Snowflake Has a Hot New Product

Snowflake reported revenue growth accelerating to 33% and raised guidance, driven by its new Cortex Code AI product, though the business still relies heavily on third-party compute and remains unprofitable on GAAP terms. Mortgage refinancing volume fell 18% week over week as 30-year mortgage rates rose to 6.65%, pressuring housing-related stocks and consumer spending. Caesars agreed to be acquired by Fertitta Entertainment for about $31 per share, highlighting cash-flow-rich gaming assets and potentially improving Vici Properties’ tenant quality.

Analysis

The market is starting to reward software that is not just “AI-enabled” but AI-native inside proprietary data moats. Snowflake’s setup is strongest where the agent layer increases seat expansion and workflow stickiness, but the hidden issue is margin architecture: as usage shifts toward compute-intensive AI products, incremental gross margin can compress even while revenue accelerates. That makes this look less like a clean SaaS reacceleration and more like a temporary re-rating on top-line surprise, with AWS/other infrastructure owners retaining structural pricing power over time.

The housing/refi slowdown is more important for second-order demand destruction than for direct loan originators. When cash-out refis disappear, spend on discretionary home projects tends to lag by a quarter or two, which is a softening catalyst for anything tied to renovation, outdoor materials, and big-ticket home upgrades. The sharper implication is balance-sheet inventory pressure: builders and intermediaries may protect volume with incentives, but that usually transmits into lower gross margin before it shows up in unit declines.

In gaming real estate, the real signal is that stressed legacy operators are becoming consolidators of stabilized cash flows, not destroyers of them. The Caesars transaction likely improves tenant quality for VICI while reducing refinancing risk embedded in the lease stream, which should support a lower cap-rate narrative for trophy casino/entertainment real estate over the next 6-12 months. The contrarian miss is that the market still prices casino-adjacent assets as if they are cyclical operating businesses, when several of these names increasingly behave like long-duration infrastructure with inflation-linked cash flow.

Key Decisions for Investors

• Long SNOW as a tactical 3-6 month momentum trade on AI product monetization; use the post-print strength to establish a starter position, but size modestly because infrastructure dependency and SBC can cap multiple expansion.
• Prefer AMZN over SNOW on a 12-month basis as the cleaner beneficiary of AI compute demand; pair long AMZN / short SNOW if valuation re-rates further without a clear margin inflection.
• Avoid or underweight RKT, UPST, SOFI, COMP, TREX, and HD for the next 1-2 quarters; refinance weakness and slower home-equity extraction should pressure growth assumptions and margin leverage.
• Long VICI vs short a basket of higher-beta lodging/gaming operators over 6-12 months; thesis is improving tenant quality and cheaper capital structure optics without needing strong Vegas traffic data.
• Add MGM on pullbacks as a relative-value expression on public casino real estate operating leverage; risk/reward improves if the market starts capitalizing cash flows at a lower yield than current multiples imply.