Rating: Cautiously Bullish | Target Price: HKD 120–140 (12-month fair value range) | Current Price: HKD 86.05 (close, 2026-10-28) | Margin of Safety: approx. +39% (vs. lower bound of base-case range) | Time Horizon: 12 months
NetEase Cloud Music is currently a "triple-play" asset: an extremely cheap balance sheet, a deteriorating competitive narrative, and returns whose delivery remains to be proven. At the current price of HKD 86.05, the market cap is HKD 18.10 billion, of which net cash of approximately HKD 15.45 billion (RMB 13.131 billion, as of 2026-06-30) accounts for 85.4%, or HKD 73.46 per share in net cash — the current price carries only a 17.1% premium over that figure. The market is effectively paying only about RMB 2.25 billion for the operating business, equivalent to 1.37–1.67x ex-interest recurring net profit — implying that two extreme assumptions hold simultaneously: "operating profit trending toward zero" and "cash essentially never returned to shareholders." The operating reality, however, is: gross margin of 37.2% at a post-IPO high, revenue +3.4%, and RMB 1.6–2.0 billion in recurring net profit generated annually. What is compressing the valuation is the share-loss panic driven by Soda Music's MAU overtaking NetEase Cloud Music, combined with a surge in customer-acquisition spending (2026H1 selling expenses +80.7%), compounded by nearly 5 years of zero dividends since listing and buybacks that stopped below the current price — clear behavioral counter-evidence. We believe the price already reflects excessive pessimism, but the first variable for a re-rating is the delivery of capital returns (buyback resumption or a first-ever dividend) and stabilization of subscription growth. Verification windows are dense (2026-11-19 Q3 segment data, monthly returns filings, and the 2027-02 annual report). Until delivery, this is a catalyst-dependent "cheap but requires waiting" opportunity; hence the Cautiously Bullish rather than Aggressively Bullish rating.
Nature of the opportunity: Catalyst-dependent deep value. Evidence grade B (core balance sheet and earnings figures are audited/reviewed first-hand disclosures, but share and ARPPU data rely on third-party paraphrases with conflicting versions; capital return expectations carry no company commitment). Market environment: headwinds — the industry has shifted from a duopoly into a "three-way battle" restructuring phase, with sector valuation centers shifting down across the board.
Key change axes over the next 6–12 months (current state → fork → verification point):
Primary share price driver: Capital return policy (buybacks/dividends) — net cash of HKD 73.46 per share is 85.4% of market cap; every 10 percentage point change in the cash discount equates to roughly HKD 7.3 per share, making this the only self-initiated catalyst that does not depend on fundamental improvement. Expectation gap: the current price is about 41% below the average sell-side target price of HKD 146.54 (all 11 brokers rated Buy, per Investing.com as of 2026-10-28; most are pre-interim-results stale figures subject to rating inertia), and still 25% below Jefferies' post-interim downgraded HKD 115. What the market is really pricing in is not operational collapse (H1 clean net profit of RMB 809 million, tax rate already normalized), but a dual discount of "zero capital return track record + Soda Music share panic." The falsification path is equally clear: if the December monthly return still shows zero buybacks and the annual report declares no dividend → cash discount becomes entrenched, weighting shifts toward the bear-case scenario.
Key evidence:
Equivalent framing: if the operating business were valued at a conservative comparable 8x ex-interest earnings, then the market is pricing the RMB 13.13 billion of net cash at only about 17% of face value. Hedges that must be candidly acknowledged: first, the "1.11x/1.36x net profit" basis including interest double-counts roughly RMB 450 million/year of deposit interest alongside the cash itself; on an ex-interest basis the multiple is actually 1.37–1.67x — still extreme, but less sensational; second, under the red-chip/VIE structure, distributing onshore cash to the Cayman layer involves dividend-chain friction (withholding tax, FX procedures), and the controlling shareholder's track record of zero dividends in nearly 5 years as a listed company is itself empirical proof of a "governance discount" — counting cash at face value presupposes governance improvement; third, cash is a flow rather than a static stock — the HKD 967 million (approx. RMB 840 million) of buybacks in 2026H1 already reduced total cash-like assets by about RMB 440 million from end-2025.
