Oracle’s 64% Rout Poses a Question No Analyst Can Answer: When Does the Spending Stop?

Oracle's AI infrastructure spending spooks investors as shares drop 39% YTD, trading 64% below record highs despite strong cloud growth and government contracts.

The numbers coming out of Oracle are the kind that usually send a stock soaring. Cloud infrastructure revenue surged 93% last quarter. The company just locked in a $7 billion Pentagon contract. And yet, shares closed Friday at €101.20 — a fresh 52-week low and a 4.08% single-day drop that erased even the modest relief rally sparked by the government deal.

The disconnect is staggering. Oracle has lost 27.11% over the past month and 39.11% since the start of the year. Investors who bought in 12 months ago are sitting on losses approaching 51%. The stock now trades 64% below its September 2025 record high of €280.70, and 36.43% beneath its 200-day moving average.

What’s driving the sell-off isn’t a failure of execution. It’s a crisis of confidence in the financial model underpinning the company’s AI ambitions.

The $95 Billion Question

Oracle is spending aggressively to build out the data centers needed to serve clients like OpenAI. Capital expenditures for fiscal 2026 have been raised to €55.66 billion, and the company plans to raise roughly €40 billion more through debt and equity issuance in the current fiscal year. The total investment in AI infrastructure for fiscal 2027 alone is pegged at around $95 billion.

The market isn’t rewarding that spending. It’s punishing it.

The Pentagon contract — a 10-year deal worth up to $7 billion covering software, cloud services, and consulting for the Department of Defense, Coast Guard, and U.S. intelligence agencies — amounts to roughly 3% of Oracle’s outstanding debt load. Spread over a decade, it’s a drop in the ocean compared to the tens of billions the company is tapping capital markets for. The initial five-year tranche is valued at $3.31 billion, with an extension option bringing the total to nearly $7 billion.

A separate 10-year, $395.8 million contract with the U.S. Office of Personnel Management for an AI-powered HR platform adds another revenue stream, but does little to shift the balance sheet arithmetic.

Technicals Signal Exhaustion, Not a Turnaround

The chart tells a story of extreme overselling. The 14-day relative strength index has fallen to 27.6, deep in oversold territory. Readings below 30 have historically preceded short-term bounces or at least consolidation. The stock is also trading 36.43% below its 200-day average, a level that often attracts value hunters.

Yet the trend remains firmly downward. Friday’s close at €101.20 sits just 0.14% above the 52-week low of €101.06, a level that, if breached, could trigger another wave of selling. No clear catalyst is visible for the week starting July 27 — the next quarterly results aren’t due until mid-September, leaving the stock at the mercy of sector sentiment and technical positioning.

The Analyst Gap That Defies Explanation

The average analyst price target for Oracle stands at €219.10, implying upside of 116.5% from current levels. That kind of gap between institutional expectations and market reality is rare. Either the analysts are fundamentally wrong about the company’s prospects, or the market is dramatically overpricing the balance sheet risk.

The core tension is simple: Oracle’s cloud revenue grew 47% last quarter, with the infrastructure segment leading the charge at 93%. Its legacy software business, meanwhile, shrank 2% — a clear sign of the ongoing transformation. But the cash flow from that growth isn’t keeping pace with the capital required to sustain it. The company has cut roughly 21,000 jobs to help fund the AI push, a move that protects margins but also underscores the operational strain.

A Stability Anchor That Can’t Carry the Weight

The Pentagon contract does serve one important function: it reinforces Oracle’s role as a trusted technology provider for government clients, a segment insulated from the volatility of the commercial cloud market. That’s a stability anchor, not a growth driver. It won’t move the needle on a $55.66 billion annual capex plan.

For investors, the next real test comes September 14, when Oracle reports quarterly earnings. The market will be looking for evidence that the company’s massive AI contract backlog is converting into cash flow fast enough to justify the spending. Until then, the stock remains a bet on two competing narratives: that the sell-off is a buying opportunity in an oversold name, or that the balance sheet risk is real and the 64% decline still has room to run.

The answer hinges on one unresolved question: Can Oracle raise the €40 billion it needs without diluting existing shareholders to the breaking point?