This Tech Stock Is 1 of the Most Shorted Stocks of 2026: Is Now the Time to Buy?

Oracle shares are down ~53% from their Sept. 10, 2025 all-time high of $324 to about $153, despite a large backlog of $638B and a reported $300B OpenAI cloud deal. The pullback is attributed to heavy AI/data-center capex—$21B last year, projected $55B in fiscal 2026—driving negative free cash flow of $23.7B and raising funding needs (another $40B planned in fiscal 2027 via debt/equity); leverage is also elevated with $167B debt and a 388% debt-to-equity ratio. Valuation looks cheaper (P/E 25, forward P/E 18, PEG 0.85), and analysts are broadly supportive (82% “buy,” median target $241 implying ~57% upside), but execution and counterparty risk (OpenAI) keep the outlook cautious.

Analysis

Oracle is increasingly trading like a leveraged financing vehicle for AI capacity rather than a software compounder. That matters because the equity is now hostage to two spreads at once: incremental debt cost versus asset yield, and the market’s willingness to finance capex before the new data-center revenue is visible in cash flow. If credit investors demand wider spreads or equity dilution becomes more likely, the multiple can compress again even without a fundamental miss.

The second-order effect is a rotation within AI beneficiaries. Balance-sheet-light winners such as MSFT and META are better insulated if the market keeps rewarding disciplined free-cash-flow conversion, while NVDA and AMD still benefit from capex flow but are less exposed to Oracle-specific financing risk. By contrast, SMCI remains vulnerable to any “AI spend is not yet self-funding” narrative because its valuation is tethered to the same hyperscale buildout cycle and to customers’ ability to keep funding purchases.

The contrarian point is that the selloff may already be discounting peak fear on dilution and debt, while underpricing backlog conversion optionality over 6-18 months. The key reversal catalyst is not another headline contract; it is proof that capex growth slows faster than revenue growth flattens, allowing cash flow to inflect before the market forces a recap. If Oracle can show a credible path to positive FCF and avoid punitive financing terms, the short can unwind quickly because positioning is crowded.

Near term, expect the stock to trade on credit-market tone and any guidance around capex/funding rather than backlog rhetoric. The thesis is falsified if Oracle secures large financing at tight spreads, trims capex without hurting growth, or posts sequential FCF improvement before the next quarterly reset.

Key Decisions for Investors

• Short ORCL on any relief rally over the next 1-3 weeks; use a 3-6 month horizon because the catalyst is financing and FCF, not backlog headlines. Risk/reward is attractive while the market is still debating dilution, but cover if management shows capex moderation and positive FCF inflection.
• Pair trade: long MSFT / short ORCL for 1-3 months. MSFT has the cleaner AI monetization path and less balance-sheet strain; ORCL carries execution and funding risk. Falsify if ORCL financing comes materially cheaper than expected or MSFT guidance softens on cloud demand.
• Trade ORCL credit versus equity: prefer to watch ORCL bonds or CDS for widening as an early warning before the stock reprices. If spreads keep widening while equity holds, that is an alert to add to the short; if spreads tighten, trim exposure.
• For AI hardware exposure, prefer NVDA over SMCI on any AI infrastructure pullback. NVDA retains pricing power and ecosystem control, while SMCI is more vulnerable to a capex air-pocket and customer payment-risk narrative over the next 1-2 quarters.