Key evidence:
Interpretation: the two main drivers of the P&L decline (customer acquisition spend, tax base) do not change the underlying cash-generating ability of the business, but "stabilization" is an artifact of six-month aggregation — the latest quarter has already flattened, and the trend's direction must await the Q3 segment data on November 19. Another quality signal to monitor: H1 receivables and notes rose +34.7% in six months (vs. revenue +3.4% only), with DSO rising to approximately 66 days, unexplained by the company (see chapter on "Financial Engineering and Cross-Period Consistency").
Key evidence:
Interpretation: TME has launched the "Free Music" app and a free Kugou version; the industry is shifting from paywall competition to free traffic pool competition. NetEase Cloud Music's ARPPU of about RMB 5.7/month is less than half of TME's (RMB 11.9), and subscription growth has decelerated from +22.2% to +5.3% over three years. However, the decline in MAU ranking coexists with the company's disclosed "DAU/MAU > 30% with YoY and QoQ increases, retention and renewal rates stable to rising"; there is no direct evidence yet of MAU leaderboard elasticity transmitting to paid revenue — Kugou's long-standing #1 MAU ranking with weaker monetization than QQ Music is precisely a counterexample. The final arbiter of the share narrative is QuestMobile's original monthly reports and quarterly subscription growth, not republished headlines.
Key evidence:
Interpretation: there is no obstacle to buyback authorization (approx. 13.73 million shares remaining under the AGM general mandate), and both the capability and mechanism for dividends are in place; the parent company's dividend forms a stark contrast — but the behavioral counterexample of halting buybacks at HKD 86 must be taken seriously, and NetEase's own Q2 net profit was under pressure (investment losses of approx. RMB 3.0 billion), raising the group's incentive to retain subsidiary liquidity. This thesis is therefore conditional rather than predictive: observation windows for delivery are the monthly returns filings (filings disclosed as early as T+1) and the February 2027 annual report; if the December filing still shows zero buybacks and the annual report declares no dividend, the valuation anchor should be materially cut for "entrenched cash discount" rather than maintaining a static undervaluation conclusion.
| Metric (RMB 100 million) | FY2023 | FY2024 | FY2025 | 2026H1 (latest) |
|---|---|---|---|---|
| Revenue | 78.67 | 79.50 | 77.59 | 39.59 |
| Revenue YoY | -12.5% | +1.1% | -2.4% | +3.4% |
| Net profit attributable to shareholders | 7.34 | 15.62 | 27.48 | 8.09 |
| Attributable net profit YoY | — | +112.7% | +76.0% | -57.1% |
| Adjusted net profit (company basis, adding back SBC) | 8.19 | 17.00 | 28.60 | 8.60 |
| Recurring net profit (this report's estimate, excluding one-off items) | — | — | 20.22 | 8.09 (tax rate normalized) |
| Gross margin | 26.7% | 33.7% | 35.7% | 37.2% |
| Net margin | 9.3% | 19.6% | 35.4% | 20.4% |
| Operating cash flow | 1.89 | 17.75 | 16.18 | 4.58 |
| Free cash flow (OCF − PP&E purchases) | — | ≈17.7 | 16.07 | 4.57 |
| Cash + cash-like assets (period-end) | — | — | ≈135.8 | 131.31 (2026-06-30) |
| Interest-bearing debt | 0 | 0 | 0 | 0 |
| Debt-to-asset ratio (period-end) | — | 25.0% | 21.8% | — |
| Membership subscription revenue | — | 44.59 | 50.53 | 26.01 |
| Subscription revenue YoY | — | +22.2% | +13.3% | +5.3% |
Note: FY2025 recurring net profit = attributable 27.48 − one-off deferred income tax asset recognition gain of 7.47 + add-back of associate impairment of 0.20 ≈ RMB 2.022 billion; excluding government grants (approx. RMB 180 million) further reduces it to approx. RMB 1.84 billion. 2026H1 net cash basis: cash 24.07 + bank deposits 94.44 + wealth management products 12.79 (at end-2025 there were still RMB 1.40 billion in deposits with maturities over one year; by 2026-06-30 these had matured into current holdings, now zero); restricted cash approx. zero, zero interest-bearing debt.
