883 — CNOOC Limited

Deep-value analysis — CNOOC Limited (HKEX: 883 / 0883.HK, SSE: 600938)

Analysis date: 09/11/2026 · Reference price: … (yfinance/data pack price from 09/11/2026; the last available close, 09/10/2026, was … — the margin-of-safety figures below are therefore ~2.6% generous) · Market cap: … = 989.1 bn CNY · Shares: 47,529.95 mil (H + A) · Reporting currency: CNY; trading currency: HKD; rate used … = 0.8536 CNY (HKDCNY=X, 09/11/2026)

Primary sources: 2025 Annual Report (HKEXnews, 04/09/2026, audited by Ernst & Young), 2026 Interim Report (HKEXnews, 09/08/2026, unaudited), the 2023 and 2024 annual reports, all downloaded locally in SEC-Filings/883/. Deterministic data pack: rapoarte/deep/data-pack-0883.HK-20260911.md. Monte Carlo simulation: rapoarte/deep/mc-883-20260911.json.

A currency caveat: the issuer reports in CNY and trades in HKD. All returns relative to market cap in this report are calculated after converting market cap into CNY (989.1 bn), not by dividing a CNY flow directly by an HKD market cap. yfinance’s P/B and P/E figures (1.158 and 7.15) hold up under verification because the provider converts book value into HKD; yfinance’s reported EV (903.1 bn) does not hold up and was not used.

Executive summary (1 page: thesis, estimated value, verdict)

Thesis in brief. CNOOC Limited is the listed arm of China’s offshore oil monopoly: 62.13% of shares are held by CNOOC Group (state), the company has the exclusive right to operate on the Chinese continental shelf, produces at an all-in cost of …/BOE (2025) — among the lowest in the global industry — and has grown production for seven consecutive years, to 777.3 million net BOE in 2025 (… vs 726.8 in 2024) and 398.7 million BOE in H1 2026 alone (…). The balance sheet is an anomaly in a sector with chronic leverage: as of 06/30/2026 the company had 161,700 mil CNY of cash plus 152,678 mil CNY of term deposits, against interest-bearing debt of 68,088 mil CNY — that is, net cash of 246,290 mil CNY, equivalent to almost a quarter of market cap. The gearing ratio (interest-bearing debt / total capital) is 8.0%, down from 10.9%.

What is not in order. Three things. First: the price of oil. H1 2026 was exceptional — realized price …/barrel versus … for all of 2025 and … in H1 2025 — and it was exceptional for a geopolitical reason (Brent hit … in March 2026, fell to 72.9 in June, and is at 104-… today). Any valuation base calculated on TTM figures is procyclical. Second: net cash grows, but does not reach the minority shareholder. The stated policy is a payout of ≥45% of net profit for 2025-2027; the actual total return realized over 2022-2025 was 49.1% of cumulative net profit. The rest accumulates — net cash has risen from ~108,000 mil CNY (end 2024) to … (end 2025) and 246,290 (06/30/2026), with no meaningful buybacks (661 mil CNY in 2024, ZERO in 2025) and no special dividend since 2022. Third: governance. Zero shares held by the executive directors, zero options, total board remuneration 5.89 mil CNY per year, and the chairman (Zhang Chuanjiang, July 2025) and CEO (Huang Yongzhang, March 2026) were both appointed in the last 14 months, from outside the company, through state rotation.

The FCF bridge and the valuation base. For 2025: CFO … − capex … = 97,487 “simple FCF”; minus interest paid reported in financing (…) and lease principal (…) = FCFE … mil CNY. The divergence between the two definitions is 6.7%, below the 10% threshold — at CNOOC, the placement of interest does not change the conclusion, since debt is small and cheap (average cost 4.1%). However, the base I ran the DCF on is neither of these two, but one normalized to a mid-cycle price of …/barrel: … mil CNY, derived starting from TTM (122,985 FCFE, at an average realized price of …/barrel) and corrected for working capital and tax-timing effects from H1. The reason is explicit: TTM sits on a price peak, and 2025 sits on a trough.

The estimated value. Five triangulated models give a wide range, from … to … margin of safety versus …. The median is …, i.e. median intrinsic value …/share — after net cash was cut by λ = 0.50 (see the valuation chapter). Monte Carlo over 20,000 scenarios: P10 …, P25 …, P50 …, P75 …, P90 …; probability of undervaluation …%, probability of a margin above 30% is …%. Broker consensus (15 analysts, average target price …-…) falls right on top of my median — a sign of arithmetic soundness, not originality.

Verdict. BUY, but with a position sized off a range, not a median. What you buy at … is an oil producer with a very low breakeven cost, 10 years of proven reserves and a dividend yield of ~6.1% trailing / ~6.9% forward, at 7.15× trailing profit and 1.16× book value. What you pay is an untreatable exposure to the crude price, a book value of which ~17% is cash that the controlling shareholder has never fully distributed, and an ownership structure in which no one in management holds a single share. The stock is no longer cheap in its own historical terms: the average P/E over 2021-2025 was 4.83×, and today it is 7.15× — the highest multiple in six years. Whoever buys is buying the normalization of a multiple depressed by sanctions, not a fresh re-rating.