Is Oracle's 29% crash the best AI bargain on Wall Street right now?

Oracle’s brutal 2026 selloff is forcing Wall Street to answer a harder question than the usual buy-the-dip reflex.

Shares of the cloud and database giant were still down about 29% for the year as of March 31, even after a brief rebound.

Yet the collapse has not produced a consensus.

Instead, it has split analysts into two camps: those who see a rare chance to buy a major AI build-out at a discount, and those who think investors are finally pricing in just how expensive that ambition may be.

Oracle stock: What went wrong?

Oracle’s fall has not been driven by a collapse in demand.

In March, the company reported fiscal third-quarter revenue of $17.2 billion (approx. €15 billion) up about 22% from a year earlier and ahead of Wall Street expectations.

Oracle lifted its fiscal 2027 revenue target to $90 billion (approx. €78.5 billion).

Its remaining performance obligations climbed to $553 billion (approx. €482.4 billion) in the quarter after hitting $455 billion (approx. €396.9 billion) in the fiscal first quarter of 2026.

On the surface, those are not the numbers of a company losing momentum.

The problem is what investors now have to stomach in order to get that growth.

Oracle has said it expects fiscal 2026 capital expenditure of $50 billion (approx. €43.6 billion), more than double the prior year, and in February, it said it planned to raise $50 billion (approx. €43.6 billion) through debt and equity to build more cloud capacity.

By December 2025, Oracle already had around $100 billion (approx. €87.2 billion) in debt, and by April, investors were still scrutinizing the company’s swelling balance sheet and cash burn.

The bull case

For the bulls, the selloff has become an opportunity.

JPMorgan upgraded Oracle to Overweight in March with a $210 price target, arguing that the stock’s collapse had changed the setup more than the business itself.

In Mark Murphy’s phrasing, the “drastic” or “severe” selloff improved the risk-reward because investor sentiment had swung from confidence to pessimism.

That matters because Oracle is still one of the few companies trying to build AI infrastructure at a real scale outside the cloud leaders.

The bullish argument also leans on backlog. Oracle’s contract book has exploded, with RPO jumping from $455 billion $455 billion (approx. €396.9 billion) in September 2025 to $553 billion $553 billion (approx. €482.4 billion) by March 2026.

The bear case

The bear case is simpler and probably easier for retail investors to grasp.

Even if Oracle is right on demand, the path from here to attractive shareholder returns could be long and messy.

Melius analyst Ben Reitzes downgraded the stock to Hold with a $160 target in February, arguing that a company with no meaningful free cash flow until the 2030s should be valued more like an infrastructure builder than a software company.

That is a sharp warning because software stocks usually command premium multiples precisely because they are asset-light and cash generative.