Are Markets Still Worried About an AI Bubble?

Investors searching for signs of an artificial intelligence bubble found a divided reality in the latest quarterly earnings reports, as Wall Street sharply diverged in its reaction to tech giants Meta and Microsoft. While bubble fears persist amidst broader market volatility in early 2026, the market is currently favoring companies demonstrating aggressive investment and immediate revenue impacts from AI, rewarding Meta’s "full steam ahead" strategy while punishing Microsoft for perceived caution and slowing cloud growth.

Key Points

• Market Divergence: Meta shares jumped 8% following a revenue beat and aggressive AI roadmap, while Microsoft stock fell 5% due to slowing Azure growth and backlog concerns.
• Meta's Aggressive Pivot: CEO Mark Zuckerberg announced a massive increase in capital expenditure, signaling a willingness to use debt financing to secure dominance in AI infrastructure and wearables.
• Microsoft's Infrastructure Bottleneck: Despite a $625 billion sales backlog, investors are concerned about Microsoft's ability to build infrastructure fast enough to capture demand, as OpenAI expands partnerships to other providers.
• Hardware Boom: Memory chip giants Samsung and SK Hynix reported doubling and tripling of profits, driven by severe shortages and high demand for AI-related high-bandwidth memory (HBM).

The Tale of Two Tech Giants

The central theme of the current earnings season is the separation of market sentiment regarding AI capitalization. While the debate regarding valuations continues, investors are looking for concrete evidence that AI spending is translating into growth. This dynamic was most visible in the contrasting after-hours reactions to Meta and Microsoft.

According to Bloomberg’s Caroline Hyde, the market verdict was clear: "Meta wins, Microsoft loses." Both companies beat earnings and revenue expectations, and both signaled increased capital expenditures (capex). However, investors struggled to see the immediate rewards of Microsoft’s spending, contrasting sharply with Meta’s ability to articulate how AI is actively driving its 24% revenue growth through improved ad systems and video impressions.

Meta: Aggressive Spending Meets Revenue Growth

Meta’s earnings call was defined by Mark Zuckerberg’s unyielding commitment to AI dominance, regardless of the immediate cost. The company delivered a year-over-year growth rate of 24%, shattering analyst estimates. Zuckerberg credited this success to AI investments improving the company’s recommendation engines and advertising efficacy.

The company issued guidance for a massive increase in capital expenditure, planning to spend as much as $135 billion on data centers this year—nearly double the previous year's spend and 20% higher than median forecasts. Additionally, expenses are projected to rise by 40%, reflecting talent acquisition for AI research.

Zuckerberg also laid out a hardware-centric vision for the future, positioning the Meta Ray-Ban smart glasses as the successor to the smartphone.

"Billions of people wear glasses or contacts for vision correction. And I think that we're at a moment similar to when smartphones arrived... It's hard to imagine a world in several years where most glasses that people wear aren't AI glasses."

While Meta’s generative AI models have yet to dominate consumer consciousness, the company views 2025 as a "rebuilding year" for its foundational models. Zuckerberg indicated that new models and "agentic shopping tools" would roll out in the coming months, aiming to provide uniquely personal experiences based on Meta’s proprietary social graph data.

Despite the massive price tag and the shift toward debt financing—which some analysts noted could deplete cash reserves rapidly—the market rewarded the clarity of Zuckerberg’s vision. By frontloading capex, Meta aims to avoid the risk of underspending, a strategy investors currently favor.

Microsoft: The Cost of Caution

In contrast, Microsoft faced a market pullback despite solid fundamentals. Azure revenue grew at a 38% pace, a deceleration of only one percentage point from the previous quarter. However, in the current hyper-growth environment, even slight decelerations are viewed critically.

The primary concern for Microsoft appears to be a disconnect between demand and infrastructure capacity. The company reported a massive sales backlog of $625 billion, more than double that of the previous year. Notably, $250 billion of that backlog comes from a commitment by OpenAI, which now accounts for 45% of Microsoft's total backlog.

Critics suggest Microsoft has lost its "early mover" advantage. Analysts noted that while Meta is overbuilding to ensure capacity, Microsoft’s CFO Amy Hood has historically maintained a stricter approach to matching data center construction with forecasted demand to avoid waste. This caution may now be costing the company revenue.

Bloomberg columnist Dave Lee highlighted a shifting narrative:

"The AI sparkle that illuminated the market value that more than doubled has diminished... Microsoft stands accused of letting its advantages dwindle."

Furthermore, the perception of Microsoft's exclusivity with OpenAI is fracturing. Reports indicate OpenAI is increasingly seeking compute power from Oracle, Google, and Amazon ("AI polyamory"), diluting the perceived value of Microsoft's foresight. CEO Satya Nadella attempted to pivot investor focus toward "AI Agents" as the new software platform, but the market remained fixated on the infrastructure bottlenecks.

Semiconductors: The Real Winners of the Gold Rush

While software giants battle over narrative and capex, the hardware manufacturers supplying the infrastructure are seeing unprecedented financial gains. South Korean memory giants Samsung and SK Hynix reported massive profit surges, confirming that the physical constraints of the AI boom are far from resolved.

• Samsung: Profits doubled year-over-year to $13.7 billion, with the memory division seeing a 5-fold increase.
• SK Hynix: Operating profit doubled to $13.5 billion, driven by sales of High Bandwidth Memory (HBM) used in AI accelerators.

Both companies signaled that shortages are likely to persist. Citigroup analysts expect DRAM prices to rise by 120% and NAND chips by 90% this year. Unlike the software giants, these hardware firms remain conservative with their own capex, seemingly content to enjoy the high margins driven by supply scarcity.

Sanjeev Rana, head of research at CLSA Securities Korea, summarized the unique nature of this cycle:

"The companies are spending real money on real stuff. We're in uncharted territory in terms of valuations, share prices, the demand cycle. Everything is unprecedented."

Market Implications

The diverging fortunes of Meta and Microsoft suggest a maturation in how markets evaluate AI. It is no longer enough to simply announce AI initiatives; investors demand to see either immediate monetization (as seen in Meta's ad revenue) or uninhibited infrastructure growth to capture future market share.

For now, the "AI bubble" narrative has not popped; rather, it has evolved. The market is validating a "gold rush" mentality where the risk of underspending is viewed as greater than the risk of overspending. As 2026 progresses, the focus will likely remain on whether Microsoft can convert its massive backlog into revenue and whether Meta’s aggressive spending can effectively transition into consumer-facing AI products beyond advertising.