TCEHY › analyze
TCEHY · Analyze
Date: 2026-07-15 · Price: ~$61.92 (OTC ADR, close 7/15) · Sector: Communication Services / Internet Content · Fiscal year ends December Market cap: ~$560B (ADR-implied) · Primary listing: 0700.HK (Hong Kong) · Reporting: IFRS, RMB Verdict: Highest-quality name in the China-internet complex — a genuine wide-moat compounder priced fair, not cheap. The "high price" that worried you is largely a P/E optical illusion; on cash flow and net of the investment portfolio it's reasonable. Conviction [7.5].
Statement figures in RMB (CNY) as reported. Practical note up front: the US line TCEHY is a thinly-traded OTC ADR (~0.9M shares/day, and it lagged the HK close by double digits at times). For real positions the Hong Kong ordinary (0700.HK) or a broker ADR-conversion is the better vehicle — assess valuation on the ordinary, not the OTC tick.
1. Fundamentals — Health Scorecard
| Metric | FY2024 | FY2025 | Read |
|---|---|---|---|
| Revenue (¥B) | 660.3 | 751.8 | +13.9% — acceleration |
| Gross profit (¥B) | 349.2 | 422.6 | +21% |
| Gross margin | 52.9% | 56.2% | ▲330bps — mix shift to high-margin ads/video/mini-games |
| Operating income (¥B) | 208.8 | 248.8 | +19% — operating leverage |
| Operating margin | 31.6% | 33.1% | Elite for its size |
| Net income (¥B) | 194.1 | 224.8 | +16% |
| Net margin | 29.4% | 29.9% | — |
| ROE / ROIC | 22.8% / 12.7% | 23.6% / 13.3% | Rising, high-quality |
| Operating cash flow (¥B) | 258.5 | 303.1 | +17% |
| CapEx (¥B) | 96.0 | 112.9 | +18% (AI build-out) |
| Free cash flow (¥B) | 162.5 | 190.2 | +17%, positive and growing |
| FCF/share (¥) | 17.53 | 20.93 | +19% |
| Diluted EPS (¥) | 20.63 | 24.32 | +18% |
| Shares out (B) | 9.142 | 9.033 | ▼1.2% — real buybacks |
| Book value/share (¥) | 96.2 | 111.2 | +16% |
Growth quality: This is the cleanest income statement in the group. Revenue +14% with gross margin expanding 330bps and operating profit +19% — growth is more profitable each year, the opposite of BABA/JD/Meituan. FCF +17% to ¥190B and ahead of accounting earnings (OCF/NI ~1.35×).
Balance sheet: Net debt ¥246.7B, net-debt/EBITDA ~1.0×, interest coverage ~18×. Modest leverage carried against a huge asset base. Debt/assets ~20%. Comfortable.
Capital allocation — best-in-class among Chinese tech: - Buybacks ¥75.5B (FY25) + ¥103.8B (FY24) — retiring ~1%+ of shares annually - Dividends ¥37.5B (¥4.13/sh, ~1% yield); dividend +33% YoY - Rising AI capex (¥113B) funded comfortably out of ¥303B OCF — no FCF sacrifice, unlike BABA - Total shareholder yield ~3–4% (buyback + dividend), plus in-kind distributions of investee shares (e.g. the Meituan stake giveaway)
Capital-allocation verdict: Tencent funds a full AI build-out, pays a growing dividend, and buys back stock — all from operating cash flow, while FCF still grows. That is the signature of a mature, cash-generative compounder. Contrast with BABA, which torched FCF and cut buybacks 90% to fund its bet.
2. Moat & Competitive Advantage — the widest in Chinese tech
Quantitative base: 56% gross margin and 13% ROIC, both rising — a moat throwing off increasing returns, not eroding.
The moat stack (multiple, reinforcing): - Weixin/WeChat — the deepest network-effect + switching-cost moat in China. ~1.4B MAU super-app (messaging, social, payments, mini-programs, Video Accounts). It's the operating system of Chinese digital life; nothing else comes close. Video Accounts + Mini-games are now high-margin ad/commerce surfaces monetizing that base — the source of the margin expansion. - Games — world's #1 publisher. Honor of Kings and Peacekeeper Elite at record gross receipts; owns/backs Riot (League of Legends), Supercell, and a stake in Epic. Domestic games +6%, international +13% (Q1'26). Durable IP + live-ops annuity. - Advertising +20% — structurally under-monetized inventory (Video Accounts, Search, Mini-programs) with years of runway; highest incremental margin. - Fintech (Weixin Pay) — payments duopoly with Alipay. Plus wealth management and cloud/business services (intl cloud +40%). - AI: HunYuan LLM integrated across products (ad targeting, game content, Weixin). AI is a demand pull on the ad/cloud engine, not a labor-arbitrage threat.
Adversarial stress-test: A well-funded rival cannot replicate Weixin's 1.4B-user graph, the games IP library, or the payments rails — these are among the most defensible assets in global tech. The credible threats are not competitive; they are regulatory/political (below). Note Tencent sat out the instant-commerce subsidy war that gutted its peers — a sign of discipline and of a business that doesn't need to fight there.
Evergreen assessment: Yes — social/communications + entertainment + payments is a forever-franchise, and the AI/ads monetization cycle extends the runway. The binding risks are external.