Reasons for metric changes (items with |YoY| ≥ 20%, all with company explanations):
Revenue of RMB 3.959 billion, +3.4% YoY — the first return to positive growth since 2023 — but structurally Q1 was +6.6% and Q2 roughly +0.4% (NetEase segment basis "stable"), with momentum high early and flat later. Gross margin hit another record high of 37.2% (content cost control + rising share of self-produced/independent musician content), validating that the industry logic of structurally declining copyright costs continues to play out in the core online music business. What truly deteriorated was the expense line: to counter Soda Music's free-model impact, the selling expense ratio jumped from 4.27% to 7.46% YoY; operating profit was RMB 746 million (-11.6%), operating margin -3.2pp; attributable profit of RMB 809 million (-57.1%) was mainly due to the high base of the prior year's RMB 849 million deferred tax credit — excluding it, roughly -22%. Cash flow quality weakening deserves a callout: operating cash flow of RMB 458 million (-13.1%), with receivables rising by a net RMB 371 million in six months (+34.7%), far outpacing revenue growth — unexplained by the company. The market's reaction to these interim results was an intraday drop of as much as -18% the next day — pricing in the combination of "burning cash on the customer-acquisition war + growth deceleration," not the absolute level of profits. This report's judgment: the interim results neither confirmed "operational collapse" (revenue, gross margin, and subscriptions all growing positively) nor falsified "share pressure" (Q2 flat, expenses surging); the verdict rests with the Q3 segment data on 2026-11-19.
Business model in brief: Asset-light subscription platform. Revenue = online music services (primarily subscriptions, 84.8% of segment) + social entertainment (residual livestreaming tipping, under proactive contraction). Subscription revenue accounts for 65.7% of total revenue (2026H1), highly recurring; pricing power is weak — growth relies on member count rather than price increases, with ARPPU persistently diluted by membership mix (the company has used the word "diluted" in self-descriptions for three consecutive reporting periods). Content licensing/revenue-sharing costs are the largest cost item, and capex is nearly zero (FY2025 PP&E purchases of only RMB 11 million; CapEx/D&A of 0.89).
Cash content of earnings test: OCF/attributable net profit for 2023–2026H1 was 0.26 → 1.14 → 0.59 → 0.57 — highly volatile and persistently below 0.8 since 2025 — but there are two main reasons: first, FY2025 net profit includes RMB 747 million of non-cash deferred tax gains (inflating the denominator); second, 2026H1 receivables surged by RMB 371 million, dragging collections. Excluding one-off tax items, OCF/recurring net profit for FY2024–FY2025 was approx. 1.0–1.14 and 0.80 respectively (16.18/20.22 = 0.80); book profits can still broadly convert to cash, but the 0.57 in 2026H1 and the receivables trend are deterioration signals requiring annual report verification. Recurring earnings test: of FY2025 attributable net profit of RMB 2.748 billion, approx. RMB 747 million came from the one-off deferred tax asset recognition (plus approx. RMB 180 million of government grants), leaving core operating profit of approx. RMB 1.84–2.02 billion — roughly 30% of the headline +76% profit growth was a tax illusion. Cross-check against company-reported basis: the company's adjusted net profit (RMB 2.860 billion) adds back only approx. RMB 110 million/year of SBC — a relatively conservative basis with no aggressive expense add-backs observed.
Return on capital: estimated ROIC of approx. 10.7% (FY2025) — but the denominator is severely diluted by net cash at the RMB 14 billion level; ex-cash, operating capital returns are far higher, and the incremental capital returns of the subscription business are actually excellent (content costs scale linearly with revenue, no heavy-asset investment).
