The Long Term Investor

After reading what WXYZ wrote that triggered my curiosity and decided to use Gemini AI and wrote a quick prompt:

Create a table listing the 25 companies with the highest market capitalization in the S&P 500 as of today, in descending order. In the same table, add two columns: one with the current P/E ratio and the other with the forward P/E ratio. Thank you.

Add the Market Cap value to the previous table, placing this information in column 3.
Can you export that table to PDF?

In seconds you have a list, with some interesting data to analise. Sadly I cannot add this document here but just a few notes: as WXYZ well notes a few names on that list are not expensive at all.

Even if you start investing today and simply pick 10 names on this list I doubt if you will not beat SP in 6 years.

Cysco (#22) PER/ F PER around 16.
Caterpillar (#21) PER/ F PER around 15.
Exxon Mobil (#18) PER/ F PER around 13.
JPM (#13) PER/ F PER around 13.
MU (#9) PER/ F PER around 18/12.
etc

 

Let's unpack what's really happening in this stream of consciousness.

The media, objectively, are morons. We don't need more data points to unequivocally prove the media, as a group, have reasonable intelligence and wield weaponized ignorance. And, clearly, something was written you don't like.

Consider this: If you agreed with what the media were saying, you would be losing money. Short term trading is a redistribution of wealth. Long term, wealth can be generated but. short term, it's a shell game. Not everyone can make money in a shell game. Not everyone can break even, either.

The more myopic the media becomes, the easier it is to make money for people who analyze companies and markets. If you consider my returns for the last two years, it's undeniable they wouldn't have been possible if the media were remotely accurate or truthful.

The dumber the media is, the better.

 

BTW, Micron positioned itself for this run in the middle of 2025. I had been following them for years but didn't purchase until last summer.

At this point, most of the gains have been made. I don't see a lot of upside but I do see a little upside and I like how the company is run so I'm hanging on for the short term.

Micron now is like Tesla was in 2022. Sure, they could go up but they could also go down. Most of the meteoric rise has already taken place.

The question of the moment is, what company will benefit most from CXL?

Micron is the horse that did well in the last race. Right now, we are watching the parade to post for the next race. The data is all there...

Place your bets. I know I have.

 

SAD, But ow so true, WXYZ, the media just want the DRAMA , if it's boring, they ignore it.
Try to find FACTUAL, TOTALLY UNBIASED, NON DRAMATIZED NEWS REPORTING,
It's getting harder and harder, heck most of the time I just skip the national news and only listen to the local news, just in case there is local recall of lettuce or something like that.

As far as MU , that one is up in the air, I think TomB is correct, the question I am asking is, where is my bale out point ?
I'm thinking the end of this year, for some of it, to lock in some profits, then maybe a year from now, when they are actually harvesting the profits everybody has been forecasting ? With Washington's cap gains tax I will/or would have to pay an additional 10% state tax on profits over $200K.......

 

It is interesting to see how AI's data gathering and analytics compares to what I am reading and how I see things shaking out over the next 5-10 years in the tech world.

I pit ChatGPT against Gemini with the following vague prompt:

Based on all available but verifiable data, choose 3 technology stocks to be the best performers in the next 5-10 years. Be as accurate as possible based on current AI trends.

ChatGPT's answer:

I’ll ground this in current company filings, AI infrastructure demand, valuation context, and execution risks—not just hype. I’ll treat “best performers” as likely total-return upside over 5–10 years, not safest mega-cap ownership.

The early evidence points toward semiconductors and AI infrastructure—not consumer apps—as the highest-upside area. The hardest call is whether to include Nvidia: it is still the dominant AI company, but its size makes “best performer” harder than “best business.”

As of June 29, 2026, my 3 highest-conviction AI-driven technology stock picks for best 5–10 year performance are:

1. Broadcom (AVGO)
2. AMD (AMD)
3. Marvell Technology (MRVL)

These are not the “safest” three. They are the three I think have the best mix of AI revenue growth, durable demand, upside from current expectations, and strategic positioning. This is not financial advice, and any of these could be very volatile.

1. Broadcom — best overall AI infrastructure pick
Why I’d choose it: Broadcom is positioned at the center of the shift from general-purpose Nvidia GPUs toward custom AI accelerators, AI networking, and hyperscaler-specific silicon. That trend matters because Microsoft, Google, Meta, Amazon, OpenAI-style labs, and sovereign AI projects all want lower-cost, more power-efficient AI infrastructure.

