Oracle Raised Its Earnings Target. The Stock Fell Anyway. Here's Why the Math Doesn't Match the Price.

Oracle raised its full-year earnings guidance after posting a 30% revenue surge and a 121% jump in cloud infrastructure income. The stock fell 5.4% today.

The disconnect between Oracle's operating performance and its market price has gone from a curiosity to a gap most investors would call structural. The stock is down roughly 54% from its 52-week high of $329.50 to around $153 today. Not because revenue growth has stalled — it accelerated to 30% in the just-reported first quarter of fiscal 2027. Not because earnings disappointed — non-GAAP EPS beat estimates by more than 10%, and management raised the full-year target.

The reason the stock has been crushed is a single number: $28.5 billion in quarterly capital spending. That pushed free cash flow to negative $5.4 billion, and investors have priced Oracle as a company hemorrhaging cash to fund an AI data center bet that might not pay off. The math tells a more complicated story.

What actually happened

Oracle's fiscal 2027 first quarter (ended August 31) delivered on virtually every top-line and bottom-line metric:

• Total revenue hit a record $19.35 billion, up 30% year over year, beating the consensus estimate of roughly $19.13 billion.
• Cloud Infrastructure (IaaS) revenue surged to $7.39 billion, up 121%. A year ago, this number was $3.35 billion.
• Non-GAAP EPS came in at $1.92 versus estimates of $1.74, a 10% beat.
• Remaining Performance Obligations — essentially the contract backlog — climbed to $664 billion, up $209 billion from a year ago. Oracle booked more than $30 billion in new AI cloud contracts during the quarter alone.

On the back of these results, Oracle reconfirmed its full-year revenue guidance of at least $90 billion and raised its non-GAAP EPS target from $8.05 to $8.10.

That EPS raise is modest on its face — a nickel. But consider what it means in context. The company just spent a quarter of a trillion-dollar-sized company's annual budget on data centers and GPUs, completed a $20 billion equity offering, and still delivered 30% revenue growth and 55% GAAP EPS growth. Raising the full-year earnings target after that kind of quarter is a signal that management sees margins holding, not collapsing.

The cash burn and why the market reacted to it

Here is the real problem that drove the stock down from $329 to $153: capital intensity at a scale most software investors have never seen.

Oracle's quarterly capex of $28.5 billion is nearly four times the same quarter a year ago ($8.5 billion). The full fiscal year 2026 capex ran to approximately $75 billion, against $47 billion in operating cash flow. That left free cash flow at roughly negative $24 billion for the year.

The balance sheet carries $236 billion in total debt and net debt of $88 billion against $36 billion in cash. S&P Global downgraded Oracle's credit rating from BBB to BBB- in July, one notch above junk status, explicitly citing infrastructure spending and exposure to OpenAI — which represents a substantial portion of that $664 billion backlog.

Investors have been through this movie before. Every time a capital-light software business pivots to building physical infrastructure, the question is whether the economics justify the cash burn. Data centers and GPUs are expensive, take time to deploy, and compete against Amazon, Microsoft, and Google who have deeper cash reserves. Oracle is financing this with a mix of equity — the completed $20 billion share sale — and debt, which is a heavy load for a company one notch above junk.

The legitimate bear argument is simple: if cloud infrastructure revenue growth decelerates before capex normalizes, Oracle faces falling free cash flow, rising interest costs on $236 billion of debt, and shareholder dilution from equity raises. The stock price reflects that possibility.

The valuation gap

But here is the divergence between the narrative and the math.

Oracle's raised full-year non-GAAP EPS target of $8.10 implies that management expects to deliver that earnings level despite the massive capex. At the current stock price of approximately $153, that puts the stock at roughly 18.8 times forward non-GAAP earnings. That is not an expensive multiple for a company growing revenue at 30% and cloud infrastructure at triple-digit rates. By comparison, Oracle's price-to-sales ratio sits at 6.1 on trailing revenue, and the EV/EBITDA multiple is 15.6. These are not bargain-bin numbers, but they are not the multiples of a company whose growth is in question either.

The PEG ratio — forward P/E divided by the growth rate — is sitting at 0.51 on trailing metrics. In plain English: the stock is priced as though it's growing at half the rate it actually is.

What the market has done, explicitly or not, is apply a severe discount to Oracle's earnings power because the cash flow looks broken. The stock went from roughly 40 times non-GAAP earnings at its 52-week high to roughly 19 times today. That halving of the multiple is the entire story of Oracle's 54% decline.

The key question is whether the market's discount is permanent or temporary. The earnings raise — even a small one — suggests it is temporary. Management is saying: we're spending a fortune right now, but the revenue and profit are flowing fast enough that the full-year earnings target doesn't just hold, it gets better.

The OpenAI concentration risk

S&P's downgrade highlighted a real structural concern. Oracle's $664 billion backlog is not evenly distributed across thousands of customers. A very small number of clients — with OpenAI being the most prominent — account for a large share. OpenAI is itself an unprofitable startup on an uncertain funding path. If one of these mega-customers scales back, the revenue pipeline shrinks and the entire capex case gets stress-tested.

This is not a hypothetical. It is the reason S&P pulled the rating a notch, and it is the reason many institutional investors have been selling. But concentration risk is not the same as revenue risk — not while the backlog is growing $209 billion per year and Oracle is booking $30 billion in new AI contracts per quarter. The concentration matters more if growth slows, less while it accelerates.

Where the numbers land

Oracle is not a straightforward buy-and-hold story. The debt load is real, the capital intensity is extreme, and the customer concentration is a genuine vulnerability. A deceleration in AI infrastructure demand — whether sector-wide or specific to Oracle's largest clients — would hit this balance sheet harder than it would Amazon or Microsoft.

But the stock's 54% decline has priced in a version of Oracle where the AI infrastructure bet is failing. The first quarter of fiscal 2027 doesn't support that version. Revenue growth is accelerating, not decelerating. Cloud infrastructure more than doubled. The backlog is growing at a pace that suggests demand is not a question mark. And management just raised the earnings target.

The thesis is not that Oracle's capex spending is harmless. It is that the stock's multiple compression — from roughly 40x to roughly 19x non-GAAP earnings — has overreacted to temporary cash flow destruction that is tied to revenue-generating capacity, not operational failure. If cloud infrastructure continues to grow at even a fraction of this quarter's pace, the $8.10 EPS target looks like a floor, not a ceiling. And at $153, the stock is priced to disappoint against those numbers.