Oracle’s shares have fallen roughly 50% over the past 12 months amid investor concerns that elevated AI spending won’t pay off. Capex surged to $55.7B for FY ending May 31 (vs. $21.2B a year earlier) while long-term liabilities rose to $176.9B from $114.7B, and the company projects raising $40B for current fiscal year AI cloud investment via debt and equity, including a $300B cloud deal with OpenAI. Despite the pullback, the article argues the stock is “cheap” on a forward P/E below 16 versus ~22 for the S&P 500, suggesting a contrarian upside case but with near-term risk.
Oracle is being re-rated less like a software compounder and more like a leveraged infrastructure financier. The market is correctly discounting that reported earnings can overstate economic value when incremental growth is funded by heavy capex and rising liabilities; in that setup, the key variable is not P/E but free-cash-flow conversion and refinancing capacity. If AI demand remains strong, the first-order beneficiaries are the picks-and-shovels names and financing providers, not necessarily ORCL shareholders.
The second-order winner is NVDA and adjacent AI infrastructure suppliers: Oracle’s spend cadence implies continued demand for accelerators, networking, storage, and datacenter buildout even if Oracle’s own equity multiple compresses. That said, the risk is that Oracle becomes a demand-deserting customer if financing costs rise or utilization disappoints, which would turn its buildout into a slower-moving balance-sheet problem over 6-18 months. The bond market will likely be the cleaner early warning signal than the stock price; spread widening would tell you the equity story is being masked by accounting growth.
Near term, the stock can bounce if management reiterates backlog conversion and defends the capex plan, but the 1-3 month catalyst path is fragile: any moderation in cloud spend, slower OpenAI-related ramp, or softer commentary on financing would force multiple compression. The consensus is missing that a low forward P/E is not cheap if the denominator is being propped up by aggressive investment and leverage. Contrarian bulls are right that the stock can be oversold, but only if utilization and customer concentration risk prove manageable.
The cleanest trade is conditional: if Oracle does not cut FY capex or improve FCF guidance, the setup favors being short ORCL on rallies rather than buying the dip. If you want exposure to AI capex without the balance-sheet risk, the relative long is NVDA versus ORCL. Falsifier: a clear downward revision to capex or a material improvement in operating cash flow would invalidate the bearish thesis and force a cover.
AI-powered research, real-time alerts, and portfolio analytics for institutional investors.
Request TrialOverall Sentiment
mildly negative
Sentiment Score
-0.35
Ticker Sentiment