Alphabet's earnings growth makes its valuation look attractive. It trades at 21x forward earnings and a 0.47 PEG while EPS grows 34% year over year; shares are down almost 9% year to date. Search still grows 12% and Cloud grows 103%, undercutting the concern that Search will sink the business. x.com  x.com 

Alphabet's rapid cloud growth and diverse profit sources make it a long-term buy below 20x forward earnings. Cloud grew 103% this quarter, despite short-term fears about search and heavy spending. x.com 

Palantir already shows companies returns on AI as they shift from testing it to demanding results. JPMorgan's $262 target is about 32% above the current price; the multi-model signal is Long, with three of four models agreeing ahead of Monday's earnings. x.com 

Alphabet's vertical integration and moat make its current valuation attractive. Its forward P/E is 19.78, below the 22.51 median shown since 2019. Alphabet is also valued below Tesla and SpaceX combined, despite the chart showing Google's quarterly revenue far exceeding theirs; that comparison does not add up. x.com  x.com 

A ten-year low in forward P/E makes hyperscaler valuations look more than reasonable. Goldman Sachs' chart tracks Amazon, Alphabet, Meta, Microsoft and Oracle as a basket. x.com 

Usage-driven AI stocks are less exposed to frontier-model companies as the AI trade separates winners from losers. All rose, all will fall, and only some will recover; favor usage-driven over frontier-biased stocks. DigitalOcean is GAAP profitable; its token usage is a nonlinear revenue proxy. SiTime is GAAP profitable; robotics, defense and self-driving are each growing 50% year over year. CrowdStrike and Rubrik also do not rely on frontier models, though valuation is a concern. x.com  x.com 

CrowdStrike and Palo Alto Networks need 50%-plus growth to deliver even a 10% IRR over a realistic horizon, but almost no conceivable case gets them there. They feel as dumb as Zoom and Peloton in 2021. Even if revenue doubles over four to five years, Palo Alto would still trade at 10x revenue and CrowdStrike at 20x, leaving little prospect of a real IRR; Palo Alto is growing in the teens today. x.com  x.com  x.com 

Oracle looks mispriced because $11 of current consolidated earnings power puts it at 11x earnings without a breakup. It trades below its software business's value despite the highest gross retention in software; separating pre-AI OCI, AI operations and AI development makes it an obvious long. Future funding will cost more, but current leverage is no issue and all contracts are well below spot, leaving nowhere near neocloud risk. Fundamentals may not matter during panics. x.com  x.com  x.com 

Zeta Global, ServiceNow, Datadog and CrowdStrike look increasingly attractive among software names as AI disrupts applications: businesses with proprietary data that build on AI and deliver intelligence and measurable outcomes could win. Cybersecurity valuations remain a concern. x.com 

Low forward earnings multiples across eleven stocks cannot persist indefinitely. Nvidia trades near 17.5x, its lowest in over a decade; Oracle near 15x, its lowest in six years. Uber is at 17x, Broadcom 19x, Vistra 15x, AppLovin 14.5x, TSMC 19x, Forgent Power Solutions 18x, Nokia 19.5x, American Express 17x and Western Digital 14x. x.com  x.com  x.com  x.com  x.com  x.com  x.com  x.com  x.com  x.com  x.com