Oracle stock falls 3%: why this analyst still sees upside

Oracle shares fell about 3.4% after reports that the company canceled a large Nvidia hardware order for 300 to 400 GB300 NVL72 racks, a deal estimated at $1.1 billion to $1.4 billion. The move raises near-term questions around Oracle's AI infrastructure spending, though analysts still view its long-term AI position as intact. The news is likely to pressure Oracle stock more than the broader market.

Analysis

This read-through is less about one canceled order and more about Oracle signaling that it may be rebalancing the mix between owned/committed infrastructure and customer-backed capacity. If the company is pulling back on speculative front-end supply, that can be a mild negative for near-term revenue recognition at hardware-adjacent vendors, but it also reduces execution risk if demand timing is lumpy. The market is likely extrapolating too much from a single procurement decision into AI capex deceleration, when the more important question is whether Oracle is shifting toward higher-return, tighter-committed deployments.

The immediate losers are SMCI first, then NVDA at the margin: SMCI loses the cleanest near-term order flow, while NVDA faces a second-order headline risk that hyperscaler-like spending is becoming less linear. But the bigger underappreciated effect may be on pricing power across the rack supply chain—if one large buyer pauses, lead times and component mix can reset quickly, pressuring OEMs with high fixed-cost leverage more than chip vendors. That makes this a positioning event for SMCI far more than a fundamental impairment to NVDA’s broader AI demand curve.

Contrarian take: the selloff in ORCL may be overdone if investors are equating capex discipline with weakening AI intent. In a year where many firms are racing to secure compute, the more durable winners are those that can delay purchases until utilization is visible and financing terms improve. If Oracle later recommits via customer prepayments, lease structures, or a different architecture, the current headline will look like timing optimization rather than demand destruction.

Near term, the key catalyst is whether this is followed by additional cancellations or by new, more structured AI infrastructure commitments over the next 1-2 quarters. If no follow-on negatives emerge, the stock impact should mean-revert quickly; if it does cascade, SMCI is the higher-beta expression of that downside because its order book is more sentiment-sensitive and less diversified. NVDA should be treated as a relative-not-absolute short only if broader capex checks deteriorate.

Key Decisions for Investors

• Short SMCI into any relief bounce over the next 1-3 sessions; use a tight stop above the post-news gap high. Risk/reward is attractive because the stock is most exposed to order timing and sentiment, while upside requires proof of replacement demand.
• Relative-value: long NVDA / short SMCI for 2-6 weeks. If AI spend is intact but being re-phased, NVDA should hold up better on broader ecosystem demand while SMCI absorbs the order-specific hit.
• Do not short ORCL aggressively on this headline alone; instead wait for the next quarterly update or management commentary. The cleaner trade is a tactical long on ORCL weakness only if the market confirms this was capex optimization rather than demand deterioration.
• Buy short-dated ORCL puts only if the stock fails to reclaim the event-day range within 2-3 trading days. That setup offers better asymmetry than chasing the initial move, with catalyst risk tied to follow-through headlines.
• For multi-week positioning, prefer a basket short in hardware-sensitive AI names versus a broad short of AI infrastructure. The risk is that one reordered deal becomes a supply-chain shuffle rather than a genuine demand miss.