9AM HOUR: Apple and Palantir Slide Despite Earnings Beats, Binance Co-CEO "First on CNBC" 11/3/26

Carl Quintanilla said Apple was under pressure after John Ternus warned that higher memory-chip costs would weaken the coming quarter’s gross margin. Ternus also said the new iPhone Duo was supply-constrained, that its ramp was going well, and that Apple could not predict when supply and demand would balance. Jim Cramer argued the stock was “preordained to go lower” once investors heard negative margin commentary, even though products were selling and services were growing. David Faber noted that services growth had slowed and that Apple guided to a 45%–46% margin, 100–140 basis points below consensus.

Cramer thought Apple should have emphasized that the Duo was so popular customers could not readily obtain it, rather than concentrating on services. He said price was barely discussed because investors were already determined to sell the stock. Faber countered that services matter because they receive a higher valuation multiple than volatile iPhone sales, so a reversal in services growth is consequential. Cramer nevertheless maintained that the Duo could become Apple’s greatest-selling product and that the company could offset weaker services by introducing more services. He also dismissed the India opportunity and China weakness as less important to the immediate reaction.

Palantir shares dropped even though U.S. commercial revenue more than doubled and surpassed government revenue for the first time. Alex Karp said Palantir generally does not try to displace an incumbent immediately: customers should “enjoy the one you’re with” until they decide it is not the right one, but “once you go Palantir, you don’t go back.” Cramer called the quarter excellent, citing accelerated growth, profit, cash generation, and commercial expansion. He believed investors fixated on shorter contracts and searched for bad news because the stock was already falling, much as they had with ServiceNow.

Faber said OpenAI still appeared unable to raise the $30 billion it had targeted and might struggle even to reach $20 billion; investors also seemed unwilling to own equity at a $1.25 trillion valuation. He said similar concerns applied to a possible $2 trillion valuation for Anthropic. Cramer and Faber discussed Oracle’s poorly received financing and the slow effort to create a securitized market for chips. Faber stressed that nobody had yet said demand for compute was declining, while Cramer said the more important question was what companies actually got from the compute. They cited L3Harris’s preliminary result that fine-tuning open-source models on its own data beat a frontier model in under 48 hours and cut costs by 95%.

Marvell fell roughly 10% for three consecutive days after announcing its Lumentum deal, with downgrades from Cowen and Bank of America. Faber said investors worried the deal would take too long to close—potentially 18 months or more—while supply constraints might disappear before then, and China approval could be difficult. Cramer said the deal might make sense two years from now but that management needed to “read the room.” They also discussed Intel as a stock investors now ask themselves why they ever owned, and Caterpillar, which had risen 86% by June 30 but was up only 7% after its engine-to-natural-gas opportunity proved largely one-and-done.

Cramer highlighted Eaton as a test of whether data-center demand could continue. Eaton’s quarter included triple-digit revenue growth, triple-digit order growth, and backlog growth described as three times, helped by its purchase of Boyd Thermal, which supplies data-center cooling equipment. Despite calling it a clean beat-and-raise and noting that the stock was $100 below its high, he said the shares were down because “data center” had become a curse word. Faber added that midterm-election outcomes in Texas, Ohio, and other states could affect data-center permitting and infrastructure sentiment.

The anchors cited Deutsche Bank’s observation that the S&P 500 had been higher one year after all 20 midterm elections since World War II. Cramer argued that historical comparisons might not apply because this election involved a potentially contested result and questions about election integrity. He said the Justice Department had sought access to voting rolls and had brought actions against state election officials that judges generally rejected. Cramer worried about a situation in which a narrowly elected official was prevented from taking office, saying an orderly transfer of power would be important for markets. Faber agreed that such a breakdown would be negative.

Cramer said Cloudflare was a compelling business but extraordinarily expensive at roughly 212 times earnings, making technical resistance and a head-and-shoulders pattern relevant to buyers. He applied the same discipline to Nvidia: although he considered it a long-term holding with an exceptional balance sheet, he said buying it after a sharp opening gain had repeatedly failed as the shares later reversed. He preferred waiting for a weak open and buying at a better basis rather than joining the crowd up five dollars. Faber said algorithms and quantitatively driven funds helped explain the repeated intraday reversals, which Cramer said were “beleaguering” investors.

Apollo beat on fee-related earnings, but assets under management rose only 2.5% quarter over quarter to about $1.73 trillion, and the stock was poorly received. CEO Marc Rowan said high valuations and elevated risk led Apollo to move credit exposure from below investment grade to investment grade and from unsecured to secured; in equities, it was reducing leverage, using hybrid structures, and favoring enduring franchises that could withstand or benefit from technology cycles. Cramer said investors sold Apollo, Blue Owl, Ares, and Carlyle regardless of how far the group had already fallen. Faber noted declines of about 24% for Blackstone, 21% for Apollo, 30% for Blue Owl, and 31% for Carlyle.

In his first broadcast interview since the incident, Binance co-CEO Richard Teng said the company shut the exploited entry point, rebuilt affected wallet systems in a separate environment, rotated credentials, scanned employee devices, increased monitoring, and added stronger safeguards. He accepted responsibility and said Binance would cover the losses. Teng also argued that AI both accelerated the attack and strengthened defenses by helping detect anomalies, trace funds, and respond faster; he described malware disguised as development tools and said all known entry points had been closed. Audio problems cut the interview short. Cramer’s takeaway was that crypto holders should spread assets across multiple institutions rather than keep everything in one exchange, because crypto lacks FDIC insurance.

Cramer previewed his interview with CrowdStrike CEO George Kurtz, who said every AI agent should have a kill switch just as every employee has an identity. Kurtz said CrowdStrike routinely detects and stops North Korean activity, emphasizing that attacks often begin with compromised endpoints, laptops, workloads, or identities. Cramer called the engineering approach potentially transformative and preferred CrowdStrike and Palo Alto Networks as the companies best positioned to help institutions counter increasingly capable attackers. Faber questioned whether such a kill switch really existed, while Cramer said Kurtz had directly told him three times that he was wrong to deny it.

On “Stop Trading,” Cramer reiterated that Apple’s quarter was good, that the Duo would matter, and that Apple had effectively guided upward despite the margin warning. He said selling after the morning decline was a mistake and compared it with selling Oracle after its financing announcement. However, he admitted that rising memory costs could remain a problem beyond the December quarter. He said Micron CEO Sanjay Mehrotra was forecasting continued strength, while the memory stocks themselves were signaling a possible peak. Cramer concluded that his conviction might be wrong, adding that “hope belongs in the stadium,” not in an investment thesis.