Red flags: ① Core operating KPIs (MAU/paying users/ARPPU) have not disclosed specific figures since the FY2024 annual report, reducing trackability; ② earnings dependence on non-operating items: FY2025 interest income of RMB 446 million was 21.6% of pre-tax profit (20.4% in 2026H1, down -14.8% with falling deposit rates); ③ customer concentration rising for three consecutive years (top five customers 10.9% → 14.2% → 17.6%, with Customer A at 15.1% of 2026H1 revenue).
Say-Do Consistency: Partially Delivered. Evidence 1: FY2024 annual report promised "optimizing profitability through standardized operations" → FY2025 gross margin +2.0pp, selling expenses -33.2%, adjusted net profit +68.2% — delivered. Evidence 2: Subscription growth promises delivered consecutively but decelerating tier by tier (+22.2% → +13.3% → +5.3%) — direction consistent, momentum fading. Evidence 3: FY2025 annual report promised "sustained cost optimization and improved operating efficiency" → 2026H1 selling expenses +80.7%, operating margin -3.2pp — first half went the opposite way. The qualitative outlook carries no numerical guidance; the company has never issued a verifiable quantitative commitment — a virtue in terms of low exaggeration risk, a flaw in terms of low transparency. Overall judgment: pragmatic, no hype, but the strategic focus in 2026 has shifted from "profit first" to "defensive customer acquisition," contradicting the annual report narrative without an updated explanation.
Shareholder Friendliness: Neutral to Unfriendly. Zero dividends in nearly 5 years since listing (none paid in FY2024/FY2025/2026 interim); the HK$967 million buyback from February to May 2026 (approximately 3.47% of share capital, already cancelled) was the only positive move, but the average price of HK$132.47 is 54% above the current price, and purchases were suspended for four consecutive months in the HK$86 range with zero explanation. No post-listing refinancing dilution (RSUs settled with existing shares). Overall: capable, has precedent, but no persistence — the absence of a return policy is a direct cause of the valuation discount.
Risk Signals: William Ding serves as both Chairman and CEO (deviation from the Corporate Governance Code, disclosed with explanation); related-party transactions with NetEase Group are controllable (FY2025 purchases of RMB 265 million, within annual cap); no record of controlling shareholder selling, nor buying.
| Segment | Revenue Share | 2026H1 YoY | FY2025 YoY | Business Logic in One Sentence |
|---|---|---|---|---|
| Online music services (subscriptions 84.8% + digital albums/advertising, etc.) | 77.5% (2026H1); 55.3% in 2023 → 67.4% in 2024 → 77.3% in 2025 | +3.4% | +12.0% | Subscription model driven by licensed catalog and community-based recommendation; profit core and growth engine |
| Social entertainment services and others (live-stream tipping, etc.) | 22.5% (2026H1); 44.7% in 2023 → 32.6% in 2024 → 22.7% in 2025 | +3.7% | -32.0% | Proactive contraction under tightening regulation; high revenue-sharing costs, historical drag on gross margin |
The company's CODM manages by a single segment and does not disclose gross margin by business line — segment-level gross margin differences are structurally uncomputable, and this is noted here as fact. Profit driver judgment (inferred via revenue_share × gross margin structure): online music subscription is the sole source of profit and cash flow; after two years of clearing, social entertainment's revenue share has fallen to ~20%, and 2026H1's +3.7% represents stabilization, not renewed expansion. Structural change worth noting: the revenue mix has fully shifted from "music + live-streaming dual engines" to "pure music subscription" (online music share up 22pp in two years) — the company is purer in quality, but its competitor has also become ByteDance, the master of the free model, rather than a tacitly layering TME.
Accounting Red Flags (pattern/severity/original-text evidence):
Cross-Period Consistency:
Overall judgment: no hard accounting manipulation of the fabricated-revenue or aggressive-recognition type found, but the combination of "one-off gains inflating the base + KPI disclosure suspension + unexplained receivables surge" means the apparent improvement in the financials warrants a sustained quality discount.
