Oracle Corporation [ORCL]
Scope limitation, stated first. This is a Tier-2 memo. There is no .xlsx workbook. Every figure in the memo was computed in reproducible Python against cached SEC XBRL facts and cached Alpaca bars, and this file is the audit trail for those computations. There is therefore no balance check, because there is no forecast balance sheet — tie-out is to the filed statements only, which is what a Tier-2 scope discharges.
Reproduction scripts (committed under work/): fetch_facts.py, facts.py, dump.py, bs.py,
filings.py, pr.py, mentions.py, mult.py, anchor.py, calc.py, calc2.py.
FY2026 10-K (accession 0001193125-26-277521, filed 2026-06-22), all figures $m:
| Line | Model | Filed | Δ |
|---|---|---|---|
| Total revenues FY2026 | 67,357 | 67,357 | 0 |
| Total revenues FY2025 | 57,399 | 57,399 | 0 |
| Total revenues FY2024 | 52,961 | 52,961 | 0 |
| Operating income FY2026 | 20,606 | 20,606 | 0 |
| Net income FY2026 | 17,087 | 17,087 | 0 |
| Net income available to common | 16,984 | 16,984 | 0 |
| Diluted EPS FY2026 | 5.83 | 5.83 | 0 |
| Diluted shares FY2026 | 2,914 | 2,914 | 0 |
| Operating cash flow FY2026 | 31,977 | 31,977 | 0 |
| Capex FY2026 | 55,663 | 55,663 | 0 |
| Interest expense FY2026 | 4,599 | 4,599 | 0 |
| Income tax expense FY2026 | 2,467 | 2,467 | 0 |
| Gain (loss) on investments FY2026 | 2,811 | 2,811 | 0 |
| Cash and equivalents 2026-05-31 | 31,289 | 31,289 | 0 |
| Total borrowings 2026-05-31 | 129,541 | 129,541 (also stated as "$129.5 billion" in Item 1A) | 0 |
| Preferred stock 2026-05-31 | 4,954 | 4,954 | 0 |
| Gross PP&E 2026-05-31 | 122,651 | 122,651 | 0 |
| Net PP&E 2026-05-31 | 99,957 | 99,957 | 0 |
| RPO 2026-05-31 | 638,000 | "$638 billion" | 0 |
Cross-check required by the brief: 17,087 ÷ 2,914 = 5.864 against a filed diluted EPS of
5.83; using NetIncomeLossAvailableToCommonStockholdersBasic of 16,984 the quotient is
5.829. Ties.
| Figure | Formula | Result |
|---|---|---|
| Q4 FY2026 revenue | FY2026 − (Q1 + Q2 + Q3) = 67,357 − (14,926 + 16,058 + 17,190) | 19,183 (press release: "$19.2 billion") |
| Q4 FY2025 revenue | 57,399 − (13,307 + 14,059 + 14,130) | 15,903 |
| Latest-quarter YoY | 19,183 / 15,903 − 1 | +20.6% (release: "up 21%") |
| 3-year revenue CAGR | (67,357 / 49,954)^(1/3) − 1 | 10.5% |
| 5-year revenue CAGR | (67,357 / 40,479)^(1/5) − 1 | 10.7% |
| FY2027 guided growth | 90,000 / 67,357 − 1 | +33.6% |
| Capex intensity FY2026 | 55,663 / 67,357 | 82.6% |
| Free cash flow FY2026 | 31,977 − 55,663 | −23,686 (release: "negative $23.7 billion") |
| FCF − net income | −23,686 − 17,087 | −40,773 |
| Liquid assets | 31,289 + 605 | 31,894 |
| Net debt | 129,541 + 4,954 − 31,894 | 102,601 |
| Net debt incl. finance leases | 102,601 + 7,701 | 110,302 |
| Enterprise value | 119.95 × 2,880.5 + 102,601 | 448,077 |
| EV / TTM revenue | 448,077 / 67,357 | 6.65x |
| EV / TTM EBIT | 448,077 / 20,606 | 21.75x |
| OCI share of revenue | 18.1 / 67.4 | 26.9% |
| OCI share of revenue growth | (18.1 − 10.2) / (67.4 − 57.4) | 79.1% |
| Total cloud share of growth | (34.0 − 24.5) / 9.96 | 95.3% |
