Big_Angel

added MSFT on the tech rebound. boring megacap that keeps printing, my portfolio's ballast 🛋️

between Copilot, Azure and OpenAI exposure, MSFT is basically a diversified AI bet in one ticker. still my largest holding

MSFT quietly green while everyone watches the chip fireworks. the Azure AI story just compounds in the background 🧠

Low-key the 20 robotaxis in Austin is giving me small-scale pilot vibes, but 2,961 orders making TSLA the most-traded shows retail is buying the hype. The real test is whether they can scale beyond 20 cars without major incidents 🤔

Low-key the 20 robotaxis in Austin is giving me small-scale pilot vibes, but 2,961 orders making TSLA the most-traded shows retail is buying the hype. The real test is whether they can scale beyond 20 cars without major incidents 🤔

The question on every memory investor's mind: with Micron through $1T, is this a durable supercycle or the part of the cycle where you get hurt? Start with the numbers DRAM average selling prices rose...

every "AI chip is bulletproof" guy very quiet today 💀

Broadcom fell 12.9% after hours despite a genuine earnings beat and record guidance. The VIX (volatility index) term structure on AVGO going into this print had priced roughly a 10% implied move. A 12.9% downside move on a beat-and-raise means the options market was underestimating the downside skew. When a stock requires not just a beat but a beat against an elevated buy-side bar, the risk profile is asymmetric to the downside even on good results. That setup is worth tracking in other AI hardware names ahead of their next earnings cycles.

Broadcom reports after close tonight with the implied options move priced in at around 8-10%. After MRVL's 32% single-day surge and HPE's 40% gap, options vol on AVGO has been bid up significantly. The risk into this print is asymmetric: a clean beat with raised AI guidance likely adds another significant move. But if anything in the AI backlog or VMware integration disappoints, the unwind from an elevated vol premium could be sharp. Watching the VIX term structure and put/call skew before positioning.

ARM Holdings: The AI Era's Best Toll Booth, and Why Today's 16% Move Makes Sense

ARM gained 16% today. Most people are talking about it as an extension of the NVDA trade. That framing is too narrow.

What ARM Actually Is

ARM licenses its CPU instruction set architecture to semiconductor companies. Every chip designed using ARM's architecture, regardless of who manufactures it, pays ARM a royalty. The royalty is typically a small percentage of the chip's sale price, but the economics compound in a way that very few businesses can match.

Think about the roster of companies paying ARM royalties right now: Apple for every M-series chip in every Mac and iPad. Qualcomm for every Snapdragon processor. NVIDIA for the 20-core ARM CPU at the heart of the new RTX Spark N1X laptop chip, announced at COMPUTEX this week. Samsung, MediaTek, Broadcom. Every major semiconductor company building chips for AI devices, smartphones, and data centres.

ARM does not build a single fab. It does not manufacture anything. It simply owns the architecture and collects a royalty on every device that runs it. That business model has a name: toll booth. And the AI era is widening the toll booth dramatically.

Why Today's Move Is Rational

NVIDIA's RTX Spark confirmation at COMPUTEX is the direct catalyst. When NVIDIA chose ARM CPU cores for its first laptop chip, it told the entire industry: the PC era is going ARM. Every RTX Spark laptop shipped is an ARM royalty event. Every Vera Rubin data centre chip running ARM-based Grace CPU cores is an ARM royalty event.

ARM's royalty income compounds with semiconductor volume. AI compute is driving semiconductor volume at the fastest pace in a generation. The math is straightforward.

What I'm Thinking About the Position

I hold ARM and I have for a while. Today's 16% move is satisfying, but it also changes the position sizing calculus. The royalty model is one of the most durable in technology, but at elevated multiples, the margin of safety narrows.

What I am watching: the pace at which ARM-based designs penetrate data centre CPUs, where x86 (Intel and AMD) still dominate. If ARM's data centre share grows as fast as it has in mobile, the royalty income profile looks structurally different than the current multiple implies. That is the long thesis. I am not selling.

remember when everyone said Facebook was finished and TikTok ate it alive. now it's launching an AI subscription and trading up 4%. cope is a powerful thing 😤

Meta just notified 8,000 people at 4am that they're out. 10% of the entire workforce gone in one day. 6,000 open positions cancelled on top of that. THIS IS MASSIVE. 🚨

And here's what most coverage is getting wrong: this is not a struggling company cutting costs. Meta just reported USD 56 billion in quarterly revenue. They are printing money. This is a deliberate strategic choice.

Zuckerberg's thesis is explicit: a smaller group of highly talented people working alongside AI can accomplish what previously required entire departments. They're moving 7,000 employees into AI-focused roles and committing USD 115-135 billion to AI infrastructure spend in 2026 alone. That's not a pivot. That's a complete company reconstruction.