Current Market Data (closing price 2026-10-28): HK$86.05; total shares 210.33 million; market cap HK$18.10 billion; average daily turnover approximately HK$34 million (about 0.19% of market cap — low liquidity, with impact costs for large orders). Multiples and percentiles (as of 2026-10-27, historical percentiles based on series since listing): PE(TTM) on reported basis 9.2–9.5x (roughly 23rd percentile over the past 5 years — earnings contain one-off items, mechanically distorted; percentile for reference only); about 8.7x excluding one-off items; PB 1.26x (2nd percentile since listing); PS(TTM) 1.95x. FY2026E recurring net profit of RMB 1.66 billion implies forward PE of 9.3x; ex-net-cash, the operating business implies only about 1.67x EBIT earnings (2026E) / 1.37x (FY2025) — horizontal PE comparisons lose meaning given the 85% cash weighting; this is the correct way to value this name.
Peer Comparison:
| Company | Market Cap | PE(TTM) | PS(TTM) | Revenue Growth | Paying Ratio/ARPPU | Notes |
|---|---|---|---|---|---|---|
| NetEase Cloud Music | HK$18.1 billion | ≈9.2x (8.7x ex-one-offs) | ≈1.95x | +3.4% (2026H1) | 25.2% / RMB 5.7 | Ex-cash, operating business at only 1.4–1.7x EBIT earnings |
| Tencent Music (TME) | ≈US$13.3 billion | ≈9.9x | ≈2.6x | +15.8% (2025) → +5.8% (2026Q2) | 24.1% / RMB 11.9 | ≈7.6x ex-cash; ~66% revenue share |
| Spotify | ≈US$120.3 billion | ≈31.5x | ≈5.5x | +14% (2026Q3, constant currency) | 38.8% / ≈€4.71 | Global mature benchmark |
Market-Implied Expectations: Market cap HK$18.10 billion − net cash HK$15.45 billion = HK$2.65 billion (approximately RMB 2.25 billion) for the operating business. This bid requires: either operating profits trending to zero (reality: RMB 1.6–2.0 billion recurring net profit per year, record-high gross margin), or cash essentially never returned to minority shareholders (reality: zero dividends in nearly 5 years since listing — this assumption is not unreasonable; it is the rational core of the pricing). In other words, the market is not mispricing facts; it is pricing a binary fear of governance and competition. Our disagreement is only on magnitude: both extreme assumptions must fully hold simultaneously to justify a bid of 1.4–1.7x EBIT earnings.
Three-Layer Value (EPV): ① Asset value (floor) = net cash HK$73.46/share; ② EPV-zero growth: normalized FCFF of RMB 1.05 billion/year (annualized operating profit RMB 1.49 billion + depreciation 0.2 − maintenance capex 0.3 − working capital increment 1.0, after tax; leaving buffer for receivables drag vs annualized 2026H1 OCF of RMB 920 million), cash flow of HK$5.87/share ÷ 10% discount rate = HK$58.7 operating layer, plus net cash HK$73.46 → EPV ≈ HK$132.2/share (range 117–140, corresponding to discount rates of 10.5%–9.5%); interest income excluded from operating cash flow and cash treated separately to avoid double counting. ③ At the current price of HK$86.05, the discount to EPV is 34.9%, and the growth option value is negative — the market is not only paying nothing for growth above zero growth, but bidding only about 21% (HK$12.59/share) of the operating layer's zero-growth value (HK$58.7/share). The long-term gap of paying ratio (25.2%) and ARPPU (RMB 5.7) versus Spotify (38.8%/≈RMB 37) is a free option at the current price. EPV is a floor reference, not a target price: it assumes cash belongs to shareholders at face value, yet the capital return record is zero — EPV will not be automatically realized, which is precisely where the scenario divergence lies.
