Oracle Stock Hasn't Been This Cheap in 3 Years -- But Is It a Buy Right Now?

$553 billion remaining performance obligations (RPO) at Oracle signals massive demand for AI data-center capacity, while fiscal Q3 revenue was $17.2B (+22% YoY) and Oracle Cloud Infrastructure revenue jumped 84% to $4.9B. The stock has fallen ~57% since last September, and the balance sheet shows ~ $124B in long-term debt; roughly $300B of the RPO was reported tied to OpenAI, which has only ~$25B in annualized revenue, raising counterparty/fulfillment risk. Given the combination of rapid cloud demand and concentrated, potentially unreliable customer commitments plus high leverage, the article recommends a cautious stance for investors until market conditions stabilize.

Analysis

The market is pricing a concentrated counterparty and capital structure risk into Oracle far more than a pure cloud-capacity story. When a material share of contracted demand sits with a single, cash-burning customer, the real exposure is counterparty credit + execution of Oracle’s capex program: a 10-20% shortfall in fulfilled RPO over 12–36 months would cut forward free cash flow by multiples of near-term interest expense and could force asset-class level impairment cycles for data-center buildouts.

Second-order supply effects matter: if Oracle cannot monetize contracted GPU demand, that capacity will either sit idle (stranding capex) or be absorbed by rivals — a scenario that re-prices GPU supplier cadence (NVDA/AMD) and shortens their lead time for enterprise sales. Conversely, a credible refinancing or prepayment by large AI customers would de-risk Oracle’s balance sheet quickly and re-rate it higher; credit-market signals (spread movements, rating agency commentary) will likely lead equity moves before revenue prints.

Timing splits the trade: near-term (days–weeks) the dominant risks are funding news and short-term macro risk from geopolitics that dampens AI spend; medium-term (3–12 months) the tests are customer payment discipline and utilization metrics from Oracle’s new farms; long-term (2–4 years) the question is structural ROI on automated RDMA-heavy builds versus hyperscaler in-house or hybrid alternatives. Watch GPU inventory reads, Oracle bond yields vs IG peers, and any large-scale customer prepayments as binary catalysts.

Key Decisions for Investors

• Pair trade (3–9 months): Short ORCL equity (target entry $135–$155) and long NVDA (target entry $800–$950) 1:0.25 size — rationale: hedge fundable bankruptcy/credit scare in ORCL with sustained secular GPU demand benefiting NVDA. Risk/reward: downside in ORCL >30% if counterparty fails vs NVDA upside ~25–60% if GPU demand centralizes; cap losses if macro reverses and both rally.
• Credit play (6–18 months): Buy protection via ORCL senior bonds (sell-or-repurchase strategy) or buy ORCL CDS notional $50–100mm equivalent — only if CDS spreads widen >150bps from current levels. Risk/reward: asymmetric — limited carry vs large payoff on material covenant/default event; manage liquidity and duration to expected capex/refi windows.
• Options hedge (0–6 months): Buy ORCL 6–9 month puts (e.g., Sep/Dec) OTM ~10–15% as tail insurance against a missed RPO realization or downgrade; fund via selling covered NVDA call spreads (3–6 months) to monetize premium. Risk/reward: controlled max loss = premium; payoff >3x if ORCL downside accelerates while NVDA remains bid.
• Event arbitrage (weeks–months): Monitor large-customer funding announcements (OpenAI or equivalents). If a credible multi-year prepayment or committed facility emerges, rapidly flip to long ORCL equity for 3–12 month capture — target 20–40% upside, cut if covenant terms opaque. Risk/reward: binary catalyst with high information asymmetry; require legal/contractual clarity before sizing.
• Selective supplier long (6–18 months): Accumulate AMD (and selectively NVDA exposure via call spreads) on GPU shortage fears easing — entries: AMD below $120, NVDA $850–950 call spreads. Risk/reward: benefits from GPU cycle normalization and re-acceleration of enterprise orders; downside if demand collapses broadly with recession.