Broadcom’s latest results are exceptional: Q2 fiscal 2026 revenue grew 48% year over year to $22.2 billion, free cash flow was $10.3 billion, and AI semiconductor revenue grew 143% year over year to $10.8 billion. Management also guided Q3 AI semiconductor revenue to $16.0 billion, up more than 200% year over year.

Why it can outperform: Broadcom is not just selling chips; it is selling into a structural custom-silicon cycle. As AI models move from training to massive-scale inference, hyperscalers have strong incentives to design custom accelerators and networking around their own workloads. Broadcom is one of the few companies with the ASIC, networking, and software cash-flow base to benefit.

Main risk: Customer concentration. If a few hyperscale customers reduce custom ASIC orders, growth could slow quickly.

2. AMD — highest upside among large AI chip challengers
Why I’d choose it: AMD is the best public-market “catch-up” bet against Nvidia. It does not need to beat Nvidia outright; it only needs to take meaningful share in AI accelerators, rack-scale systems, and AI server CPUs.

AMD’s Q1 2026 revenue rose 38% year over year to $10.3 billion, and Data Center revenue rose 57% to $5.8 billion, driven by EPYC server CPUs and Instinct GPU shipments. AMD also guided Q2 revenue to about $11.2 billion, implying roughly 46% year-over-year growth.

The clearest catalyst is that AMD is gaining more serious hyperscaler traction. AMD disclosed that Meta plans to deploy up to 6 gigawatts of AMD Instinct GPUs, with the first 1 GW powered by a custom MI450-based GPU, and AMD also cited growing MI450/Helios customer forecasts.

Why it can outperform: AMD has multiple shots on goal: AI GPUs, EPYC CPUs, rack-scale AI systems, semi-custom designs, and edge/embedded AI. Its upside is much larger if inferencing becomes more cost-sensitive and customers actively seek alternatives to Nvidia.

Main risk: Software and ecosystem. Nvidia’s CUDA/software moat remains very strong, so AMD must keep improving ROCm, developer support, supply, and system-level execution.

Advanced Micro Devices Inc. (AMD)

$539.49
+$17.90(+3.43%)Today
$540.02+$0.53(+0.10%)After Hours
Open522.50
Volume26.7M
Day Low496.00
Day High541.99
Year Low133.50
Year High562.99
Market Cap (TTM)890.16B
EPS (TTM)3.05
P/E Ratio (TTM)176.88
3. Marvell — smaller, higher-risk AI networking/custom silicon bet
Why I’d choose it: Marvell has more execution risk than Broadcom or AMD, but it may have more percentage upside because it is smaller and highly levered to AI networking, optical interconnect, custom silicon, CXL, and data-center connectivity.

Marvell’s Q1 fiscal 2027 revenue reached a record $2.418 billion, up 28% year over year, and the company guided Q2 revenue to $2.7 billion, implying 35% year-over-year growth. Management said AI-related bookings were exceptional and raised its fiscal 2027 and 2028 outlook, citing 800G/1.6T optics, 51.2T Ethernet switches, optical scale-up/scale-out products, and custom XPU solutions.

Marvell also announced the availability of an industry-first 102.4 Tbps switch for AI and cloud data-center infrastructure, which supports the broader thesis that AI clusters increasingly bottleneck on networking, memory movement, and optical interconnect—not just GPUs.

Why it can outperform: AI data centers are becoming networking factories. If GPU clusters scale from tens of thousands to hundreds of thousands or millions of accelerators, Marvell’s optical, switching, and custom silicon exposure becomes more valuable.

Main risk: Marvell is more dependent on future design wins and customer ramps. It is not as financially proven as Broadcom.

Why I did not put Nvidia in the top 3
Nvidia is still the best AI business. Its Q1 fiscal 2027 revenue was $81.6 billion, up 85% year over year, with Data Center revenue of $75.2 billion, up 92% year over year, and gross margin around 75%. That is extraordinary.

But for best stock performance from today, Nvidia’s challenge is size. It is already a multi-trillion-dollar company, with current market data showing a market cap around $4.76 trillion. It can absolutely keep winning, but a smaller AI beneficiary such as AMD or Marvell may have more room to multiply if execution is strong.

Near misses
Nvidia (NVDA): Best company, but not my best return pick from today’s scale.
TSMC (TSM): Possibly the most essential AI supply-chain company, with Q1 2026 revenue of $35.9 billion and Q2 guidance of $39.0–$40.2 billion, but geopolitical risk and capital intensity reduce the upside multiple.
Micron (MU): Huge AI memory/HBM beneficiary, with recent results boosted by tight supply and reported multi-year commitments, but memory remains cyclical and the stock has already surged dramatically in 2026.

