This page tracks every stock Ticker Symbol: YOU has discussed on YouTube, extracted from AI transcription of their 4 most recent videos covering 22 tickers. They are currently bullish on GOOGL, ASML, TSM, LRCX, KLAC, VRT and others. Each take below includes their stated reasoning and, where given, a price target.
The AI infrastructure spending cycle is intensifying dramatically, with Google alone committing $200B in capex for 2026 and $811B in total future purchase agreements including energy contracts running to 2054. The creator frames this as a massive land-grab phase where vertically integrated companies can justify extraordinary spending, while the market is punishing near-term cash burn. Jevons Paradox is in full effect: as AI inference costs fall (Google reduced serving costs 78%), usage and total spending rise even faster.
| Ticker | Action | Why | Target | When |
|---|---|---|---|---|
GOOGL●●● | buy | Google is the only company that owns every layer of its AI stack—models, chips, data centers, networks, and browser—giving it unmatched vertical integration and cost optimization. Google Cloud is accelerating at 82% YoY growth with a $514B backlog and has just started selling physical TPU chips to other companies, a revenue stream neither Microsoft nor Amazon has. While Wall Street sees a cash-burning company, the creator sees the only company on Earth that can justify this level of AI infrastructure spending. | — | 2d ago |
ASML●●● | buy | ASML is the only company on earth that makes EUV lithography machines, giving it an irreplaceable monopoly. Demand is growing faster than their ~30% annual capacity expansion, meaning customers can't rush, replace, or negotiate down ASML, allowing it to set prices. | — | 1w ago |
TSM●●● | buy | TSMC's COOS advanced packaging is sold out into 2027, giving them pricing power with 67% gross margins and 77% earnings growth. The 2nm node transition creates short-term margin pressure but offers 30% power savings, a critical advantage for AI data centers. Risks include Strait of Hormuz exposure and 2nm ramp costs, but the creator sees the 15% drawdown as a long-term buying opportunity. | — | 1w ago |
LRCX●●● | buy | Lam Research supplies deposition and etching machines essential to every fab expansion by TSMC, Intel, Micron, Samsung, and SK Hynix. Buying the 25% dip is a bet that AI chip demand and production will keep accelerating, though if fab expansion slows, LAM will feel it first. | — | 1w ago |
KLAC●●● | buy | KLA holds over 50% of the semiconductor process control market and 80%+ in optical wafer inspection, with its next competitor at just 10%. Its advanced packaging inspection business is expected to grow 50%+ YoY to ~$1B this year. Once KLA's machines are integrated into a fab, switching costs make them extremely sticky, reinforcing market share. | — | 1w ago |
NVDA●●● | buy | NVIDIA holds over 90% of the data center GPU market and demand is accelerating, not slowing. The stock is down over 10% from earnings despite every signal showing supply-constrained demand, making it a great buy. | — | 3w ago |
AVGO●●● | buy | Broadcom designs custom ASICs for Google, Meta, Anthropic, OpenAI and others, and just delivered OpenAI's Jalapeno chip from concept to prototype in nine months versus the typical 2-3 years. The stock is down over 20% from earnings despite demand speeding up, making it a great buy. | — | 3w ago |
CRWV●●● | buy | CoreWeave rents GPU compute to companies that don't want to build their own infrastructure, and demand is so high that every GPU they bring online is rented. Meta's cloud ambitions don't threaten them because Meta has no excess capacity and has $35B committed to CoreWeave as a customer. | — | 3w ago |
NBIS●●● | buy | Nebius has up to $27B in commitments from Meta and is running at full capacity. The deal structure assumes demand stays high, and Alex believes AI demand is only going up, so Nebius will keep renting GPUs even if Meta eventually enters the market. | — | 3w ago |
IREN●●● | buy | As a neocloud, IREN is renting out every GPU it can bring online at full capacity. Alex is still buying all three neocloud stocks as their prices continue to fall. | — | 3w ago |
LITE●●● | buy | Lumentum makes lasers and optical transceivers critical for AI data center interconnects, and NVIDIA invested $2B in equity to lock in capacity. The stock is down 10-20% recently despite having cash backing and multi-year revenue locked in. | — | 3w ago |
