Graphic and Text | Sister Tang

On January 28, after the market closed, Meta, Microsoft, and Tesla simultaneously unveiled their earnings reports.

The CEOs of all three companies delved into discussions about AI during their earnings calls, yet the market's reactions were starkly different: Meta's stock soared by 6.6%, Microsoft's took a 6% hit, and Tesla's edged up by 1.7%.

Meta's performance comprehensively outstripped expectations, with Zuckerberg announcing a staggering capital expenditure of up to $135 billion in 2026, nearly doubling the previous amount, to "accelerate AI" and keep pace with rivals. The market embraced this bold move. Microsoft also surpassed revenue and profit forecasts, but Azure's growth rate dipped from 40% to 39%, with next quarter's guidance of 37%-38% barely meeting expectations and a lower profit margin outlook. The combination of slowing growth and margin pressure resulted in a lackluster market response.

Tesla fared the worst among the trio, with both annual revenue and profit declining, marking the first-ever annual revenue drop in the company's history. However, Musk revealed the discontinuation of the Model S and X, with production lines being repurposed for the Optimus robot, and stated that Austin's Robotaxi had commenced operations without safety drivers.

Three companies, three scenarios, three bets. Yet, some common trends emerge from these earnings reports:

1. AI infrastructure investment continues to surge. Meta doubled its capital expenditures to $135 billion, Microsoft spent $37.5 billion this quarter, a 66% year-over-year increase, and Tesla also doubled its 2026 capital expenditures to $20 billion. Altogether, these three companies plan to invest over $200 billion in 2026 alone on data centers, chip acquisitions, and capacity expansion. The AI arms race is far from over.

2. The impact of AI on profit margins is becoming evident. Microsoft's gross margin fell to 68%, the lowest in three years, due to substantial AI infrastructure investment. Although Meta pledged that operating profit would still grow in 2026, its expense guidance reached as high as $162 billion-$169 billion, a year-over-year increase exceeding 20%. Whether high investment can translate into high returns is a question all AI companies must answer in 2026.

3. AI is reshaping corporate structures. Zuckerberg stated that AI enables "one person to accomplish the work of a team," with Reality Labs laying off thousands this month; Musk directly eliminated two vehicle production lines, replacing them with robot production lines. Cost reduction and efficiency enhancement are no longer mere slogans but tangible organizational reforms.

01 Meta: A $135 Billion Wager on 'Personal Superintelligence'

There's little to critique about META's earnings figures themselves—they comprehensively exceeded expectations: revenue of $59.9 billion, a 24% year-over-year increase; EPS of $8.88, compared to the market expectation of $8.19; advertising revenue of $58.1 billion, accounting for 97%; annual revenue surpassing $200 billion, with 3.58 billion daily active users; Q1 revenue guidance of $53.5 billion-$56.5 billion, also higher than anticipated.

However, Zuckerberg clearly wasn't interested in discussing these numbers during the earnings call, repeatedly emphasizing "AI acceleration"—2026 capital expenditure guidance of $115 billion to $135 billion, nearly double the previous year and $20 billion higher than Wall Street expectations. The market wasn't deterred, instead rewarding it with a 6.6% after-market rise.

The reason behind the significant capital expenditure expansion isn't hard to deduce—any investor tracking AI knows that Meta is somewhat lagging in large models. Over the past year, OpenAI, Google, and Anthropic have sequentially released new models, while Meta's Llama has gradually faded into obscurity. Zuckerberg invested $14.3 billion mid-year to acquire Scale AI and recruited founder Wang Hao, but after half a year of effort, no competitive product has emerged yet.

This $135 billion is essentially an attempt to reclaim lost time.

According to the earnings report, the majority of this investment will go towards the "Meta Superintelligence Lab and core business," encompassing data centers and computing power. Company CFO Susan Li admitted during the earnings call that while capacity expansion is rapid in 2025, demand is growing even faster, and most of 2026 will still be constrained by capacity.

Zuckerberg's vision is termed "personal superintelligence": users opening Meta apps will encounter not just recommendation algorithms but an AI that truly understands them; he also believes glasses are the ultimate medium for this vision. However, he himself acknowledged that the initial batch of models is merely decent, with the focus on demonstrating a trajectory of rapid progress—in other words, there's nothing groundbreaking to showcase yet.

Meanwhile, AI is already transforming Meta's internal operations. Zuckerberg stated that engineers using AI tools have witnessed a 30% productivity increase, with heavy users experiencing an 80% rise—"what a team used to accomplish can now be done by one person." Reality Labs laid off thousands this month. The flip side of efficiency gains is layoffs, which will become the norm in 2026.

Despite doubling capital expenditures, the CFO still anticipates operating profit to exceed 2025 levels. However, in reality, what the market truly cares about is whether Meta can deliver a competitive large model to further intensify the already fierce competition. If $135 billion doesn't yield results, the market won't be so lenient.

02 Microsoft: Cracks Emerge in the AI Narrative

Microsoft's situation is relatively more intricate.

Examining the numbers first: revenue of $81.27 billion, a 17% year-over-year increase, surpassing expectations; EPS of $4.14, also exceeding expectations; intelligent cloud revenue of $32.9 billion, a 29% increase; personal computing was the only segment that fell short of expectations, but it's not a major concern; M365 Copilot paid seats exceeded 15 million; commercial bookings grew by 230%. Nadella stated that Microsoft's AI is still in the early diffusion stage, but its business scale already surpasses some of the company's largest traditional business lines.

