Cramer Dumped a Chipmaker Over $80 Billion in Debt, Kept Nvidia
Jim Cramer exited an AI chipmaker over an $80 billion debt raise while defending Nvidia for the same practice. Nvidia last changed hands at 217.55, down 4.58% on the day.

Jim Cramer said he sold an AI chipmaker after it raised $80 billion in debt but defended Nvidia for the same financing practice, an argument that rests on Nvidia's earnings report due after the close.
Jim Cramer has drawn a line through the AI hardware complex, and it runs straight down the middle of a balance sheet. He sold one AI chipmaker because it raised $80 billion in debt. He then declined to apply the same judgment to Nvidia (NVDA), which uses the same financing practice, according to 24/7 Wall St. Whether the distinction survives contact with reality depends on what Nvidia reports after the bell.
The chipmaker Cramer exited was not named in the segment as reported. What was named was the number: $80 billion of borrowed money, raised by a company whose growth story is supposed to be so good that customers pay cash up front. That is the tension. In a boom, debt is a tool. In a boom that turns out to have been front-loaded, debt is the thing that decides who is still standing.
Why an $80 billion raise reads differently at different companies
Debt is not a moral category, and Cramer's split verdict is more defensible than it looks at first glance. Two companies can issue identical paper for opposite reasons. One borrows because customer prepayments and operating cash flow no longer cover the fabs, the packaging capacity and the memory it has to lock up years ahead of demand. The other borrows because money is cheap relative to the return on the capacity, and because it would rather not sell equity or drain a cash pile it uses for buybacks.
The test is coverage, not size. Does operating cash flow service the interest without assuming next year's growth? Are the maturities spread out, or does a wall arrive in the same window as a possible order-book pause? Is the borrowed cash going into assets with a long useful life, or into prepayments to suppliers that vanish if a customer defers?
Those questions are answerable from filings. The reason Cramer's position is being framed as a double standard rather than an analysis is that the answer at Nvidia has, until now, been taken on faith by a market that has been paid handsomely for the faith. Debt-funded capital spending in AI has spread well beyond the chipmakers themselves — into data center developers, power providers and the neoclouds that buy the chips — and each layer of borrowing raises the sensitivity of the whole chain to a single quarter of soft guidance.
What the tape said before the report
Nvidia was not calm going into the print. The stock last traded at 217.55, down 4.58% on the day from a prior close of 227.98, with a session range of 216.81 to 229.26. That is a close near the bottom of the day's range — the pattern of a stock being sold into the bell rather than bought on dips.
It was also a clear underperformance. The Nasdaq 100 tracker (QQQ) closed at $716.43, off 0.65%, while the S&P 500 proxy (SPY) finished at $769.35, down 0.23%, and the Dow tracker (DIA) was essentially flat at $535.06, down 0.03%. Nvidia's decline was roughly 3.9 percentage points wider than the Nasdaq 100's on the same session — the index was drifting, and Nvidia was being repositioned. All figures are as of the last trade at 20:00 GMT on Friday, 28 August 2026.
The move in points was about 10.43 on the single session. For a company of Nvidia's size, that is not a rounding error; it is the market pricing in a wider distribution of outcomes for the report than it did a week earlier.
The specific line items that settle the argument
For a company of Nvidia's size, that is not a rounding error; it is the market pricing in a wider distribution of outcomes for the report than it did a week earlier.
If Cramer's defense is to hold, the report has to do more than beat on revenue. Investors watching the debt question should look for:
- Operating cash flow versus capital commitments. If cash generation comfortably funds supply prepayments and capacity, the borrowing is opportunistic. If it doesn't, the borrowing is necessary — a different animal.
- Inventory and purchase obligations. Committed spending that cannot be cancelled is the sharp end of an AI slowdown. It shows up in the notes, not the headline.
- Customer concentration. A handful of hyperscalers funding their own buildouts partly with debt means the credit risk is not eliminated by being someone else's, only moved one step away.
- Guidance shape. A quarter that beats but guides cautiously would validate the sellers who marked the stock down 4.58% into the event.
- Gross margin. Memory and advanced packaging costs are the pressure point where a supply-constrained boom starts to cost the seller money.
The read-through for everything downstream
Nvidia's report is not just a company event. It is the closest thing the market has to a scheduled referendum on the AI capital cycle, which is why an index that barely moved could still deliver a nearly 5% single-day drawdown in its most-watched constituent. Equipment makers, foundry suppliers, memory manufacturers, power developers and the leveraged data center builders all take their cue from the same guidance paragraph.
Cramer's split decision is, in effect, a bet that the market can still tell the difference between borrowing from strength and borrowing from need. That distinction has held for most of this cycle. It tends to stop holding the moment a large enough player admits that demand was pulled forward. Nothing in the numbers available before the report answers that question — which is precisely why the stock closed at the low end of its range.
For anyone holding either name, the practical takeaway is unglamorous: read the cash flow statement and the obligations note before deciding whether $80 billion of debt is a red flag or a financing decision. The answer is not the same at every company, and it is not the same at the same company in every quarter.
Key facts
- Nvidia (NVDA) last trade: 217.55, -4.58%, as of 20:00 GMT Fri 28 Aug 2026
- Debt raise that triggered the sale: $80 billion, at an unnamed AI chipmaker
- Nasdaq 100 (QQQ) close: $716.43, -0.65% on the day
- Catalyst: Nvidia earnings due after the closing bell
Frequently asked questions
Which chipmaker did Jim Cramer sell?
The company was not identified in the reported segment. What was stated is that Cramer exited an AI chipmaker after it raised $80 billion in debt, and that he declined to apply the same negative judgment to Nvidia, which uses the same financing practice. Naming the seller beyond that would be speculation.
Why would the same debt raise be acceptable at one company and not another?
Because size alone says little. What matters is whether operating cash flow services the interest without relying on future growth, how maturities are spread, and whether the borrowed money funds long-lived assets or cancellable supplier prepayments. Identical borrowing can signal opportunism at one issuer and necessity at another.
How did Nvidia trade going into the earnings report?
Nvidia last changed hands at 217.55, down 4.58% from a prior close of 227.98, with a session range of 216.81 to 229.26. The close sat near the low of the day, a pattern consistent with selling into the event rather than accumulation. Figures are as of 20:00 GMT on 28 August 2026.
Did the broader market fall as much as Nvidia?
No. The S&P 500 tracker SPY closed at $769.35, down 0.23%, the Nasdaq 100 tracker QQQ at $716.43, down 0.65%, and the Dow tracker DIA at $535.06, down 0.03%. Nvidia's decline was far wider than the indexes, indicating a stock-specific repositioning ahead of results.
What should investors look for in the earnings release?
Operating cash flow measured against capital commitments, inventory and non-cancellable purchase obligations, customer concentration among hyperscalers, gross margin given memory and packaging costs, and the shape of forward guidance. Those items, not the headline revenue number, determine whether debt-funded capacity is a strength or a risk.
Why does one company's report matter to the wider AI trade?
Nvidia sits at the center of the AI capital cycle, so its guidance is read by equipment makers, foundries, memory suppliers, power developers and leveraged data center builders. A cautious outlook would ripple through all of them, which is why a nearly flat index session still produced a sharp single-day drop in the stock.
Sources
Photo: Joshua Mayo · Pexels Licence — source


