Microsoft and Oracle: The AI Expectation Gap

The recent selloff in major software stocks is a classic case of expectations being violently reset. The numbers tell the story of a market in retreat. As of late February, Microsoft stock is down 28% from its all-time high set last fall, while Oracle stock is down 55% from its record high in late 2025. This isn't a minor correction; it's a fundamental reassessment of the AI investment thesis. The core question now is whether this decline is a rational, forward-looking reset or an overreaction to a perceived existential threat.

Two competing narratives are driving the sell-off, each representing a different kind of expectation gap. The first is the fear of obsolescence. As AI tools like Anthropic's Claude Code mature, they allow non-coders to build software-a-trend known as "vibe coding." This raises a direct threat to the core moats of application software firms. As fund manager Nick Evans warns, "We think application software faces an existential threat from AI". In this view, the market is pricing in a future where software companies are rendered redundant by their own clients.

The alternative narrative is one of overspending without payoff. Here, the threat isn't AI making software obsolete, but rather that the massive capital required to build and deploy AI will not yield sufficient returns. This creates a dual threat: companies are spending heavily on AI infrastructure and talent, potentially straining cash flows, while the promised productivity gains and new revenue streams fail to materialize quickly enough to justify the investment. The market is weighing the risk of a costly, unproductive AI arms race.

The depth of the sell-off suggests the consensus may have overestimated the near-term disruption. The selloff is indiscriminate, hitting even the giants like Microsoft. This points to a broader "AI bubble" narrative where the initial euphoria has given way to a more cautious, value-based assessment. The market is now asking if the massive investments in AI will actually protect software moats or simply erode them. The expectation gap is clear: the market was priced for AI as a pure growth enabler, but it's now pricing in the very real risks of disruption and capital inefficiency.

The Whisper Number: What is the Market Actually Pricing In?

The market's current "whisper number" is a cautious one. It's not pricing in a sudden collapse, but a significant slowdown in the growth trajectory that made software stocks kings. The consensus view is that the easy money from AI-driven expansion is fading, replaced by a period of adjustment and scrutiny.

The clearest signal of this reset is the known headwind in core demand. Oracle's own results show a software revenues fell 3% in its latest quarter. This isn't an unexpected shock; it's the baseline reality the market is now factoring in. For a stock like Oracle, which saw its valuation crater by 55% from its late-2025 peak, this decline is a tangible weight. It means the market is no longer assuming software revenue will grow in lockstep with AI hype. The expectation gap has narrowed from "unstoppable growth" to "growth, but at a slower, more uncertain pace."

This sets the stage for the analyst divide. The market is simultaneously pricing in two competing fears, each representing a different kind of expectation gap. The first is the existential threat: that AI tools like Anthropic's Claude Code will make traditional software obsolete by enabling "vibe coding." The second is the capital risk: that companies have overspent on AI and won't see sufficient returns to justify the burn. Both narratives point to a future where the software business model faces pressure, but from opposite directions. The market's current price reflects this uncertainty-it's not betting on either story being fully true, but rather that both contain elements of risk.

The bottom line is that the market is pricing in a potential slowdown. The selloff in Microsoft and Oracle is the market's way of saying the high-growth, high-margin future it was priced for is now in question. The whisper number is one of moderation: expect AI to bring change, but not necessarily the kind of pure, unimpeded growth that fueled the last bull market.

Reality Check: AI as a Revenue Driver vs. a Disruptive Force

The market's fears of AI obsolescence are being tested by the hard numbers of deployment. The evidence shows AI is not a looming threat to be feared, but a powerful engine being actively built into the core of these companies' revenue models. This is the reality check against the panic narrative.

For Microsoft, AI is scaling as a direct revenue product. The company's AI franchise has reached 15 million commercial Microsoft 365 Copilot seats, turning a software subscription into a layered, high-margin service. This isn't just a feature; it's a new pricing lever that drives ARPU expansion. The company's Cloud revenues crossed $50 billion last quarter, with Azure growing 39%, and guidance pointing to continued durable demand. In other words, AI is being monetized at scale, becoming a core pillar of the business rather than a disruptive force.

Oracle's story is one of robust enterprise demand for its AI infrastructure. While software revenue dipped, the company's Oracle Cloud Infrastructure (OCI) revenues grew 68% last quarter, with GPU-related sales surging 177%. More telling is the record remaining performance obligation (RPO) of $523 billion, up 438% year-over-year. This backlog, fueled by commitments from giants like Meta and NVIDIA, signals that enterprise clients are not retreating. They are actively investing in Oracle's AI stack, validating its cloud-neutral strategy and its role as a critical infrastructure provider.