Final ranking
Rank Stock Why it makes the top 3 Risk level
1 Broadcom (AVGO) Best mix of AI ASICs, networking, margins, cash flow, and hyperscaler demand Medium
2 AMD (AMD) Biggest large-cap upside if it takes AI accelerator and rack-scale share Medium-high
3 Marvell (MRVL) Smaller, higher-beta AI networking/custom silicon winner High
My strongest single pick would be Broadcom. My highest-upside pick would be Marvell. My best “AI challenger” pick would be AMD.


Gemini's Answer:

The consensus data among financial analysts, market trackers, and structural AI trends identifies three technology stocks best positioned to dominate over the next 5 to 10 years: NVIDIA Corporation (NVDA), Microsoft Corporation (MSFT), and Broadcom Inc. (AVGO). [1, 2]
These three companies represent the absolute foundational layers of the AI economy. They capture profits from hardware infrastructure, enterprise software deployment, and custom cloud-scaling chips. [1, 2, 3]
1. Hardware Monopolizer: NVIDIA Corporation (NVDA)
NVIDIA remains the undisputed gatekeeper of the AI infrastructure wave. Despite short-term market consolidation, the fundamental demand for raw compute power ensures its multi-year upside. [1, 2, 3, 4]
NVIDIA Revenue & Chip Architecture Trajectory
┌────────────────────────────┐
│ FY 2027: Blackwell Gen │ ──► Projected 81% Revenue Growth
└────────────────────────────┘
│
▼
┌────────────────────────────┐
│ FY 2028: Vera Rubin Gen │ ──► 10x lower inference cost / 4x lower training cost
└────────────────────────────┘

• Architectural Supremacy: The launch of the Vera Rubin chip architecture drops AI inference costs by 10x and training costs by 4x. This ensures high capital expenditure efficiency for buyers. [1, 2, 3]
• Massive Capex Backing: Hyperscalers are projected to deploy over $650 billion in data center capex, directly feeding NVIDIA's top-line growth. [1]
• Market Consolidation: Holding a consensus "Strong Buy" rating from Wall Street analysts, the firm continues to see strong demand that limits cyclical downside risks. [1, 2, 3]

2. Enterprise Monetizer: Microsoft Corporation (MSFT)
While other companies spend heavily to build AI, Microsoft has successfully converted those expenses into recurring software revenue, turning itself into the de facto "utility layer" of enterprise AI. [1, 2, 3]

• Sustained Cloud Run Rate: Microsoft Azure’s cloud division exhibits 40% year-over-year growth, outpacing standard competitors by capturing major cloud workloads. [1, 2]
• Proven Consumer Monetization: The company's AI business has achieved an annualized revenue run rate of $37 billion, growing at a 123% clip due to 20 million+ paid Copilot seats. [1, 2]
• Structural Backlog: A commercial remaining performance obligation (RPO) of $627 billion provides deep financial insulation and visible revenue streams over the coming decade. [1, 2, 3, 4]

3. Silicon Architect: Broadcom Inc. (AVGO)
As cloud platforms look to reduce their heavy dependence on NVIDIA's general GPUs, Broadcom has emerged as the premier partner for custom, application-specific integrated circuits (ASICs). [1, 2]

• The OpenAI Validation: Broadcom recently collaborated with OpenAI to launch "Jalapeno," OpenAI's first custom inference processor designed to scale their 10GW compute network. [1, 2]
• Predictable Financial Visibility: The company possesses an immense $73 billion total AI backlog scheduled for delivery, yielding concrete revenue predictability. [1, 2, 3]
• Custom ASIC Dominance: Management has outlined a clear path toward surpassing $100 billion in cumulative AI chip revenue by 2027, establishing a wide competitive moat in the custom cloud space. [1, 2]

Comparative Overview of Core Growth Vectors
Stock Ticker [1, 2, 3, 4, 5, 6, 7, 8, 9] Primary AI Growth Engine Long-Term Strategic Edge Current Financial Health Indicator
NVDA Data Center Infrastructure Vera Rubin processing efficiency Estimated 81% revenue growth for FY27
MSFT Enterprise SaaS & Cloud Massive horizontal integration $627B Commercial Backlog
AVGO Custom ASICs & Networking Collaborative co-design ecosystems $73B secured 18-month AI backlog


I have had my eye on Marvell for a while due to their leadership in ASICs and interconnects. I have been aware of Broadcom, ahem, Avago forever. I have a lot of leftover anger with them over what they did to PLX switch chip prices a handful of years ago. It's good to see the Nvidia and AMD love as well.

