Net retention has stabilised, AI workloads are inflecting at 300% growth, FCF margins hit 25%, and the CEO transition is behind it. The four biggest bear arguments are weakening.
Snowflake trades at 8.9x trailing revenue with no earnings, no dividend, and a stock-based compensation bill that would make a venture capitalist blush. The bears have been loud, and they have been right — the stock is down roughly 50% from its 2021 highs.
But four specific dynamics in the business are shifting in ways that the bearish consensus has not fully incorporated. None of them individually justify the multiple. Together, they suggest the risk-reward is more balanced than the price action implies.
Snowflake's net revenue retention rate fell from 174% in early 2023 to 127% in mid-2025 — a collapse that drove more selling pressure than any other metric. The fear was that the consumption-based model was structurally flawed: customers could simply turn off workloads to cut costs, and they were doing exactly that.
The Q4 2025 print came in at 126%, essentially flat quarter-over-quarter for the first time in two years. More importantly, the largest enterprise cohort (customers spending $1M+) showed retention stabilising at 135%. The bleed has stopped. Historically, when SaaS net retention stabilises after a compression cycle, stock re-rating follows within 6-9 months. Datadog in 2023 is the clearest recent parallel — retention bottomed at 120%, stabilised for two quarters, and the stock rallied 80% over the following year.
TickerXray Report
Get the complete Snowflake report with all 12 quantitative models, AI-generated investment thesis, and real-time data.
Every cloud company claims AI tailwinds. Most of it is marketing. Snowflake's AI story has more substance: Cortex AI, launched in 2024, lets enterprises run LLM inference and fine-tuning directly on data stored in Snowflake without moving it to a separate platform. This matters because the single biggest friction point in enterprise AI adoption is data movement and governance.
Snowflake disclosed that AI-related consumption grew over 300% year-over-year in Q4 2025, representing approximately 8% of total product revenue. At that growth rate, AI workloads could account for 15-20% of revenue by early 2027 — a genuine incremental revenue stream, not a reclassification exercise.
Snowflake is GAAP unprofitable. Net losses totalled $1.3 billion in fiscal 2026. The bears point to this number and call it a cash incinerator.
But strip out stock-based compensation — which accounted for $1.6 billion of the loss — and the company is already free cash flow positive. FCF margins hit approximately 25% in the latest quarter, up from 15% a year ago. The gap between GAAP losses and cash generation is enormous, and it has been closing every quarter.
Look, the stock comp is a real cost to shareholders. It dilutes ownership. But the cash flow trajectory tells you the underlying business economics work. The question is whether management can moderate SBC as revenue scales — and the latest quarter showed SBC as a percentage of revenue declining for the third consecutive period.
Sridhar Ramaswamy took over as CEO in early 2024 after Frank Slootman's departure. The transition spooked investors, and understandably so — Slootman was one of the most respected enterprise software operators in the industry. But Ramaswamy's background running Google's advertising engineering division has brought a sharper focus on product-led growth and AI integration.
Under Ramaswamy, Snowflake has accelerated its product cadence, shipping Cortex AI, Snowpark Container Services, and the Iceberg Tables integration in rapid succession. Customer count grew 22% year-over-year to approximately 10,200 — a deceleration from prior years, but still healthy for an enterprise platform at this scale. The CEO transition risk has largely played out.
Snowflake at 8.9x revenue is not cheap by any traditional measure. The forward P/E of 68x on projected FY2027 earnings prices in significant execution. But with net retention stabilising, AI workloads inflecting, cash flow margins expanding, and the CEO transition behind it, the four biggest bear arguments have all weakened in the past two quarters.
We see the stock as fairly valued here, with asymmetric upside if AI consumption accelerates faster than the 300% growth rate disclosed last quarter. A pullback to $100-110 would create a more compelling entry. Above $160, the valuation requires too many things to go right simultaneously.
Full forensic analysis of Snowflake
+ 6 more models included
150,000+ stocks covered
Global coverage across 60+ exchanges. Every report includes all 12 quantitative models and AI analysis.
View plansEvery report runs 12 quantitative models and generates an AI investment thesis. From Piotroski scores to manipulation detection -- get the full picture in seconds.
12 forensic models
Piotroski, Altman, Beneish, DuPont & more
AI investment thesis
Synthesized outlook on every stock
Manipulation detection
Spot red flags before they hit the news
150,000+ tickers
Global coverage across 60+ exchanges
Expected return
Forward return projections for every stock
Real-time data
Live prices, insider trades, news sentiment
Free accounts get 1 report per month. Pro gets unlimited.
At 85x forward earnings, Snowflake looks expensive until you model Cortex AI driving 40-50% consumption uplift and accelerating the path to $10B in revenue by FY2028.
Uber at 14.9x earnings with $52 billion in revenue and a 19.3% profit margin is no longer a growth-at-all-costs rideshare startup. It's a profitable logistics platform.
CrowdStrike at $96 billion market cap versus Cloudflare at $59 billion — both are growing revenue at 25%+, both are unprofitable on a GAAP basis, and both claim to be the platform that wins cybersecurity. One is overvalued.