COHR●●● | buy | Coherent makes optical transceivers, lasers, and amplifiers for data center and telecom networks. NVIDIA invested $2B in equity to secure capacity, and the stock is down 10-20% despite strong cash backing and committed revenue. | — | 3w ago |
GLW●●● | buy | Corning makes glass and fiber optic cables for AI data center networks and has secured supply deals worth up to $6B with Meta, $3.2B with NVIDIA, and a multi-billion dollar agreement with Amazon. The stock is down 10-20% despite years of locked-in revenue. | — | 3w ago |
QCOM●●● | buy | Qualcomm nearly doubled its non-smartphone revenue target to $40B by fiscal 2029, with $15B from AI data centers via the Dragonfly C1000 CPU. Meta and Microsoft signed supply agreements, and the stock is down 15%, essentially pricing the data center business at zero. | — | 3w ago |
SKHY●●● | buy | SK Hynix dominates HBM with 60% market share, is tripling revenue year over year at 70%+ operating margins, and the entire HBM market is sold out into 2027 with shortages possibly extending to 2030. Its ADR listing on NASDAQ (July 10) gives US investors direct access to the memory supercycle leader for the first time. | — | 3w ago |
Google is the only company that owns every layer of its AI stack—models, chips, data centers, networks, and browser—giving it unmatched vertical integration and cost optimization. Google Cloud is accelerating at 82% YoY growth with a $514B backlog and has just started selling physical TPU chips to other companies, a revenue stream neither Microsoft nor Amazon has. While Wall Street sees a cash-burning company, the creator sees the only company on Earth that can justify this level of AI infrastructure spending.
ASML is the only company on earth that makes EUV lithography machines, giving it an irreplaceable monopoly. Demand is growing faster than their ~30% annual capacity expansion, meaning customers can't rush, replace, or negotiate down ASML, allowing it to set prices.
TSMC's COOS advanced packaging is sold out into 2027, giving them pricing power with 67% gross margins and 77% earnings growth. The 2nm node transition creates short-term margin pressure but offers 30% power savings, a critical advantage for AI data centers. Risks include Strait of Hormuz exposure and 2nm ramp costs, but the creator sees the 15% drawdown as a long-term buying opportunity.
Lam Research supplies deposition and etching machines essential to every fab expansion by TSMC, Intel, Micron, Samsung, and SK Hynix. Buying the 25% dip is a bet that AI chip demand and production will keep accelerating, though if fab expansion slows, LAM will feel it first.
KLA holds over 50% of the semiconductor process control market and 80%+ in optical wafer inspection, with its next competitor at just 10%. Its advanced packaging inspection business is expected to grow 50%+ YoY to ~$1B this year. Once KLA's machines are integrated into a fab, switching costs make them extremely sticky, reinforcing market share.
Vertiv provides power and liquid cooling for AI data centers and is an Nvidia partner for the new 800V DC power architecture ahead of Kyber racks. However, a potential Kyber delay would hit Vertiv, and the creator is waiting for their upcoming earnings report before buying the ~20% dip.
Meta is launching Meta Compute to sell excess AI compute capacity, and as a highly profitable trillion-dollar company it can build without borrowing or diluting shareholders. However, Zuckerberg said they have no excess capacity yet and still have $35B committed to CoreWeave and up to $27B to Nebius, so Meta Compute is a hedge, not an imminent threat.
NVIDIA holds over 90% of the data center GPU market and demand is accelerating, not slowing. The stock is down over 10% from earnings despite every signal showing supply-constrained demand, making it a great buy.
Broadcom designs custom ASICs for Google, Meta, Anthropic, OpenAI and others, and just delivered OpenAI's Jalapeno chip from concept to prototype in nine months versus the typical 2-3 years. The stock is down over 20% from earnings despite demand speeding up, making it a great buy.
CoreWeave rents GPU compute to companies that don't want to build their own infrastructure, and demand is so high that every GPU they bring online is rented. Meta's cloud ambitions don't threaten them because Meta has no excess capacity and has $35B committed to CoreWeave as a customer.
Nebius has up to $27B in commitments from Meta and is running at full capacity. The deal structure assumes demand stays high, and Alex believes AI demand is only going up, so Nebius will keep renting GPUs even if Meta eventually enters the market.