Three Scenarios and Odds (12 months):
| Scenario | Probability | Fair Value Range (HK$) | Midpoint vs Current | Key Drivers |
|---|---|---|---|---|
| Bear | 30% | 63–80 | -16.9% (to midpoint 71.5) | Subscription growth turning negative + customer acquisition war not converging (FY2027E recurring net profit falling to ~RMB 1.4 billion) + cash discount widening to 30–40%; earnings decline alone cannot break through the current price — the bear case requires the absence of capital returns on top |
| Base | 45% | 120–140 | +51.1% (to midpoint 130) | Low-single-digit revenue growth, partial expense normalization (FY2027E recurring net profit RMB 1.79 billion, EBIT RMB 1.48 billion); operating business at 7–9x EBIT earnings (anchored to TME's 7.6x ex-cash, discounted for growth differential, not the company's own current multiple) + net cash discount converging to 10–15% (first signal of capital returns) |
| Bull | 25% | 160–190 | +103% (to midpoint 175) | Share stabilization (Soda's monetization stalling) + buyback/dividend formally implemented: zero cash discount + operating business at 11–13x EBIT earnings, converging above EPV 132 |
Probability-weighted fair value approximately HK$123.7 (+43.8%); base-case floor of HK$120 implies a 39% margin of safety; risk-reward ratio approximately 3.0:1 (upside to base-case midpoint +51.1% vs downside to bear-case midpoint -16.9%). The distribution is significantly right-skewed, but realization depends on specific events — not an automatic mean-reversion process.
Our Own Earnings Forecast (alongside management guidance and sell-side expectations): The company provides no quantitative guidance; the latest post-interim sell-side figures are Jefferies at HK$115 (Buy) and Investing.com consensus target of HK$146.54 (mostly pre-interim stale values). This report's forecast:
| Period | Revenue (RMB 100M) | Recurring Net Profit (RMB 100M) | Key Drivers |
|---|---|---|---|
| FY2026E | 79.5–80.1 (+2.5% to +3.2%) | 16.2–17.2 | Subscriptions +5–6%, social entertainment +3–5%, gross margin 37.2–37.5%, selling expenses elevated at RMB 580–600 million, tax rate 13–14% |
| FY2027E | 81.8–84.0 (+3% to +5%) | 17.2–18.5 | Subscriptions +6–7%, partial pullback in promotion spend releasing RMB 100–200 million, interest income contracting with falling rates |
Conclusion: Undervalued, target price HK$120–140. Separating quality from price: the balance sheet and core-business cash generation are hard (net cash at 85% of market cap, record gross margin, positive subscription growth); the governance and competitive margins are soft (zero return record, Soda Music impact, KPI disclosure suspension). The price has already fully priced in both sets of pessimism; the odds are favorable, but "cheap" by itself is not a catalyst — this is a cautious bullish view, not a full endorsement.
Industry Size: Three measures side by side. ① China's digital music industry size: RMB 102.746 billion (2024, Music Industry Development Research Center of Communication University of China, +15% YoY); ② China's recorded music revenue grew +20.1% YoY in 2025, overtaking Germany as the world's fourth-largest market (IFPI Global Music Report 2026; among the fastest-growing major markets; global recorded music +6.4% in 2025); ③ online music users: 729 million (2024) — the user base has essentially peaked. Guosen Securities (2025-06) expects the online music industry to sustain a ~15% CAGR over the next three years; TME's medium-term guidance is paying penetration to 25% and ARPPU to RMB 15 within 3–5 years. The math of industry growth has shifted from "user numbers" to "paying ratio × ARPPU": Chinese platform paying ratios at 20–25% vs Spotify's 38.8%; ARPPU at RMB 5.7–11.9 vs Spotify's ≈RMB 37 equivalent — the headroom is real, but realization is constrained by both purchasing power and free alternatives; overseas penetration cannot be linearly extrapolated.
Industry Chain and Value Distribution: Upstream are rights holders (the three majors Universal/Sony/Warner + thousands of independent labels + independent musicians). China's structural characteristics: the majors' share is lower than globally and the upstream is fragmented; since the exclusive licensing regime was ordered dismantled in July 2021, platforms' bargaining power over upstream is significantly stronger than overseas platforms — this is the institutional driver behind NetEase Cloud's gross margin recovery from -12.2% (2020, content costs at 97.8% of revenue) to 37.2% over five years, and also a structural dividend of industry-wide gross margin flowing back to midstream platforms. Midstream platforms handle catalog operations, recommendation/distribution, and monetization; downstream C-end willingness to pay is the ultimate bottleneck. NetEase Cloud sits at the midstream platform layer: strong bargaining power over upstream (but still constrained by TME's first-mover position on top-tier catalogs such as Jay Chou's), weak bargaining power over downstream users (ARPPU gains can only come from benefit upgrades, not direct price increases). AI-generated content is eroding the traditional moat of "catalog size": Suno users generate roughly 7 million tracks per day; on Deezer, AI-generated tracks account for about one-third of daily uploads — catalogs are no longer scarce, and the competitive focus shifts to recommendation trust and community governance, which for community-born NetEase Cloud is both a reconfirmation of its moat and a rise in content governance costs.
