Cramer Says Nvidia and Salesforce Broke Two Bear Cases
Jim Cramer says Nvidia and Salesforce results dismantled two bear cases weighing on tech. Nvidia closed up 8.74%, Salesforce up 22.58%, as the Nasdaq 100 proxy gained 1.37%.
CNBC's Jim Cramer said strong quarterly results from Nvidia and Salesforce shattered two of the biggest bear cases hanging over technology stocks, on a day Nvidia closed 8.74% higher at 227.98 and Salesforce closed 22.58% higher at 252.05.
Two of the loudest arguments against owning technology stocks ran into earnings this week, and by Thursday's close both were harder to make. CNBC's Jim Cramer said results from Nvidia (NVDA) and Salesforce (CRM) shattered two major bear cases that had been weighing on the group, a view the tape backed with unusual force.
Nvidia finished the session at 227.98, up 8.74% from the prior close of 209.66 — a gain of 18.32 points — and traded in a band of 220.90 to 230.47. Salesforce closed at 252.05, up 22.58% from 205.62, a move of 46.43 points, with a day range of 230.05 to 254.48. Both figures are as of the last trade at 20:00 GMT on Thursday, 27 August 2026. Markets are closed.
The two arguments that had been doing the damage
The bear case against Nvidia has not really been about demand for its chips. It has been about the durability and the plumbing of that demand: whether the hyperscalers and the newer AI-native buyers can keep raising capital expenditure at the pace they have been, and whether some of the money circulating through the AI supply chain is being recycled — vendors and customers financing one another in ways that flatter reported growth. If the capex cycle flattens, the argument runs, the compression in Nvidia's growth rate is sudden rather than gradual.
The bear case against Salesforce has been almost the opposite shape. It is not that spending dries up but that the product category itself is undermined. Software-as-a-service — subscription software billed per user, or per "seat" — depends on headcount. If artificial intelligence agents do work that once required more people at a desk with a licence, seat counts stop compounding and the whole subscription model deflates from the inside. That fear has been the single most persistent overhang on enterprise software valuations.
Cramer's argument, as reported by CNBC, is that the reported quarters were strong enough to dismantle both stories at once — the AI infrastructure trade and the AI-displacement trade, addressed in the same 24 hours.
What the price action actually says
Price reaction is not proof, but a 22.58% single-day move in a company of Salesforce's size is not a rounding error in sentiment. Moves that large in mature large-cap software usually mean the market was positioned for something materially worse, and the bar was cleared rather than nudged. The stock closed at 252.05, near the upper end of its 230.05 to 254.48 range, which suggests buyers held the gain into the bell instead of fading it — often a sign that institutions, not just fast money, were doing the work.
Nvidia's 8.74% gain is smaller in percentage terms but arguably harder to produce, because the company's market value is so large that the dollar amount attached to that percentage is enormous. It too closed inside the top half of its 220.90 to 230.47 range.
The index math followed. The Nasdaq 100 proxy (QQQ) closed at $721.11, up 1.37% on the day and within a whisker of its session high of $721.35. The S&P 500 proxy (SPY) closed at $771.10, up 0.66%. The Dow 30 proxy (DIA) closed at $535.22, up 0.19%. That spread — tech-heavy benchmark up roughly twice the broad market and many times the industrial average — is the signature of a day driven by two large technology names rather than by a macro headline. When a rate decision or an inflation print moves the market, the three tend to travel together.
Where the bear cases still have room to live
A single quarter does not settle either question, and investors would be wrong to treat Thursday as a verdict. The capex argument against Nvidia is inherently a multi-year one: the question is not whether the current quarter of orders is strong — it plainly can be — but whether the buyers of AI compute will earn enough of a return to justify the next round of spending. That is a 2027 and 2028 question, and no report published in August 2026 can close it. Every strong quarter raises the base off which the comparison must eventually be made.
The seat-count argument against Salesforce is subtler and slower. Enterprise contracts renew on multi-year cycles, so any erosion in per-user pricing power would show up gradually, in renewal terms and net revenue retention, long before it showed up in a headline growth rate. A strong quarter tells you the current cohort renewed. It does not tell you what a customer will negotiate in 2028 when its own AI deployment is mature.
What did change is the burden of proof. Until this week, the bear had the easier job: point at the structural risk and let the market discount it. After a 22.58% and an 8.74% day, the bear has to be more specific — which customer, which contract, which line of spending.
What to watch from here
A single quarter does not settle either question, and investors would be wrong to treat Thursday as a verdict.
Three things will tell you whether Thursday marked a turn or a squeeze. First, follow-through: do these two names hold their gains over the following sessions, or does the market give back a meaningful share of the 18.32 and 46.43 points added? Reversals within days are common after outsized earnings pops and would suggest short covering rather than a re-rating.
Second, the read-across. If the bear cases really have been dented, other AI infrastructure suppliers and other subscription software companies should benefit without reporting anything themselves. A narrow rally confined to the two reporters means the market believes something company-specific, not something sector-wide.
Third, the next round of hyperscaler capital expenditure commentary. Nvidia's revenue is, functionally, someone else's capex line. Whether that line keeps rising is the only version of the bear case that ultimately matters, and it will be settled by the buyers, not the seller.
For now, the scoreboard is clear enough: two of the most widely held anxieties in technology investing were tested against actual reported results, and the market's answer was to pay up on both.
Frequently asked questions
What did Jim Cramer say about Nvidia and Salesforce?
CNBC's Jim Cramer said strong quarterly earnings from Nvidia and Salesforce shattered two major bear cases that had been weighing on technology stocks. He framed the two reports as addressing separate anxieties — one about the durability of AI infrastructure spending, the other about AI undermining subscription software — in the same short window.
How much did the two stocks move?
Nvidia closed at 227.98, up 8.74% from a prior close of 209.66, with a day range of 220.90 to 230.47. Salesforce closed at 252.05, up 22.58% from 205.62, ranging between 230.05 and 254.48. Both are last-trade figures as of 20:00 GMT on 27 August 2026, with markets closed.
What is the bear case against Nvidia?
It centres less on demand and more on the plumbing behind it: whether large cloud buyers and AI-native customers can keep raising capital expenditure at the current pace, and whether some capital circulating through the AI supply chain is being recycled between vendors and customers in ways that flatter reported growth rates.
What is the bear case against Salesforce?
Software-as-a-service is typically billed per user, or per seat, so revenue growth depends on customer headcount rising. The fear is that AI agents perform work that previously required more licensed employees, so seat counts stop compounding and subscription revenue deflates from within, regardless of overall IT spending levels.
How did the broad market close on the same day?
The Nasdaq 100 proxy QQQ closed at $721.11, up 1.37%, near its session high of $721.35. The S&P 500 proxy SPY closed at $771.10, up 0.66%, and the Dow 30 proxy DIA closed at $535.22, up 0.19%. The gap between them points to a tech-led session.
Does one strong quarter settle either bear argument?
No. The Nvidia capex question is a multi-year one about whether buyers of AI compute earn returns sufficient to justify future spending. The Salesforce seat question plays out across multi-year renewal cycles. A single quarter shifts the burden of proof toward the bears without resolving either structural concern.
Sources
Photo: McKay Savage from London, UK · BY 2.0 — source
