whAt aBouT the sUbSidies???
Thanks for listening.


