Supply, Demand, and Competitive Landscape: On the demand side, a stock game (TME's 2025Q4 online music MAU -5% YoY to 528 million; NetEase Cloud down from a peak of 184 million four years ago to ~147 million). Three supply-side changes: ① after the end of exclusivity, content is non-exclusive and entry barriers for new players have collapsed; ② ByteDance's Soda Music, leveraging Douyin traffic + the "watch 80 seconds of ads for 24-hour membership" free model, went from 0 to 156 million MAU in two years (QuestMobile 2026Q1, +78%), with Tomato Music up +92.4% in the same period — ByteDance's two products combined exceed 250 million MAU; ③ the explosion of AI music supply dilutes catalog scarcity. Concentration: revenue-based CR2 ~93% (TME ~66%, NetEase Cloud ~27%, Guosen 2025-06 estimate); MAU-based ranking has been restructured — in 2026Q1, Kugou, QQ Music, and Soda rank top three, with NetEase Cloud falling to fourth. Entry barriers have shifted from licensing costs to traffic ecosystems (Douyin's daily music video exposure exceeding 100 billion) and accumulated recommendation algorithm data. Substitution threat: short video is the first entry point for music discovery, squeezing time spent on dedicated music apps; the industry has moved from "one super, one strong" into a "three-kingdoms standoff," and the divergence of three business models (TME's licensing ecosystem + fan economy premium, NetEase Cloud's community taste + low-price defense, ByteDance's content generation + integrated distribution free model) reduces the probability of a head-on price war, but the user battle is already white-hot — TME launched a standalone "Free Music" app on Huawei's channel in September 2026 in response.
Cyclicality and Regulation: The industry is in the mid-to-late growth stage / stock-cultivation phase, with paid conversion as the remaining dividend. Four regulatory tracks: ① abolition of exclusive licensing (2021) — already reshaped the cost base, shifting from a benefit for new players to a competitive headwind for incumbents; ② tightening of live-streaming/tipping regulation — directly caused NetEase Cloud's social entertainment FY2025 -32%; that segment has been largely cleared; ③ copyright protection and paid-content policies are a long-term positive for subscriptions; ④ the copyright ownership and revenue-sharing framework for AI-generated music remains unsettled (the 2026 Munich court ruling against Suno for infringement, domestic "AI song-laundering" controversies) — the biggest rule variable.
Company's Industry Position: Second pole by revenue share (~27%), fourth by MAU share — a misaligned pressure period of "second in scale, fourth in users." The solid part of the moat: the community and independent musician ecosystem has hard data (over 1.25 million registered independent musicians, 7.3 million cumulative uploaded tracks as of end-2026H1; in 2020, independent musicians' songs, at 1.67% of catalog, contributed 45%+ of plays; DAU/MAU sustained above 30%; paying penetration of 25.2% above TME's 20.6% — leading stickiness and willingness to pay). The narrowing part of the moat: absence of top-tier catalogs limits user acquisition; the low-price membership structure keeps ARPPU at only half of TME's and still diluting — the "loyalty for low price" monetization ceiling is evident. A note on market-share measures: revenue share (Guosen 2025-06 estimate) and MAU share (QuestMobile) are two different measures at two different points in time; the former is temporarily stable because Soda's monetization has just begun, while the latter has already flipped — if ByteDance launches Soda membership monetization, revenue share will come under direct pressure; this is the industry's biggest variable over the next 12 months.