The numbers appear promising, but the stock dropped by 6.6% after the market closed.

The issue lies not with the current numbers but with future trends. Azure's growth rate slowed from 40% last quarter to 39%, with next quarter's guidance further declining to 37%-38%. While the decline seems minor, two consecutive quarters of slowing growth for a stock with a P/E ratio exceeding 30 means the market expects the company to consistently outperform, not just meet or fall short of expectations.

Margin pressure is even more pronounced—Microsoft's gross margin fell to 68% this quarter, the lowest in three years, with capital expenditures of $37.5 billion, a 66% year-over-year increase. Nadella stated that the company added nearly 1GW of computing capacity this quarter, but CFO Hood mentioned that demand still exceeds supply, with Azure's growth constrained by capacity while expansion continues to erode profits—a contradiction with no immediate resolution.

Another noteworthy figure: Microsoft's commercial remaining performance obligation (RPO) reached $625 billion, a 110% year-over-year increase, but 45% of this $625 billion stems from the $250 billion cloud services contract with OpenAI, which is still unprofitable. Whether this contract can be fulfilled is questionable. In response, Hood only clarified during the earnings call that the remaining 55% of customers are highly diversified and healthy.

However, the issue is that investors don't care about the health or diversity of the other 55%—45% relying on a still-unprofitable customer means the risk doesn't dissipate just because other customers are of high quality.

Over the past few years, Microsoft and Nadella have spun a highly successful AI narrative, with Microsoft's market cap rising from $1 trillion to $3 trillion and Nadella being hailed as the "most successful CEO of the AI era." But this earnings report exposes the underlying problems: growth is slowing, margins are declining, and the largest customer isn't profitable yet...

AI is a genuine need, but how much Microsoft can earn from it and when remains uncertain.

03 Tesla: Musk Abandons Car Sales

In Q4, Tesla's revenue was $24.9 billion, a 3% year-over-year decrease; annual revenue was $94.8 billion, a 3% decline, marking the first-ever annual revenue drop in history; total deliveries fell by 16% year-over-year to 418,000 units; energy storage deployment reached a record 14.2 GWh, a 29% year-over-year increase, with energy gross margin hitting a new single-quarter high.

Overall, the automotive business declined, but the stock rose by 3% after the market closed before slightly retreating to a 1.7% gain.

It must be said that Musk has effectively "trained" the market—no one, including himself, seems to care about Tesla's automotive performance anymore. During the earnings call, he first announced changing the company's mission from "sustainable abundance" to "astonishing abundance," stating that humanity is heading toward a "future of universally high incomes"—while no longer mentioning sustainable energy.

"Sustainable Abundance" is the core vision proposed by Elon Musk in Tesla's Master Plan Part IV. Simply put, it envisions a future society where poverty is eradicated, and living standards for all humanity are significantly elevated because energy, labor, and intelligence are no longer scarce.

Then came the major announcement: Model S and Model X production will cease next quarter. "It's time to retire the Model S and X honorably because we're moving toward a future centered on autonomous driving," Musk said. "If you want to buy one, now's the time to order."

These two models once defined Tesla: the Model S, launched in 2012, proved that electric vehicles could outperform gasoline cars; the Model X, launched in 2015, with its falcon-wing doors remains a focal point at auto shows. They witnessed Tesla's transformation from a Silicon Valley upstart to a global automaker; now, their production lines will be repurposed for robots—the Fremont factory will be transformed into an Optimus production line with a target annual capacity of 1 million units.

Musk's focus has completely shifted. Regarding Robotaxi, Austin's autonomous vehicles have commenced operations without safety drivers, aiming to cover a quarter to half of the U.S. by the end of the year, with expansion to seven cities in the first half of the year; FSD subscribers doubled to 1.1 million, with the buyout option being canceled starting February 14 in favor of a full subscription model. The energy business was the only bright spot, with Q4 energy storage deployment reaching a record 14.2 GWh, and the CFO stating that demand for Megapack and Powerwall is "crazy"; Tesla's 2026 capital expenditures will be $20 billion, which seems modest compared to other giants but is already double the previous year.

Musk finally said, "The future is more exciting than you imagine." He's made similar statements before—in 2016, he predicted full self-driving by the end of 2017; in 2019, he said 1 million Robotaxis would be on the road by 2020. None of these predictions materialized. But this time is different—Musk isn't just making promises but taking decisive action—he personally discontinued the Model S and X, repurposing production lines for robots. This is an irreversible decision. He's betting Tesla's future isn't in selling cars but in Robotaxi and Optimus.

The 3% gain indicates that the market is skeptical but willing to give Musk some leeway.

04 Conclusion

Three earnings reports, three bets, but they all hinge on the same question: whether AI can translate into tangible profits.

Meta is betting on scale, investing $135 billion to catch up in the large model competition; Microsoft is betting on ecosystem, leveraging Azure + Copilot + OpenAI, betting enterprise customers will continue to pay for AI tools; Tesla is betting on form, eliminating vehicle production lines for robots, betting Robotaxi and Optimus can redefine the company.

2025 will be the year of validation. Zuckerberg needs to prove the investment wasn't wasted, Nadella needs to demonstrate growth can rebound, and Musk needs to show this isn't just another empty promise.

Among the three, at least one will be proven wrong—let's wait and see.

Disclaimer: This article is for learning and communication purposes only and does not constitute investment advice. Feel free to like, share, and repost—your support is what keeps us updating!