I have no intention of doing anything with my Nvidia holdings, but AMD and Marvell are interesting companies due to the massive desire to move away from Nvidia's stranglehold in the AI hardware arena. I've already committed to expanding my AMD holdings, but Marvell looks like it maybe on the horizon. I'll have to read more about them to see if they and AMD are a good hedge against Nvidia maturing and delivering less massive stock price growth in the future.

 

As far as CXL, I'm in a quandary over that one, maybe just get SMH (the semi's ETF) and XLK (TECH ETF) and call it good ?
OR ......Marvell ? Intel ? AMD ? NVDA ? Broadcom ? maybe a little of each ?
I'd ask TOMB, but we don't pay him enough .........

HA, we were typing at the same time, I guess "ask and ye shall receive" may be the phase of the day.

 

@roadtonowhere08 same question, now Claude answer!

I'll pull some current data before discussing this, since stock positioning and AI trends shift quickly and I want to ground this in up-to-date information rather than my training data.Worth saying upfront: predicting specific stock winners over a 5-10 year window is genuinely not something anyone — including professional analysts — can do reliably. Stock prices already bake in a lot of expected growth, technology cycles shift quickly (the AI leaders of 2026 may not be the leaders of 2031), and even great businesses can underperform if you pay too much for the stock today. I'm not a financial advisor, so what follows is factual context on companies that currently sit in strong structural positions within the AI buildout, not a recommendation to buy.

With that framing, here's how I'd think about three companies that show up consistently across analyst coverage as durable, diversified ways to have AI exposure, each representing a different layer of the AI stack:

Nvidia (NVDA) — the compute layer. Nvidia still controls an estimated 80–90% of the AI training chip market through its current GPU architectures, with a CUDA software ecosystem that creates a meaningful moat since most AI researchers write code optimized for its hardware. The bull case is that it remains the default infrastructure layer for AI training and inference. The honest risk side: it carries a high valuation and meaningful customer concentration risk, since a relatively small number of large hyperscaler accounts represent a substantial share of its data-center revenue — so any slowdown in big tech's AI capex would hit it disproportionately. There's also rising competition from custom silicon (more below).

Taiwan Semiconductor (TSM) — the manufacturing chokepoint. TSMC is structurally hard to bypass because it manufactures chips for Nvidia, Apple, AMD, Qualcomm, and virtually every other major semiconductor designer, making it both the bottleneck and the beneficiary of the AI hardware boom. High-performance computing, including AI workloads, made up roughly 58% of its fiscal 2025 revenue, and the company projects overall revenue growth of about 30% in 2026 with AI chip revenue potentially growing 50%+ annually through 2029. The main risk isn't business execution, it's geopolitics — TSMC's Taiwan headquarters introduces concentration risk tied to U.S.-China tensions, which is a real, non-trivial tail risk over a 5-10 year horizon.

Broadcom (AVGO) or Alphabet (GOOGL) — the diversified platform layer. These are slightly different bets, so it's worth distinguishing. Broadcom benefits because it's not a pure GPU bet: it sells both custom AI accelerators to hyperscalers like Google and Meta and the networking switches that connect data center clusters, giving it strong near-term revenue visibility, with AI chip sales projected to climb from $20 billion in fiscal 2025 to $60-90 billion by the end of fiscal 2027. Alphabet's case is different — it's a full-stack AI company that owns its own chips, large language models, cloud infrastructure, and consumer-facing apps, with analysts expecting revenue and EPS to grow at roughly 19% and 16% CAGRs from 2025 to 2028, and it doesn't carry the single-product concentration risk Nvidia does.

A few things worth holding in mind regardless of which names you're drawn to. First, there's a live and unresolved debate about whether AI infrastructure spending is in bubble territory — most economists surveyed in one recent newsletter said we're likely in some form of AI bubble, drawing comparisons to the dot-com era, though others point out that current valuations are still broadly supported by actual earnings and revenue growth rather than pure speculation. Second, market structure is shifting underneath the leaders: leading hyperscalers increasingly have in-house teams building their own chips, which is a long-term threat to companies that sell processors to those same hyperscalers at a premium. Third, today's highest-momentum names (memory and storage stocks have had extraordinary 2026 runs) carry real questions about durability — a key debate across the AI buildout is whether the current demand surge is durable, or whether pricing power erodes once supply catches up.

If you want, I can pull current valuation multiples (P/E, forward growth estimates) for any of these names side by side, or look at how a diversified basket across compute, manufacturing, and platform layers might balance risk versus a single-stock bet — that's generally what most of the analyst coverage I found recommends over picking just three names.

 

Well....we have some "common ground" on all models....AVGO (Broadcom). May worth a look.
NVDA also present in two recomendations.