Jim Cramer sees more pain for beaten-down AI stocks: ‘I think it’s too soon to say the coast is clear’

On Wall Street, that kind of downslide often creates an opportunity for stocks that have otherwise generated robust returns. In this case, investors can “buy on the dip” and take advantage of lagging AI stocks that should rebound.

But not so fast, warns Jim Cramer, host of CNBC’s Mad Money.

“Even after today’s bounce, I think it’s too soon to say the coast is clear, and it’s safe to dive back into the hardest-hit AI data center stocks,” Cramer said in his Investing Club column on October 18.

The long-time stock picker said multiple issues need to be cleared up before he’d wade into the AI sector.

“I just don’t think it’s safe to be aggressively bullish on the group until the charts improve, the hyperscalers report, the election’s in the rearview mirror, and Anthropic manages to come public,” Cramer added. “That’s what you need to keep your eyes on before circling back to this newly beaten down group.”

It’s not a red light, yet

Other market darlings believe shying away from AI stocks right now is a mistake. But you’ll need to keep a sharp eye on your portfolio.

Take Monte Malukas, founder of London, UK-based Stockatomics, who says he’d separate what’s happening to AI stocks from what’s happening in the physical market.

Malukas said his market analysis company’s AI Infrastructure Index, which tracks the cost of hardware used in AI data centers, is up around 70% over the past year, and memory chip prices have risen nearly fivefold.

“That doesn’t look like demand is collapsing at all,” he told Moneywise. “I’d watch hardware prices and shipment volumes. If they start falling, that could be an early sign demand is weakening — and trade data may show it before company earnings do.”

Another key AI market issue to track is how sector companies are managing their budget for chips, data centers, talent and other resources. Exhibit A is tech giant Meta Platforms (NASDAQ: META), which noted “free cash flow was negative $2.12 billion” in its third-quarter 2026 earnings report. The company estimates total capital expenditure of $135 billion to $145 billion in 2026, with a large share of that capex coming from AI spending.

“Some companies can fund it from cash flow; others are leaning more heavily on borrowing,” Malukas noted. “I’d focus on capital spending compared with operating cash flow and on how quickly component costs are rising.”

Yet AI spending could become a liability. But it’s best to take a wait-and-see approach, according to InvestWealth.com founder Harry Hamann.

“That’s not the case, but the numbers that would tell you have started to move,” he told Moneywise.

Hamann added that Microsoft, Alphabet, Amazon and Meta all point to about $700 billion in capex spending in 2026, and that scenario bears watching, too. Alphabet’s free cash flow went negative in Q2 for the first time since its 2004 IPO, and it sold $20 billion of bonds and $50 billion of equity in the same quarter.

“The spending is rational short term,” Hamann said. “Nvidia just reported $96 billion of quarterly revenue, up 106%.”

That capex liability shows up in three company financial bottom lines.

“You have free cash flow at the buyers, depreciation that lags spending by years, and the gap between Nvidia’s $59.7 billion profit and its $24.1 billion operating cash flow,” Hamann said. “When buyers borrow to spend, and the supplier’s receivables outrun its cash, capital markets are funding the cycle, not yet end demand. Telecom in 1999 is the precedent.”

Stick close to the upcoming Anthropic IPO

Like Cramer, market gurus also advise tech investors to watch the Anthropic IPO, set to kick off in mid-to-late November, with an estimated $1.5 trillion to $2 trillion valuation.

“An Anthropic IPO could give investors a useful look at how an AI company turns demand into revenue and, eventually, profit,” Malukas said. “But one successful listing wouldn’t prove that the whole sector is profitable.”

Investors would still need to look closely at margins, cash burn, computing costs, and how much growth depends on continued investment.

“A high valuation could also make people more cautious if it sets expectations the business can’t meet,” Malukas added. “And while some investors may shift money toward the IPO, I’d see that as a short-term possibility rather than proof that established AI stocks have lost their appeal.”

Perhaps the most useful data coming from an AI kingpin’s IPO is the numbers nobody has today. There are no reliable figures for what a model costs, Sid Verma, founder of Tenet Investing, told Moneywise.