The broader investment landscape confirms this is still an expansion phase, not a collapse. Since the AI boom began, over $100 billion has been invested into AI and AI companies. This capital is flowing into real-world applications, from healthcare streamlining to security testing. The consolidation trend-like Nvidia buying Run:ai or ServiceNow picking up Moveworks-reflects a maturing market where viable use cases are finding fit, not a dying hype train.

The bottom line is that the market's "AI will kill software" narrative is out of step with current execution. Both Microsoft and Oracle are embedding AI deeply into their products and infrastructure, turning it into a revenue driver and a source of enterprise demand. The expectation gap here is between the market's fearful pricing and the reality of AI being productized and scaled. For now, the data suggests AI is being absorbed, not obliterating, the software business model.

Valuation and the Guidance Reset

The market's violent reset has created a clear expectation gap between price and future performance. Despite the steep selloff, the fundamental guidance from both companies points to a durable growth runway, suggesting the current valuations may represent a mispricing driven by short-term fear rather than long-term reality.

The numbers tell a story of resilience. Microsoft's forward P/E of 22.7 is not just cheap for Big Tech-it's the lowest among the 'Magnificent Seven' except Meta. Oracle trades at a forward P/E of 20.2, below its three-year average. Both stocks are priced at attractive multiples relative to the broader market, even as their projected sales growth far outpaces the S&P 500. This disconnect is the core of the guidance reset: the market is pricing in a potential slowdown, but the companies' own forecasts show robust expansion ahead.

Look at the specific guidance. Microsoft's Azure cloud business is on a clear path, with third-quarter fiscal 2026 guidance calling for 37-38% constant-currency growth. This isn't a whisper number; it's a durable, high-single-digit growth target that underscores enterprise demand. Oracle's guidance is equally strong for its cloud engine, with OCI revenues surging 68% last quarter and a record remaining performance obligation of $523 billion providing multi-year visibility. The market is pricing in a potential slowdown in software revenue, but Oracle's software revenues fell 3% in its latest quarter-a known headwind that was already in the print, not a surprise.

The bottom line is that the market's new baseline is a lower one, but it's still a growth baseline. The selloff has priced in existential fears, but the companies' forward guidance shows they are navigating those fears while still scaling AI as a revenue driver. For investors, this creates a setup where the expectation gap has narrowed, but the valuation gap remains wide. The market is pricing in a cautious future, while the guidance reset points to a more resilient one.

Catalysts and What to Watch: The Stargate Factor

The market's current pessimism creates a clear expectation gap. The selloff has priced in fear, but the companies' guidance shows resilience. The catalysts ahead will test whether this fear is justified or overdone. The central event to watch is the Stargate Project, a $500 billion joint venture between Oracle, OpenAI, and SoftBank. This isn't just a headline; it's a massive, real-world bet on AI infrastructure that could validate the entire sector's growth trajectory. The project's execution and its financial impact will be a major signal. If Stargate delivers on its promise of revolutionizing health-tech and building a new AI backbone, it would directly counter the narrative that AI is a costly dead end. If it falters, it could confirm the market's "overspending" fears.

A second key test is any guidance reset from either company. The market is pricing in a potential slowdown, but the companies' own forecasts show robust expansion. Microsoft's guidance for 37-38% constant-currency growth in Azure and Oracle's record $523 billion in remaining performance obligations point to durable demand. However, if either company revises its AI spending plans or its projected returns downward, it would be a direct signal that the capital risk narrative is gaining ground. Any such guidance change would likely widen the expectation gap, as the market had already discounted some of the easy growth.

Finally, track the adoption rate of AI products to gauge monetization. For Microsoft, the scale of its 15 million commercial Microsoft 365 Copilot seats shows AI is being layered into subscriptions. For Oracle, the surge in GPU-related sales by 177% and enterprise commitments signal strong demand for its AI infrastructure. The pace of adoption here is critical. If uptake stalls, it would undermine the "AI as revenue driver" story and reinforce the "software obsolescence" fear. If it accelerates, it would prove AI is being productized, not just hyped.

The Stargate Factor is the ultimate test. This $500 billion project represents a unified bet on AI's future. Its success would validate the sector's growth model and likely reverse the current selloff. Its failure would confirm the market's worst fears about capital inefficiency and disruption. For now, the expectation gap is wide, but the catalysts are clear. Watch Stargate's execution, any guidance shifts, and AI product adoption. These are the metrics that will decide if the market's pricing is a buying opportunity or a rational reset.