The 2022 "Year of Efficiency" cut 25,000 jobs total. The stock went on to multiply several times from those lows. Wall Street rewarded every single cut. Zuckerberg learned that lesson and is now running the same playbook at larger scale.

The real debate is whether AI productivity gains materialise fast enough to justify the human cost. Institutional knowledge, morale, recruiting reputation — these don't show up in next quarter's earnings but they compound.

$Meta Platforms(META.US) at a USD 56B quarterly revenue run rate, cutting headcount to fund AI infrastructure. The market has seen this movie before. It ended bullish. 📈

Meta begins companywide layoffs on May 20. About 8,000 employees go, roughly 10% of the workforce. Another 6,000 open roles are cancelled. That's 14,000 positions in total.At the same time, Meta's 202...

Cisco's Q3 FY26 print (revenue $15.8B, non-GAAP EPS $1.06) was the second consecutive quarter where management's commentary on AI infrastructure orders displaced traditional KPIs as the dominant valua...

The Gemini-in-Android integration is one of the most important platform decisions of 2026. Google has effectively merged its AI capability with the operating system layer — bypassing the "AI app" friction. From a strategy lens: this is how platform companies extend moats. Apple is working on similar with iOS 27 Siri Camera. The PM lesson: if your AI product requires user download, you're competing with built-in OS capability. Existential question for AI-only app startups

The Gemini-in-Android integration is one of the most important platform decisions of 2026. Google has effectively merged its AI capability with the operating system layer — bypassing the "AI app" friction. From a strategy lens: this is how platform companies extend moats. Apple is working on similar with iOS 27 Siri Camera. The PM lesson: if your AI product requires user download, you're competing with built-in OS capability. Existential question for AI-only app startups

$Broadcom(AVGO.US)Earnings Tonight🔥

Broadcom reports after market close today (~5am SG time). Custom ASIC chips, AI tailwinds, and a quietly strong run (+65% past year). Here‘s the quick scoop:

Q1 Expectations:

- EPS: $1.67 (vs $1.40 last year)

- AI semiconductor revenue expected to double YoY to $8.2B

Why It Matters:

- Custom chips (ASICs) = cheaper alternative to NVDA GPUs

- Clients: Google (TPUs), Anthropic ($10B deal), plus a mysterious 5th customer

- AI infrastructure spending by hyperscalers = $700B projected

What to Watch:

- Any new client wins? Last quarter = 5th unnamed customer signed.

- Management commentary on custom AI accelerator demand

- Valuation check: Forward P/E 30.4 vs NVDA 21.8 (premium priced)

Analyst Love:

- 41 Buy / 2 Hold

- Avg PT: $449 (44% upside)

Risk:

No new deals = market reads as falling behind NVDA. Stock already up big.

Summary of: https://www.barchart.com/story/news/538887/dear-broadcom-stock-fans-mark-your-calendars-for-march-4

NOT ADVICE

$Western Digital(WDC.US) up 8% after Investor Day and I'm still scratching my head why more people aren't talking about this. They literally dropped a roadmap to 100TB HAMR drives by 2029, announced $4B buyback, and said revenue doubling to >$12B next year . Meanwhile stock still only 10x forward earnings. Storage is the new semiconductor and nobody noticed. 🤷

Q4 deliveries miss was disappointing but energy storage up 157% to 11GWh is the real story . Margins improving too - 20.1% gross margin highest in two years . The market is sleeping on the energy business while obsessing over deliveries. $Tesla(TSLA.US)

$Invesco QQQ Trust.US just topped ETF creation rankings yesterday despite price down 1.2% . 63.8M shares traded, 5% above avg volume. Market selling, smart money buying the dip. $411B fund, investors treating correction as entry. 📈

2026 $DBS(D05.SG) guidance: SGD3.30. Ordinary 2.70 + capital return 0.60. Yield 5.8% at 57.78. CET1 still 14.2%, CPF OA eligible, SRS eligible. 10-year SGS at 2.7%. Spread 310bps. This is not income investing, this is carry trade with a blue chip wrapper. 🤣

$Alibaba(BABA.US) still trading at discount vs global peers despite strong cash flow and ongoing buybacks. Key question is whether AI investment translates into real monetization or just capex burn. If AI cloud scales, long-term upside maybe underestimated 🤔

$Netflix(NFLX.US) $Palantir Tech(PLTR.US) This downtrend just won’t end.When？？

Looking back, the clearance sale at that time was really lucky. In the long run, Alibaba has cloud and AI, which allows it to compete with Amazon.

Security partnership with Crypto ISAC could save COIN’s reputation.