Waterdrop (WDH) Q4 2025 Earnings Call Transcript

Waterdrop reported FY2025 revenue of CNY 3.98 billion, up 43.5%, and net profit attributable to ordinary shareholders of CNY 570 million, up 54.8%; Q4 revenue was CNY 1.41 billion (+105.5% YoY) and Q4 net profit CNY 153 million (+62.7% YoY). Management declared a CNY 0.03 per ADS cash dividend totaling $10.8M (record April 24, 2026) and repurchased 60.7M ADS for about $118M. The company deployed large language models across core workflows, filed 72 LLM-related patent applications, and reported insurance-related income of CNY 1.31B (+125% YoY) and AI-driven premium uplifts (mini-program +33% sequential). ESG rating upgraded to A+ and medical platform metrics (3.68M patients, $72.3M raised via crowdfunding) underline technology-led growth and stronger sustainability credentials.

Analysis

Waterdrop’s LLM rollout is a strategic lever that can compress unit economics materially, but the path is nonlinear: expect visible margin volatility over the next 2–4 quarters as elevated customer acquisition and integration costs front-load while AI-driven conversion and retention gains compound later. The company’s move to license its AI agent stack to industry partners creates a subtle but important re‑positioning from distribution-first insurtech to horizontal B2B SaaS, which would shift its multiple from pure growth to blended growth + software margin profile if even a modest portion of revenue becomes recurring third‑party fees within 12–24 months.

The filings (72 LLM patents) matter only if they create extractable rents; in practice they are more likely to be defensive bargaining chips that reduce, not eliminate, commoditization risk from large cloud/open‑source stacks. Second‑order winners include ASIC/GPU suppliers and cloud inference hosts — sustained mini‑program premium lift implies incremental GPU inference spend per incremental premium dollar, favoring NVDA/capacity providers over pure digital brokers. Conversely, incumbent insurers and middlemen who rely on referral fees face margin erosion if Waterdrop scales its agent‑matching and supplier APIs across partners.

Key risks: Chinese AI and healthcare privacy enforcement can re‑price the business within weeks (regulatory headlines) and the crowdfunding segment remains reputationally levered — one high‑profile misuse or fraud event could trigger rapid user and regulator pullback. Watch three catalysts: (1) quarterly FYP trajectory and margin delta vs S&M run‑rate over the next two quarters, (2) any commercial licensing announcements for the AI stack (near‑term de‑risking of monetization), and (3) regulatory guidance on medical data/AI in China; these will re‑rate the name on 3–12 month horizons.

Key Decisions for Investors

• Long WDH ADS (6–12 months): initiate a 1–2% portfolio position on weakness or in tranches if the stock retraces 8–12% intraday; target +30% if management proves sustainable FYP margins and announces AI licensing deals, hard stop at -18% to limit execution risk from regulatory headlines (R/R ~1.7:1).
• Buy NVDA 3–9 month 1:1 call spread to hedge upside in AI infrastructure demand (buy near‑the‑money calls, sell calls 20–30% higher); max loss = premium, max gain ~3–4x premium if inference demand accelerates — use as directional hedge to WDH’s AI monetization path.
• Sell WDH 6‑month covered calls on an existing long or implement a put‑sell if comfortable owning more (strike ~15% above current spot): collect yield to offset near‑term elevated S&M spend while retaining upside to a licensing re‑rating.
• Event hedge: allocate a tactical 0.5% position to an inverse China fintech/insurtech ETF or short concentrated sector peers for 3 months to protect against abrupt regulatory tightening in Chinese healthcare/AI; unwind if regulators issue clarifying, friendly guidance.