Feb 28, 2026 · 44 min
AI spending forces investors to rethink scale and risk
Mad Money w/ Jim Cramer 2/27/26
The episode connects battered technology stocks, data-center construction, and massive AI commitments to a broader question about how markets value unprecedented spending.
- 1Data-center infrastructure is becoming a central investment theme as Sterling shifts toward higher-margin, mission-critical work.
- 2Flutter’s sports-betting narrative has weakened as prediction markets and disappointing results pressure the sector.
- 3OpenAI and CoreWeave’s capital demands challenge familiar assumptions about evaluating growth businesses at unprecedented scale.
Don't miss
Joseph Cutillo details how Sterling Infrastructure is building and serving modern data-center campuses while pursuing higher margins and cash flow.
The brief
February punished software, hardware, and NVIDIA while pushing investors toward established consumer brands, financials, and companies supporting the data-center economy.
Viewer calls test the market’s appetite for turnarounds and income: Starbucks faces execution questions, KeyCorp offers a dividend and management case, and Nike and Roku divide opinion.
Sterling Infrastructure illustrates the infrastructure shift, moving from lower-margin highway work toward data-center campuses, electrical systems, skilled labor, acquisitions, and stronger cash flow.
Sterling CEO Joseph Cutillo explains why reliable delivery, recruiting, margins, and a strong balance sheet matter as modern data-center projects grow in scale.
Cramer argues that OpenAI’s fundraising and chip purchases, alongside CoreWeave’s capital needs, may require investors to rethink how they assess enormous corporate spending.
What was said on this episode
37 statements · 22 positive · 9 negative · 2 mixed · 4 neutral
Greg Abel will adopt a lower-key leadership style than Warren Buffett.
“I'm betting the new CEO, Greg Abel, will opt for a much more low-key style than his predecessor.”
Listen at 2:19
Norwegian Cruise should sell itself to Disney.
“My suggestion for Norwegian Cruise? Sell yourself to Disney”
Listen at 2:51
Disney acquiring Norwegian would be strategically smart given shipbuilding delays.
“I think it's a smart move as ordering a new cruise ship takes about five years.”
Listen at 3:34
Disney should buy Norwegian and emphasize entertainment over linear television.
“So I say we buy Norwegian and put the emphasis on entertainment and not linear TV, which would be behind us.”
Listen at 3:51
Target must quickly reinvent itself to remain relevant.
“Mike's going to have to reinvent to stay relevant, I think, and he's got to do it quickly.”
Listen at 4:33
CrowdStrike stock will rise when investors again value cybersecurity expertise.
“I stand by CrowdStrike, which is why we own it for the Chapel Trust. One day soon, expertise will matter again and the stock will go higher.”
Listen at 5:39
Marvell stock is a buy before earnings.
“I think the stock's a buy going into the quarter.”
Listen at 8:17
Starbucks may suffer if it cannot close weak stores and reinvest in stronger ones.
“What can hurt them is the inability to be able to close all the stores that aren't doing that well and then put the money towards the ones that are doing well.”
Listen at 10:14
Brian Niccol will improve Starbucks’s geographic store distribution over time.
“And I think he's going to take care of that. But it just takes time.”
Listen at 10:39
Investors should buy KeyCorp if its stock falls to $19.
“I actually would advise people if this stock fell to $19, I think you pull the trigger.”
Listen at 11:35
Sterling Infrastructure shifted its strategy toward margin growth and cash flow.
“our focus shifted to margin growth and cash flow.”
Listen at 14:49
Sterling attracts site-development labor through higher pay, long hours, and powerful equipment.
“We tend to pay more. Our guys work 60 hours a week. And we have the biggest and most powerful toys in the industry, right? So we can attract that talent.”
Listen at 15:53
A large data-campus project can involve hundreds of millions in site and electrical work.
“the exterior, the site development package on that would probably be three to four hundred million dollars. The electrical package on each one of those buildings is about five hundred million dollars.”
