Oracle is already selling off on news of a capital-heavy plan from Ellison, which means years of spending financed before any revenue shows up. With the Fed hiking again and every investor bracing for 'higher for longer' rates, borrowing costs for big exp

Verdict: Oracle's bear case is credible but the trade isn't armed — wait

The bear thesis has real teeth: Oracle's fiscal 2026 (ended May 31) shows free cash flow of negative $23.7B, long-term debt of $122.3B against just $42.5B of equity, and ownership filings for the period ended June 30, 2026 (filing deadline passed) show net insider open-market selling of about $39.6 million across 21 holders. But the fundamentals push back hard — revenue grew 17.3% to $67.4B, net margin rose 3.7 points to 25.4%, diluted EPS grew 34.3%, and the dividend was raised 11.1% in 2026 — so this looks like a booming franchise making a levered bet, not a broken business. The biggest practical problem is trade readiness: the idea is bearish, yet the configured strategy's entry rules are a long trigger, and none of the three conditions are met. Price at $143.16 sits about $6.00 below the 50-day EMA at $149.16, the trend-strength reading of 15.8 is below the required 20, and the MACD histogram is still positive at +0.83. The backtest evidence is solid but scope-limited — a 10.0% return on 37 trades over 60 months with a 5.9% maximum drawdown, and a stronger 24-month window — while the 12-month window could not be evaluated because a required momentum indicator was unavailable for the recent data. On balance, wait: the entry conditions are not armed, and the most recent ownership filing predates the Ellison-plan selloff, so it may not reflect current insider posture. The single fact that would flip the verdict is all three long-entry conditions confirming while price holds above the $138.59 support.

Trade now: ORCL long setup is close but not armed

The entry setup on ORCL is a long trigger: price above the 50-day EMA, the 14-day trend-strength reading above 20, and the MACD histogram below zero, with price above the nearest support at $138.59. At the latest close of $143.16, none of those are fully in place. Price sits about $6.00 below the 50-day EMA at $149.16, so the trend filter is the closest miss; the trend-strength reading is 15.8 and needs to clear 20; and the MACD histogram is still positive at +0.83 when the setup requires it to turn negative. Two of the three conditions remain pending, so the honest call today is wait. Worth flagging: the idea's thesis is bearish on Oracle's capital-heavy spending plan, but the mechanical strategy as configured trades ORCL long, with its own stops and targets doing the risk work. The exits are explicit: a hard stop at a 2.4% loss on the position, a take-profit at 4.8%, a stop if price closes at or below the second support level of $138.59, and a profit-taking exit near the first resistance at $149.07. Against the 2.4% stop, the 4.8% profit target gives an effective reward-to-risk of roughly 2-to-1. The evidence here is a completed backtest, not a live track record. Over 60 months of daily bars the strategy returned 10.0% on 37 trades with a 45.9% win rate and a 5.9% maximum drawdown; the last 24 months were stronger, returning 14.7% on 15 trades with a 73.3% win rate and only a 2.2% drawdown. Those are the numbers backing the rules, and the recent regime clearly flattered them. Concretely, waiting means: do nothing until ORCL closes above $149.16 (the 50-day EMA), the trend-strength reading rises above 20, and the MACD histogram turns below zero, while price holds above $138.59. If ORCL breaks below $138.59 before all three conditions align, the setup is off and should be reassessed rather than chased.

Oracle's Bill Is Coming Due Before the Revenue Does

The bear case starts with cash quality, and Oracle's is deteriorating fast. In fiscal 2026 (ended May 31), the company generated $32.0B of operating cash flow but spent $55.7B on capital expenditures, leaving free cash flow of negative $23.7B — the bottom of its peer group, at the 0th percentile among 791 information-technology names. Long-term debt stands at $122.3B against just $42.5B of shareholder equity, and interest expense ran $4.6B in the year. This is exactly the setup the idea describes: a capital-heavy plan being financed now, years before the promised revenue arrives. The news flow backs the thesis — Oracle dropped on the Ellison spending plan per the Yahoo Finance movers segment on September 14, 2026, and CNBC's September 16 coverage of the Fed's unanimous 12-0 hike has investors explicitly bracing for 'higher for longer' rates. For a borrower with a $122B debt stack and negative free cash flow, every rate-hike cycle makes that financing more punishing. The ownership picture leans the same way. Across the filing period that closed June 30, 2026 — a disclosure whose reporting deadline has passed, so it reflects holdings as of that date, not today — 21 reporting insiders at Oracle showed net open-market selling of roughly $39.6 million on 8.3 million shares held. Compare that with Berkshire Hathaway, the second name in this idea, where insiders show net open-market buying of about $2.0 million. Management selling while the company borrows to build is not disqualifying on its own, but it does not support the 'bill gets paid easily' narrative. The most recent fundamentals also show strain beneath a strong headline. Revenue grew 17.3% to $67.4B and net margin improved 3.7 points to 25.4%, but operating margin slipped 0.2 points to 30.6% — meaningfully worse than the 39.5% Oracle posted in fiscal 2013 — and return on equity fell by 20.7 points, from 60.8% to 40.2%. Diluted share count rose 2.6% year over year to 2.88 billion, a reversal of…

Scores

• Conviction score breakdown: 52
• Thesis support: 62
• Trade readiness: 25
• Risk quality: 55
• Backtest evidence: 68
• Fundamentals trend: 48

Watch items

• ORCL — ADX (14)
• ORCL — MACD (12,26,9) histogram
• ORCL — Close vs 50-day EMA
• ORCL — Close vs support level 2
• ORCL — Insider net open-market activity
• ORCL — Return on equity (FY, YoY)