Mag7News — Week Ending April 2nd, 2026

Holiday-shortened week, full-size swings. Mag7 spent most of the week bouncing back from March’s mess, then had to reprice the same old question in a sharper form: how much does AI really cost when oil, power, and rates all move against you at once? Friday’s jobs report did not help the “Fed will save growth stocks soon” case, but cash equities were closed for Good Friday, so that read-through shifts into next week.

Most retirement plans underestimate two things: how long your savings need to last, and how quietly inflation erodes them along the way.

The 15-Minute Retirement Plan helps you close both gaps with practical guidance on longevity risk, purchasing power, and building a financial plan that doesn't run out before you do.

If you have $1,000,000 or more saved, download your free guide to start.

Download Your Free Guide.

Tesla had the week’s clearest company-specific miss.

What happened: Tesla said it produced 408,386 vehicles in Q1, delivered 358,023, and deployed 8.8 GWh of energy storage. That was light versus expectations, and the stock dropped 5.4% on Thursday as investors focused on soft deliveries, weak storage deployments, and another quarter where production ran ahead of sales.

Why it mattered: Tesla still gets plenty of attention for robotaxis and AI, but the core car business still pays most of the bills.

Impact: The market looked past the big-picture Tesla story and zeroed in on a simpler one: the core operating numbers were not good enough.

Nvidia reminded investors it wants to own more than just the GPU.

What happened: Nvidia announced a $2 billion investment in Marvell and expanded its NVLink Fusion ecosystem. The idea is simple in plain English: make it easier for customers to mix Nvidia’s networking and system tech with custom AI chips, not just standard Nvidia accelerators.

Why it mattered: That makes Nvidia look less like a one-product winner and more like the backbone of the wider AI data-center stack.

Impact: If hyperscalers keep building custom silicon, Nvidia wants to stay in the room anyway, and this deal showed how.

Microsoft sat right in the middle of the week’s power-and-spending debate.

What happened: Reuters reported that Microsoft, Amazon, Alphabet, and Meta were expected to spend about $635 billion on AI infrastructure in 2026, just as higher oil prices started making that math look harder. Then on Friday, Microsoft added a separate strategic signal by announcing a $10 billion Japan investment tied to AI infrastructure, cybersecurity, and workforce training.

Why it mattered: Azure demand still looks strong, but investors are no longer treating AI spend as a free good. Power, construction, and financing costs matter more now.

Impact: Microsoft still looked like a long-term AI leader, but this week made the cost side of that story harder to ignore.

Alphabet had one of the best stock weeks in the group, but the bigger story was still macro.

What happened: Google benefited from the early-week Mag7 rebound and stayed wrapped into the same AI-cost conversation as Microsoft, Amazon, and Meta. Separately, Google launched Gemma 4, its newest open model family, under an Apache 2.0 license and said Gemma downloads had already topped 400 million.

Why it mattered: Alphabet’s upside still comes from AI execution, but the market is also asking what it costs to keep scaling all of this.

Impact: Google got a stock lift from the rebound, but investors are still balancing product progress against AI buildout costs.

Amazon rose, but the week’s real Amazon story was not retail. It was infrastructure.

What happened: Amazon was part of the same group caught in the AI spending meets energy shock story, with investors newly focused on what higher power and oil costs could mean for cloud and AI expansion. As a side note, Amazon also signed Delta for Leo satellite Wi-Fi starting in 2028, a small but useful reminder that it is still building optionality outside AWS.

Why it mattered: Amazon can spend huge sums, but the market wants proof that future AI demand will justify today’s buildout bill.

Impact: For Amazon, the market cared less about one product headline and more about whether expensive infrastructure stays worth it.

Meta bounced hard, but the same investor question followed it around all week.

What happened: Meta was another direct part of the $635 billion AI spending conversation, which put power costs and capex discipline back in focus. The company also launched its first prescription-optimized AI glasses, a good product update but not a week-defining one on its own.

Why it mattered: Meta keeps trying to prove two things at once, that it can keep spending aggressively on AI and still ship consumer products that people actually want.

Impact: Meta’s stock acted better, but the market is still grading it on whether big AI spend turns into durable payoff.

Apple had a quiet week, which was the point.

What happened: There was no in-range Apple-specific headline strong enough to take over the story. Apple mostly traded as part of the broader Mag7 rebound early in the week and then as part of the oil-and-rates reset later in the week.

Why it mattered: When Apple does not bring its own catalyst, it can end up looking more like a passenger in the macro tape than a driver.

Impact: Apple was not ignored, it just did not have enough company-specific news to beat the week’s bigger cross-Mag7 forces.

Marvell (MRVL): Nvidia’s investment and partnership were the clearest adjacent-stock win of the week. Marvell rallied after investors saw it as a bigger part of the custom-chip and AI-networking stack.

Impact: This was a reminder that the AI buildout is not just a seven-stock story.

Delta Air Lines (DAL): Amazon’s Leo deal gave Delta a new satellite Wi-Fi path on 500 aircraft starting in 2028. It was not a top-tier Mag7 catalyst, but it was a real example of Amazon pushing into a new connectivity lane.

Impact: The deal mattered less for next quarter and more for showing Amazon still has useful second-act businesses outside its main profit engine.

The big takeaway was not that Mag7 had a bad week. It was that the market started treating AI like a real industrial buildout, with real power bills, real financing costs, and less room for hand-waving.

Powered by Mag7News — The Magnificent 7 Stocks, Market Momentum in 7 Minutes.