| RPO next-12-month recognition | 0.12 × 638,000 | 76,560 |
| — as % of FY2027 revenue guide | 76,560 / 90,000 | 85.1% |
| — as % of FY2026 revenue | 76,560 / 67,357 | 113.7% |
| RPO weighted-average duration | 0.12×0.5 + 0.34×2.0 + 0.34×4.0 + 0.20×6.5 | ≈3.4 years |
| Q1-FY2026 RPO increase | 455,000 − 138,000 | 317,000 |
| — as % of today's RPO | 317,000 / 638,000 | 49.7% |
| — per counterparty (3 stated) | 317,000 / 3 | ≈105,667 = 1.84x FY2025 revenue |
| Ampere/Bloom EPS impact | (2,811 − (−278)) × (1 − 0.126) / 2,914 | $0.93 |
| Clean FY2026 GAAP EPS | 5.83 − 0.93 | $4.90 |
| Effective tax rate FY2026 | 2,467 / (17,087 + 2,467) | 12.6% |
| Q4 depreciation annualised | (7,623 − 5,208) × 4 | 9,660 |
| Interest coverage | (20,606 + 7,623) / 4,599 | 6.1x (covenant ≥3.0x) |
| ATM dilution | 20,000 / 119.95 / 2,880.5 | +5.8% |
| Lease commitments, undiscounted | 41,867 + 11,460 | 53,327 |
| Unbilled share of RPO | 1 − 15,395 / 638,000 | 97.6% |
beta = cov(ORCL daily returns, SPY daily returns) / var(SPY) over 252 sessions = 1.85
Ke = 4.2% + 1.85 × 5.5% = 14.4%
w_e = 345,516 / (345,516 + 129,541 + 7,701) = 71.5%
WACC = 0.715 × 14.4% + 0.285 × 5.5% × (1 − 0.15) = 11.6%
realised vol = stdev(daily returns) × sqrt(252) = 65.8%
Stated assumptions, not derived: risk-free 4.2%, ERP 5.5%, pre-tax cost of debt 5.5%, forward tax 15%. The cost-of-debt estimate is above Oracle's FY2026 realised cash cost of ~4.2% and is deliberately conservative for marginal issuance.
Daily series, 1,505 sessions from 2020-07-27 to 2026-07-28 (Alpaca IEX daily bars, split-adjusted).
For each session: the TTM metric is the value known at that date, built by summing the last four quarterly XBRL periods and lagged to the filing date so the series is not forward-looking. Missing fourth quarters are inferred as (annual − the three tagged quarters) and carry the annual filing's filed-date. Share count and net debt are held at today's verified values.
Known bias, and its direction. Holding today's $102.6bn net debt constant across a period when Oracle's net debt was materially smaller overstates historic EV, therefore overstates historic multiples, therefore understates today's percentile. The EV percentiles (1st on sales, 17th on EBIT) are floors. The P/E series is unaffected by net debt and returns the 12th percentile, which confirms the direction independently. This is disclosed rather than corrected because correcting it properly requires a quarterly net-debt series that the scanner does not build.
Source: reports/scan_v3/_scan.json, 3,980 names with status == "OK".
Filter, applied for ORCL alone: revenue_cagr_demonstrated in [12.0%, 20.0%]; sic2 == "73";
market_cap > $5bn; 0 < ev_ebit < 200. n = 26.
Resulting set growth range 12.1%–19.9% — it brackets the subject's estimated exit-year
growth, which is the condition valuation.md imposes. Market-cap floor of $5bn was set explicitly
to avoid the logged defect of anchors built from sub-$100m shells; the realised minimum was $6.0bn.
Output: p25 16.8x, median 27.1x, p75 37.2x.
ORCL_Trade_Construction.md §6
rather than fabricated.