The decisive battle in music streaming lies not in the income statement but in three user-side levers — also the dimension most easily missed by generic financial frameworks:
Paying Ratio: NetEase Cloud 25.2% (2024, company's historical disclosure) > TME 20.6% (same comparable source) < Spotify 38.8% (2026Q3). NetEase Cloud's per-user paid conversion is leading in the Chinese market — its community/playlists/independent musician ecosystem is more efficient at cultivating paying habits than the catalog-stacking model. Each 1pp increase in paying ratio ≈ ~20 million additional paying users (on a ~200 million MAU base) ≈ ~RMB 1.4 billion in annual subscription revenue (at RMB 5.7/month) — enormous elasticity, but premised on MAU no longer eroding.
ARPPU: ~RMB 5.7/month (2024 measure, disclosure suspended since), less than half of TME's (RMB 11.9) and about one-sixth of Spotify's (≈RMB 37 equivalent). Management has self-reported for three consecutive quarters that "membership mix changes caused ARPPU dilution" — with 88VIP co-branding and student memberships, it chose new-user acquisition over dilution control. The hedge is SVIP-ization: expanding K-Pop label partnerships (RBW, StarShip, Shofar, DSP Media, etc.) + concert ticketing/merchandise benefits (the MAMAMOO 12th anniversary tour collaboration has validated the "licensing → SVIP conversion" loop) — fan groups with both high spend and high willingness to pay are the most realistic lever for ARPPU repair; TME's successful path with 20+ million SVIPs can be referenced, though NetEase Cloud has yet to disclose SVIP data.
Content Costs: Revenue sharing/licensing costs are the largest cost item, falling from 97.8% of revenue (2020 peak of the licensing wars) to 64.3% (2025). Two structural downward channels: narrowing renewal premiums post-exclusivity + rising share of self-produced/independent musician content; the reverse risk is that if majors such as Universal/Warner raise renewal prices amid intensifying competition, the gross margin recovery would be interrupted (the company completed Universal/Warner renewals in early 2026 — the short-term window is safe). The combination of these three levers determines valuation elasticity: any improvement in paying ratio × ARPPU carries high operating leverage on a 37% gross margin, while MAU erosion hits both simultaneously — this is why the market is so sensitive to QuestMobile's monthly reports.
Overall assessment: Cautiously bullish, confidence 0.55, time horizon 12 months. The price of HKD 86.05 has priced in the dual extreme scenario of "operations trending to zero + no cash return," implying a margin of safety of +39% (versus the base-case floor of HKD 120), a probability-weighted expected return of +43.8%, and odds of roughly 3:1; however, realization depends on two explicit conditions — capital returns being implemented (buybacks/dividends) and subscription growth stabilizing — and until these are verified, deep value can go deeper, so leverage should not be used. Execution constraint: average daily turnover is approximately HKD 34 million, so position building should be executed over multiple days.
Strategy: Suitable for phased participation with net cash as the floor-value logic, using November Q3 data and the buyback column in the monthly returns as signals for adding/reducing positions; if the price falls below the bear-case floor of HKD 63 (implying a cash discount of over 40% with no hedging event), the thesis should be re-examined rather than averaged down.
Risk and catalyst highlights: Upside catalysts, in order, are the resumption of buybacks in the monthly returns (earliest signal, visible at T+1), Q3 segment data on 2026-11-19 (subscription growth ≥5% with expense ratio declining would falsify the "cash-burning collapse" narrative), first dividend in the FY2027-02 annual report, and narrowing of the QuestMobile monthly data gap; downside falsification triggers, in order, are subscription revenue turning negative in a single quarter, consecutive zero buybacks spanning the annual report window, Soda Music MAU exceeding 200 million while NetEase Cloud Music falls below 140 million, and the selling expense ratio becoming entrenched above 7%. The tail risk is the VIE structure and capital accessibility (low probability, unhedgeable, yet the single fatal attack point on the "cash = value" assumption).
This report is based on publicly disclosed documents and public news; financial data are as of the 2026-06-30 interim report, and the price anchor is the 2026-10-28 closing price of HKD 86.05. The forecasts and scenarios in this report are analytical judgments and do not constitute investment advice.