Down 55% From Its High, Is Oracle a Buy?

Oracle’s fiscal 2026 revenue rose 17% to $67B and cloud infrastructure revenue jumped 75% YoY, alongside $638B of backlog for strong visibility. However, aggressive AI data-center build-outs pushed free cash flow sharply negative and increased debt, driving much of the ~55% stock decline from its $345 peak. With the forward P/E around 17 (PEG < 1) and the shares near $145, the article frames valuation as an opportunity, but flags risk if build-outs or demand weaken.

Analysis

The market is likely punishing ORCL less for growth and more for financing the growth: once a software franchise starts behaving like a utility, the relevant question becomes whether incremental AI capacity earns returns faster than the balance sheet degrades. That makes this a margin-of-safety story, not a simple revenue story. In the next 1-2 quarters, the stock should trade off FCF credibility and debt trajectory more than backlog headlines; if capex stays elevated while cash generation lags, the multiple can keep compressing even with strong top-line growth.

Second-order winners are the infrastructure suppliers that can sell into Oracle’s buildout regardless of whether ORCL’s own equity rerates. NVDA still benefits from any acceleration in GPU deployment, but the bigger beneficiary may be power, networking, and data-center ecosystem names because those budgets are harder to cut once committed. The competitive loser set is the hyperscaler cohort if Oracle is proving that price-sensitive AI workloads can be monetized outside the big three, but that only matters if utilization and renewal rates validate the backlog over the next 6-18 months.

The contrarian view is that the street may be overfocusing on negative free cash flow as if it were terminal, when for a contracted cloud business it can be temporary if capacity turns productive fast enough. The key falsifier is not the next revenue print; it is whether operating cash flow plus backlog conversion starts offsetting capex by the next couple of quarters. If not, ORCL remains a classic "good business, bad capital allocation timing" setup, where the equity can underperform despite solid headline growth.

Key Decisions for Investors

• Do not chase ORCL on valuation alone; wait 1-2 quarters for evidence that capex intensity is peaking and FCF is inflecting positive. If that does not happen, treat rallies as sellable into strength.
• Pair trade: long MSFT or AMZN, short ORCL over 3-6 months if the market keeps rewarding balance-sheet optionality and punishing levered AI infrastructure spend. Target is relative underperformance if ORCL’s FCF remains negative.
• Watch NVDA as a secondary beneficiary of Oracle’s buildout; if Oracle commentary confirms continued GPU deployment, that supports NVDA demand durability over the next 1-2 quarters.
• Set a tactical alert on ORCL credit metrics/spreads and net debt trend; widening spreads or another step-up in capex would be the cleanest signal that the equity de-rating has further to go.
• Contrarian buy condition only: accumulate ORCL on any sharp pullback if the next quarterly update shows backlog conversion improving faster than expected and capex growth decelerating; otherwise, stay neutral.