# A Closer Look at NVDA's Vendor Financing
In FY 2026, NVDA had $216B in sales. Let's look at the 10K to inspect their vendor financing.
|NVDA|FY26|FY25|
|:-|:-|:-|
|Publicly-held equity securities|17,726|381|
|Non-marketable equity securities|22,251|3,387|
|Total|39,977|3,768|
|Increase from FY25 (a)|36,209||
|Gains equities, net (b)|8,918||
|Equity Purchases (a-b)|27,291||
The first two rows is the amount of equity NVDA held. The number went from $4B to $40B for an increase of $36B. But $9B of that was from gains. Leaving about $27B in equity purchases.
|Company|$|
|:-|:-|
|Intel|5B|
|Coreweave|3B|
|xAI|2B|
|Nokia|1B|
|Random Startups|16B|
|Total|27B|
We know that the $5B in Intel (4% stake) was to help co-develop x86 CPUs with NVLink. **Leaving the remaining $22B for vendor financing.**
So roughly **10% of their 2026FY revenue was from vendor financing.**
They also had Investment Commitments of $11.4B (\~$10B in Anthropic) and Partners’ Facility Lease Obligations of $3.5B. If we include those, we get about $37B or 17%. Granted, that's a little disingenuous because those are future commitments being tied back to FY2026 revenue.
Let's look at their Q1 FY 2027:
|NVDA|**1Q27**|
|:-|:-|
|Publicly-held equity securities|39,101|
|Non-marketable equity securities|43,364|
|Total|82,465|
|Increase from FY 26 (a)|42,488|
|Gains equities, net (b)|15,936|
|Equity Purchases (a-b)|26,552|
So in the first quarter of FY27, NVDA has nearly matched their equity purchases from FY26.
In the first quarter of FY27, NVDA invested $30B in OpenAI. Which is hard to reconcile with the above number. I'm guessing that it's some combination of: my methodology being imprecise, the $30B not fully hitting in Q1, or even NVDA selling some random equities to bring down the net number.
**In my opinion, I think the headlines and Youtube scare videos about circular financing are overblown.** At least as how they relate to NVDA. I think NVDA is making smart investments. From their perspective, they invest a few billion dollars and immediately get that money back (minus cogs). And they still keep the equity.
If you want to make the case that unprofitable AI startups are likely to fail or their IPOs are bad bets, fine. But that's a separate argument. At the end of the day, a few tech startups possibly going bankrupt isn't worth getting all worked up about.