As a neocloud, IREN is renting out every GPU it can bring online at full capacity. Alex is still buying all three neocloud stocks as their prices continue to fall.
Lumentum makes lasers and optical transceivers critical for AI data center interconnects, and NVIDIA invested $2B in equity to lock in capacity. The stock is down 10-20% recently despite having cash backing and multi-year revenue locked in.
Coherent makes optical transceivers, lasers, and amplifiers for data center and telecom networks. NVIDIA invested $2B in equity to secure capacity, and the stock is down 10-20% despite strong cash backing and committed revenue.
Corning makes glass and fiber optic cables for AI data center networks and has secured supply deals worth up to $6B with Meta, $3.2B with NVIDIA, and a multi-billion dollar agreement with Amazon. The stock is down 10-20% despite years of locked-in revenue.
Qualcomm nearly doubled its non-smartphone revenue target to $40B by fiscal 2029, with $15B from AI data centers via the Dragonfly C1000 CPU. Meta and Microsoft signed supply agreements, and the stock is down 15%, essentially pricing the data center business at zero.
SK Hynix dominates HBM with 60% market share, is tripling revenue year over year at 70%+ operating margins, and the entire HBM market is sold out into 2027 with shortages possibly extending to 2030. Its ADR listing on NASDAQ (July 10) gives US investors direct access to the memory supercycle leader for the first time.
Micron is one of three major HBM suppliers for NVIDIA's next-gen GPUs and is reporting that demand is off the charts and supply-constrained.
No bearish takes.
Mentioned as a comparison point; Microsoft has custom chips (Maya), Azure data centers, and Windows/Office ecosystems, but still relies on OpenAI's models for heavy workloads rather than owning the full stack like Google.
Mentioned as a comparison point; Amazon uses Anthropic's models rather than owning its own model layer, and does not sell custom chips directly like Google now does with TPUs.
AMD is mentioned as a competitor to Qualcomm in the data center CPU market with its EPYC line, and as a stock covered on the channel, but no specific investment stance is given in this video.
Intel is mentioned only as a competitor to Qualcomm in the data center CPU market with its Xeon chips. No investment stance is expressed.
The creator argues Google (Alphabet) is the most powerful AI company on Earth because it uniquely owns every layer of the AI stack—models (Gemini), custom chips (TPUs), data centers, fiber networks, and the Chrome browser. Despite a 7% post-earnings drop driven by massive capex, negative free cash flow, and paper gains from SpaceX/Anthropic inflating headline net income, the creator believes Google Cloud's accelerating 82% growth, $514B backlog, and newly begun TPU chip sales justify the spending. He views the cash burn as strategic self-investment, not waste.
The creator identifies three simultaneous bottlenecks hitting AI stocks: ASML's limited EUV machine production capacity, TSMC's sold-out advanced packaging (COOS) into 2027, and potential delays to Nvidia's next-gen Kyber racks. Combined with geopolitical risk from the Strait of Hormuz closure affecting Taiwan's energy supply, AI stocks have sold off sharply. The creator views this as a buying opportunity for long-term investors and recommends five semiconductor and infrastructure equipment stocks that are down 10-25% over the last month.
Alex sees a major buying opportunity after a sharp sentiment swing from greed to extreme fear, with top AI stocks down 10-30%. He covers Meta's new cloud business threatening neoclouds, Broadcom and Nvidia's custom chip advances, optical networking suppliers receiving massive equity investments and supply commitments, Qualcomm's data center CPU push, and the upcoming SK Hynix ADR listing. He remains bullish across AI infrastructure and is actively buying dips in neocloud, chip, and optical networking stocks.
The creator argues that Micron (MU) is a highly undervalued AI stock that has transitioned from a cyclical commodity memory maker to a critical AI component provider. Highlighting Micron's recent record earnings, 85% gross margins, and $100 billion in take-or-pay strategic contracts through 2030, the creator believes Wall Street is still incorrectly modeling the company as a cyclical business. The creator is actively buying shares, citing a forward P/E of just 5.5 and massive growth potential driven by hyperscaler demand for high bandwidth memory (HBM).
Based on AI analysis of this creator's recent videos. Theses are paraphrases of the creator's own words, not verbatim quotes. Not financial advice.