Gil Luria

Micron at 10x earnings looks safer than higher-priced peers; memory supply constraints keep margins high, and the cycle may extend to 2030.

NVIDIA crushed earnings, has no competition in certain segments (dunking on AMD, Broadcom, Intel, Cerebras), and will continue to be the core of the AI trade with strong growth and dominant market position.

Software stocks have underperformed the broader market by the widest margin since 2000. Luria notes MSFT, NOW, SNOW, and DDOG are trading at attractive valuations relative to growth. Newman highlights CRM and NOW are proving AI is an accelerator, not a displacer. The market is pricing in "obsolescence risk" (AI replacing software), but the counter-narrative is that AI is a "labor enhancer." These companies provide the essential infrastructure for AI (MSFT/SNOW) or the "rules and rails" of business (CRM/NOW) that cannot be easily replaced by LLMs. LONG. The sell-off is a tactical opportunity to buy high-quality compounders at depressed multiples. If "Agentic AI" actually begins replacing seat-based SaaS licenses faster than anticipated.

Software stocks have underperformed the broader market by the widest margin since 2000. Luria notes MSFT, NOW, SNOW, and DDOG are trading at attractive valuations relative to growth. Newman highlights CRM and NOW are proving AI is an accelerator, not a displacer. The market is pricing in "obsolescence risk" (AI replacing software), but the counter-narrative is that AI is a "labor enhancer." These companies provide the essential infrastructure for AI (MSFT/SNOW) or the "rules and rails" of business (CRM/NOW) that cannot be easily replaced by LLMs. LONG. The sell-off is a tactical opportunity to buy high-quality compounders at depressed multiples. If "Agentic AI" actually begins replacing seat-based SaaS licenses faster than anticipated.

Memory stocks like Micron and SK Hynix are undervalued because they have locked in five-year supply contracts that reduce cyclical downside, yet trade at 8x earnings versus SOX at 40x, while AI demand continues.

Software stocks have underperformed the broader market by the widest margin since 2000. Luria notes MSFT, NOW, SNOW, and DDOG are trading at attractive valuations relative to growth. Newman highlights CRM and NOW are proving AI is an accelerator, not a displacer. The market is pricing in "obsolescence risk" (AI replacing software), but the counter-narrative is that AI is a "labor enhancer." These companies provide the essential infrastructure for AI (MSFT/SNOW) or the "rules and rails" of business (CRM/NOW) that cannot be easily replaced by LLMs. LONG. The sell-off is a tactical opportunity to buy high-quality compounders at depressed multiples. If "Agentic AI" actually begins replacing seat-based SaaS licenses faster than anticipated.

Palantir and Microsoft provide the AI orchestration layer that lets companies switch between models and avoid lock-in, positioning them as prime beneficiaries of enterprise AI adoption.

Meta’s entry into the cloud compute supply business is directly threatening Nebius and CoreWeave. Meta accounts for over half of Nebius’s backlog and more than a third of CoreWeave’s, and Meta may renegotiate or not renew these contracts, making those backlogs worth much less.

Meta's strategic pivot to sell its excess AI computing capacity to third parties transforms unused data centers into a high-margin revenue opportunity, mirroring SpaceX’s successful model. This move turns underutilized assets into a potential $30B+ annual business and is a no-brainer for the stock.

Meta’s entry into the cloud compute supply business is directly threatening Nebius and CoreWeave. Meta accounts for over half of Nebius’s backlog and more than a third of CoreWeave’s, and Meta may renegotiate or not renew these contracts, making those backlogs worth much less.

Adobe's quarter was good with a beat and raise. It has decelerated less than 80% of software peers over the years. Adobe's products are a management layer for marketing campaigns that will be used by both humans and AI agents, not made obsolete by AI. The selloff is overdone.

The three major hyperscalers—Amazon, Microsoft, and Google—all reported cloud business acceleration above expectations, driven by strong AI demand. This is a very positive sign for their earnings and stock performance.

The three major hyperscalers—Amazon, Microsoft, and Google—all reported cloud business acceleration above expectations, driven by strong AI demand. This is a very positive sign for their earnings and stock performance.

AMD is a buy because of surging CPU demand from AI agents, which creates a bottleneck that benefits both AMD and Intel, but AMD has better execution with higher gross margins (55% vs Intel's 39%) and superior manufacturing economics (TSMC-level profitability vs Intel's losses). AMD also has exposure to GPUs and a key customer in OpenAI, guiding for 35% growth through 2030. The CPU demand increase is happening faster than expected, as evidenced by Intel's strong quarter including selling previously written-down chips.

The transition from Tim Cook to John Ternus as CEO is good news for Apple because Tim Cook is retiring at a time of record iPhone sales and growth, and the new CEO is a hardware expert, which means the company will focus on hardware innovation such as glasses, pins, folding phones, and more affordable VR devices. Apple can avoid the heavy capex race in AI models since AI models will flow through Apple's market-leading premium hardware, providing relief to investors and supporting stock performance.