NVIDIA (NVDA): the criteria that hold back our maximum score
2026-06-23 · By Lubin Danilo, founder of Lubin Investment
NVIDIA is the AI star of 2026 and scores 9/10 in our screener. Three criteria prevent it from reaching 10/10: a high net collection delay (long DSO on data center clients), a current FCF multiple of 45x, and a price of $208.65 that exceeds our entry target of $170.78. The quality is undeniable, but our method stays disciplined on valuation.
NVIDIA in 2026: exceptional quality
NVIDIA has become the indispensable GPU supplier for generative artificial intelligence. Its H100 and B200 chips power the data centers of Microsoft, Google, Meta, and Amazon. In 2026, NVIDIA generates an FCF margin above 50% and triple-digit revenue growth over the previous two fiscal years. Our screener awards it 9 points out of 10 — an exceptional score that fewer than 5% of listed companies achieve.
Criterion 1: net collection delay (DSO)
DSO (Days Sales Outstanding) measures the average delay between a sale and actual cash collection. NVIDIA extends significant payment terms to its large data center clients, which inflates its accounts receivable. A high DSO means that reported revenue precedes actual cash entering the register. Our "net collection delay" criterion penalizes companies whose DSO lengthens year over year, as it is a warning signal about the quality of revenue-to-cash conversion.
Criterion 2: current FCF multiple at 45x
Our screener calculates a normalized P/FCF multiple based on the trailing twelve-month FCF. For NVIDIA, this multiple stands at 45.3x in June 2026. Our preferred acceptability zone is below 30x. A 45x multiple is not disqualifying for a high-growth company, but it falls in our high zone — where the margin for error on future growth assumptions becomes very thin. Any slowdown in GPU demand could significantly compress this multiple.
Criterion 3: valuation above our target
Our method calculates an entry target based on normalized FCF and a disciplined entry multiple. For NVIDIA, this target comes out at $170.78. The current price of $208.65 sits 18.1% above this target. In our framework, buying above the target means reducing the margin of safety — the buffer between the price paid and the calculated value that protects the investor against estimation errors or market shocks.
| Criterion | Value | Lubin threshold | Result |
|---|---|---|---|
| Overall score | 9/10 | 10/10 | Excellent |
| Net collection delay | High DSO | Stable/low DSO | Limits score |
| Current FCF multiple | 45.3x | <30x | High zone |
| Valuation | $208.65 vs $170.78 | Price < target | Above target |
Should we rule out NVIDIA?
No. A 9/10 remains a remarkable score. NVIDIA has a real technological moat: the CUDA ecosystem has locked in AI developers for fifteen years, its customers cannot easily substitute its GPUs in the short term, and its lead over AMD and Intel in inference remains considerable. Our method simply says: this quality is real, but the current price does not leave a margin of safety according to our criteria. NVDA at $170 would be a different conversation.
FAQ
What is DSO and why does it impact NVIDIA's score?
DSO (Days Sales Outstanding) measures in days the delay between invoicing and cash collection. A lengthening DSO means NVIDIA is extending more credit to its clients, delaying cash inflows. Our "net collection delay" criterion penalizes this trend because it dilutes the quality of revenue-to-free-cash-flow conversion.
Is a 45x FCF multiple really problematic for NVIDIA?
Not necessarily in absolute terms, but within our methodology yes. Our preferred zone is below 30x. At 45x, the valuation embeds very sustained growth for several years. If AI GPU demand slows — for instance if cloud clients reduce capex or competitors emerge — the multiple would compress quickly, even if fundamentals remain strong.
What is the difference between a score of 9/10 and 10/10 in your method?
A 10/10 means the company validates all ten criteria simultaneously: operational quality, balance sheet health, buyback discipline, valuation below our target, etc. A 9/10 means one criterion (sometimes two minor ones) is not validated. NVIDIA misses on valuation and DSO. The intrinsic quality remains elite.
Is your $170.78 target for NVIDIA final?
No. Our target is recalculated at each data update (generally quarterly). If NVIDIA continues to grow rapidly, our target will rise. If the market corrects and the price converges toward $170, NVIDIA could reach 10/10. Our target is not a stock price objective, it is a disciplined entry price based on our FCF valuation.
Is NVIDIA comparable to other semiconductors in your screener?
NVIDIA is atypical among semiconductors because its production model is quasi-asset-light (TSMC manufactures its chips). Its FCF margin above 50% far exceeds Intel or Texas Instruments. Analog semiconductors like AVGO or MRVL have different profiles. NVIDIA is closer to a software publisher than a traditional chip manufacturer in terms of cash generation.
Voir l'analyse NVDA sur Lubin Investment
About the author
Written by Lubin Danilo, founder of Lubin Investment. A self-taught individual investor, I have analyzed stocks through their fundamentals for several years and invest my own money with this method. I codified it into a tool that judges a company's quality and its price separately, using criteria drawn from the financial literature (Warren Buffett, Michael Mauboussin, Aswath Damodaran).