Listen at 17:26
Sterling delivers data-center projects on time and faster than competitors.
“what we deliver is a project on time, every time, faster than anybody else.”
Listen at 18:58
Sterling improves profitability through project management and technology.
“we make our profitability or our increased margins through project management and technology.”
Listen at 20:48
Sterling Infrastructure is cash-positive and has capacity for further acquisitions.
“No, we're cash positive. We've got a great balance sheet. and a lot of drive louder for more acquisition.”
Listen at 21:24
Major online sports-betting stocks have declined sharply for months.
“the stocks of the main online sports folks have been getting killed for months”
Listen at 28:22
Investors should avoid online sports-betting stocks until companies address prediction-market competition.
“I say don't catch your line until these guys come to grips with the fact that the prediction markets are the real problem.”
Listen at 28:44
Nike’s turnaround requires at least one to two years to evaluate.
“We have to wait a full maybe year, maybe two years to see what he does.”
Listen at 29:48
Nike’s performance will improve with more time under Elliott Hill.
“I'm betting that with more time. It will not be disappointing.”
Listen at 30:02
Roku is performing well because of advertising and targeted advertising.
“I think it is because it's got advertising and it's got targeted advertising.”
Listen at 30:26
Roku stock is a smart purchase.
“I think it's a very smart thing to buy.”
Listen at 30:38
Investors should avoid online sports-betting stocks until problems are addressed.
“I say you've got to stay away from the cohort until they come to grips with what's really going wrong.”
Listen at 30:51
Investors should allocate 5% to 10% of net worth to gold.
“I think between five and 10 percent.”
Listen at 32:46
A portfolio-wide own-and-never-trade strategy can cause substantial losses.
“If you have a portfolio of things that you own, don't trade, you're going to lose a lot of money.”
Listen at 33:39
Marvell CEO Matt Murphy will deliver strong results.
“Matt's going to do a big number.”
Listen at 35:18
Investors should buy more Marvell if its stock declines after earnings.
“If it drops next Monday or Tuesday, buy more Marvell.”
Listen at 35:20
Investors should sell rather than buy Lumentum after its sharp rise.
“A stock like Lumentum that is up straight line, it's just too dangerous for me. I would be a seller, not a buyer.”
Listen at 35:55
Investors should take partial profits in Applied Materials after its parabolic rise.
“I would take a little bit off. It's a parabolic chart. And then come back to it if you want to.”
Listen at 37:16
Investors may hold one speculative rare-earth position but should not hold two.
“you can buy one of these. OK, let that be your one. Do not betray me and buy two of them. That'll be too dangerous.”
Listen at 37:58
International Flavors & Fragrances has recovered after a prolonged downturn.
“IFF is back. After being in the wilderness for almost a decade, IFF is back.”
Listen at 38:47
Transocean stock could rise but carries substantial debt and is not preferred.
“The fact is that it can go higher, but I do know it's got a lot of debt. It's not my favorite.”
Listen at 39:08
Halliburton and SLB are Cramer’s preferred oilfield-service stocks.
“My favorite is Halliburton, and my second favorite is SLB. I like quality.”
Listen at 39:13
Investors should buy Coach stock incrementally, adding if it falls.
“You want to buy, buy, buy some, buy half of it, and then if it falls, buy some more.”
Listen at 39:57
Investors should view $100 billion as a relatively modest data-center investment.
“I think you need to think of $100 billion as, well, not a lot of money.”
Listen at 40:58
Data-center investors expect gains within two years of spending $20 billion.
“they're expecting a much quicker return than most of us would imagine. They would not be surprised if they lay out $20 billion for data center and start generating gains in two years time from completion. Maybe, maybe even less.”
Listen at 42:27
Hyperscalers that underinvest risk being overtaken by competitors.
“if they don't spend the money. they're going to end up being eaten alive by the competition.”
Listen at 42:43
Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.
Weekly email
State of the Pod, every Monday
The week's most-discussed names and stories from podcasts, with the moments behind them. Free, five minutes, unsubscribe in one click.