Key takeaways

· Demand looks solid, but markets now price the cost of growth as tightly as the growth itself.

· Guidance is the main event: spending plans and timelines drive the first reaction.

· The artificial intelligence cycle looks real, but payback stays uneven across ads, cloud, and cars.

Earnings nights can feel like school reports. This one feels more like a project review: who is building, who is paying, and who can prove it.

Meta, Microsoft and Tesla report earnings after the US market close on 28 January 2026. They share the same backdrop: artificial intelligence (AI) is now capital intensive, and investors want a clearer link between spending and cash returns.

Coming into the results, the market's mood is already uneven. As of the close on 28 January 2026, Meta is up 0.7% over the past month, while Microsoft is down 1.4% and Tesla is down 9.1%. The numbers matter, but the message matters more. The question is not "did they beat". The question is "did they beat in a way that makes 2026 feel easier, not harder".

The same theme, three business models

Meta sells attention. Microsoft sells computing capacity. Tesla sells hardware, but markets increasingly price it like a software and autonomy option.

That is why reading these three together works. You see the AI value chain end to end: data centres and cloud, consumer apps and advertising, and the long shot into the physical world.

It also explains why guidance dominates the first move. In an AI cycle, the profit line is the snapshot. The capital plan is the plot.

Meta: the ad engine pays, the AI engine consumes

Meta's quarter is strong at first glance. Revenue is 59.89 billion USD versus a Bloomberg consensus estimate of 58.42 billion USD, and earnings per share are 8.88 USD. Advertising impressions rise 18% year on year, well above estimates, even as the average price per ad grows 6% and misses expectations. The shares jump about 10% in after-hours, a clear sign investors are leaning into the message rather than nit-picking the mix.

In short, Meta sells more ad space than expected, but each slot is not getting more expensive as quickly as the market hoped. That is not a crisis. It is a reminder that ad growth can come from volume, pricing, or both, and 2026 needs more "both".

The forward lines do the heavy lifting. Meta guides first-quarter revenue to 53.5 to 56.5 billion USD, above the 51.27 billion USD estimate. Then comes the part the market wrestles with: higher 2026 cost ranges. Meta sees full-year expenses of 162 to 169 billion USD versus a 151 billion USD estimate, and capital expenditure of 115 to 135 billion USD versus 110.62 billion USD.

The tone is clear and slightly defiant. Mark Zuckerberg frames the strategy as front-loading the build: "We had strong business performance in 2025. I'm looking forward to advancing personal superintelligence for people around the world in 2026."

For long-term investors, the implication is simple. The ad machine funds the AI build, but the market now asks for milestones. Meta can afford ambition. It cannot afford ambiguity.

Microsoft: demand is real, but the bill is larger

Microsoft's print looks solid, but it lands in a "show me the payoff" market.

Revenue is 81.27 billion USD versus an 80.31 billion USD estimate. Azure and other cloud services grow 38% excluding foreign exchange, matching expectations. Operating income is 38.28 billion USD versus 36.55 billion USD.

If you stop there, it reads like a clean quarter. The softer after-hours reaction, with the shares down more than 4%, is not a vote against growth. It is a reminder that investors are now pricing the cost of delivering that growth just as sharply as the growth itself. It is about what it costs to deliver it. Capital expenditure including finance leases is 37.5 billion USD versus 36.26 billion USD expected by the average Bloomberg analyst. Microsoft Cloud gross margin slips to 67% from 70%.

Investors do not dislike investment. They dislike investment without a clear timetable for returns. Microsoft offers one form of visibility: backlog. Commercial remaining performance obligation rises 110% to 625 billion USD, and 45% of that growth is tied to new commitments with OpenAI.

This is where the story gets subtle. Azure meets expectations, but in crowded "AI winners" trades, meeting expectations can still feel like a miss. The market is effectively saying: if you are going to spend like the leader, you have to grow like the leader, every quarter.

Satya Nadella leans into the upside: "We are only at the beginning phases of AI diffusion and already Microsoft has built an AI business that is larger than some of our biggest franchises." The next debate is not whether demand exists. It is whether margins can stabilise while demand is being served.

Tesla: profit steadies, the story shifts to autonomy and robots

Tesla brings a different flavour of the same tension. Here the market cares less about cloud growth rates and more about whether Tesla can defend margins while funding a broader platform strategy.

Adjusted earnings per share are 0.50 USD versus a 0.45 USD Bloomberg estimate. Revenue is 24.90 billion USD versus 25.11 billion USD. The headline trade-off is clear: profitability beats expectations, while the top line still looks tight. The shares rose by more than 3% in after-hours, suggesting investors like the mix of steadier profitability and a clearer forward narrative.

The more forward-looking lines sit in the business update. Tesla says it will invest further in clean transport and autonomous robots infrastructure. It expects to ramp six new production lines, and says Cybercab, Tesla Semi and Megapack 3 are on schedule for volume production starting in 2026. It also says Optimus production lines are being installed for volume production.

That language matters because it signals where Tesla wants investors to look. The company keeps trying to shift the frame from "cars sold" to "systems built". The market rewards that narrative in after-hours, but the real test is operational: can these plans turn into volume, reliability, and cash, without leaning on price cuts.

Risks: the early warning signs to watch

The first risk is that the AI build-out turns into a margin problem. If capital expenditure keeps rising while revenue growth slows, the market will price patience lower. Watch cloud gross margin at Microsoft and operating margin at Meta.

The second risk is execution. Tesla's roadmap gets priced like an option, and options lose value quickly when timelines slip. Listen for a shift from "on schedule" to softer language, and watch for delays in volume ramps.

The third risk is policy and regulation. Meta remains under scrutiny across jurisdictions, Microsoft faces competition questions around AI partnerships, and Tesla operates in an incentive and tariff landscape that can change faster than product cycles.

Investor playbook: how to read the next quarter

· Treat capital expenditure as a signal: rising spend is fine if demand and pricing power stay visible.

· Look for monetisation proof points: ad pricing at Meta, AI attach rates at Microsoft, timeline delivery at Tesla.

· Separate "in line" from "better": in crowded trades, "meeting" can still disappoint.

· Track language, not only numbers: guidance ranges and timeline words often move prices more than earnings.

The bill, the build, and the first receipts

If you want one clean takeaway from this earnings night, it is this: AI is no longer a feature, it is an infrastructure race. Meta shows how a powerful ad engine can fund ambition, but also how quickly spending swells when you decide to build at the frontier. Microsoft shows demand is real and backlog is rising, but also that the market will punish a bigger bill even when growth looks healthy. Tesla shows investors still pay for optionality, as long as the path to volume stays credible. The winners in 2026 are not the loudest builders. They are the builders who turn spending into repeatable cash, quarter after quarter.

Source: SAXO