3. Valuation — is it actually overpriced? (your core question)
At ~$62 ADR / ~¥525 ordinary. This is the crux, so take it in three layers:
| Model | Output | Weight & note |
|---|---|---|
| Graham √(22.5×EPS×BVPS) | ¥247 vs ¥525 price → "2× overvalued" | Downweight. Graham punishes asset-light, high-ROE compounders (low book vs earnings power). Wrong lens for this business type per framework §3. |
| P/E | ~21.6× trailing (~18–20× fwd) | The number that looks expensive — but see below |
| P/FCF | ~11.9× | The number that matters. A 30%-margin, 13%-ROIC franchise growing FCF ~17% at ~12× FCF is not expensive. |
| EV/EBITDA | ~10× (Alpha Vantage) / higher on roic's op-income proxy | Reasonable for the growth/quality |
| Bogle expected return | ~11–14%: ~10–13% earnings growth + ~1% yield ± modest multiple change | Attractive if the multiple holds |
| DYT / DDM | N/A (~1% yield) — but total shareholder yield ~3–4% | Not a dividend story |
Resolving the P/E-vs-P/FCF gap and your objection: The 21.6× P/E overstates the price for two structural reasons: 1. Cash generation runs well ahead of earnings (FCF/EPS gap; OCF/NI ~1.35×). On cash — the framework's preferred measure — you're paying ~12×, not ~22×. 2. The investment portfolio. Tencent carries >¥1.0 trillion (~$143B book) in "other non-current assets" — stakes in PDD, Kuaishou, Xiaohongshu, Epic, and many unlisted names. That's ~25% of the market cap sitting inside the price at book (likely more at market for the listed slice). Strip it out and the core operating business trades cheaper than the headline. You are partly getting a $140B+ investment book for free.
So: not overpriced. Your instinct ("higher quality but I'm paying up") is half-right — you are paying a quality premium versus BABA — but the premium is smaller than the P/E implies, and it's largely earned by superior, expanding economics plus a large asset cushion. This is quality at a fair price, not "great company, bad entry."
Fair-value range (ADR): ~$58 – $78. Current ~$62 sits at the low end of fair — a reasonable entry, not a bargain. Add on dips to the low-$50s. Upside to $80+ if margin expansion + ad/AI monetization + buybacks continue and the multiple re-rates toward global peers; the investment portfolio is a downside cushion / free-ish call option (with the caveat that Beijing is nudging Tencent to shrink it — which returns cash to holders via distributions but caps SOTP upside).
4. Synthesis — Weighted Verdict
The four lenses barely conflict; the only tension is Graham (says expensive) vs FCF/quality (says fair) — and for a high-ROE, asset-light compounder the framework explicitly downweights Graham. Resolved in favor of fair-to-attractive.
- Fundamentals: best-in-class — accelerating revenue, expanding margins, growing FCF, real buybacks, comfortable balance sheet.
- Moat: the widest in Chinese tech (Weixin + games + payments), rising ROIC, and disciplined enough to skip the value-destroying delivery war.
- Valuation: looks full on P/E, is reasonable on FCF (~12×) and cheap-ish net of a ~$140B investment book. Fair price for a superior business.
Tencent vs BABA (your framing): Tencent is the better business at a fair price; BABA is the cheaper business with more contested economics and more valuation upside. For a long-term, quality-first investor transitioning toward balanced growth, Tencent is arguably the better core China holding; BABA is the higher-variance value + catalyst play. Owning both as a small combined sleeve is defensible; if forced to pick one quality anchor, it's Tencent.
🚩 Key risks (why it still isn't a table-pound)
- China regulatory — the 2021 gaming crackdown (approval freezes, minor playtime limits) is the template; gaming/fintech/data are all policy-exposed.
- VIE / ADR structure + delisting risk — same Cayman-shell wrapper as BABA; permanent discount. (0700.HK mitigates the delisting angle vs the OTC ADR.)
- OTC ADR illiquidity (TCEHY specifically) — thin volume, price lag; use the HK ordinary for a real position.
- Investment-portfolio politics — Beijing pressure to unwind stakes caps SOTP upside (though it returns cash).
- Macro China consumer / FX — RMB earnings, USD-quoted ADR; ad and fintech revenue track the domestic economy.
🟢 Golden flags
Margin expansion (GM 53%→56%), FCF +17% funding AI without cutting buybacks, ~1%+ annual share shrink, ad segment +20%, intl cloud +40%, AI (HunYuan) as a demand pull, and a ~$140B investment cushion under the price.
A portfolio-specific passage was removed from the public build.
Verdict
A wide-moat compounder, fairly priced — the objection that it's "expensive" is mostly a P/E optical effect that dissolves on cash flow (~12× FCF) and net of a ~$140B investment portfolio (~25% of the cap). Higher business quality than BABA, with less valuation upside and the same China/VIE risk wrapper. Conviction [7.5] — BUY quality-at-a-fair-price; reasonable at ~$62, add in the low-$50s. Execute via 0700.HK, not the thin TCEHY ADR. Better core China holding than BABA; BABA remains the cheaper, higher-variance value play. Not a table-pound purely because the entry is fair rather than cheap and the China/regulatory overhang is real.
Data: roic.ai (FY24–FY25 statements, ratios, per-share — deep history unavailable for this semi-annual filer), Tencent Q1 2026 results (13 May 2026), SCMP/Reuters (investment-portfolio divestitures). Cross-checks consistent.