LexinFintech Holdings Ltd. (LX) Q4 2025 Earnings Call Transcript

LexinFintech held its Q4 2025 earnings call on Mar 19, 2026; the prepared remarks were delivered by Chairman & CEO Jay Xiao, CRO Arvin Qiao, and CFO Xigui Zheng, and figures are presented in RMB with quarter-over-quarter comparisons. Management gave remarks in Chinese followed by AI-based English voice translations, and the call included analysts from UBS and Citi. The provided excerpt contains no financial metrics, results, or guidance.

Analysis

Lexin’s use of AI-driven investor and risk tools is a leading indicator, not just PR: it implies management is prioritizing durable reductions in origination and servicing unit costs. If AI models can shave 20–50% off manual underwriting/collection labor, net yield on new originations could expand by 100–300bps over 12–24 months absent higher loss rates, amplifying ROE disproportionately versus legacy banks. The immediate second-order beneficiaries will be outsourced collection firms and cloud/AI infra vendors that power model training and inference — expect rising incremental spend with a two- to four-quarter lag.

The principal near-term risks are regulatory and credit-cycle driven. A targeted supervisory campaign or tightened consumer-credit rules could wipe out pricing power within 30–90 days and force balance-sheet de-risking that depresses revenue for several quarters; conversely, macro stabilization and easing of regulatory scrutiny would validate the AI-led efficiency story over 6–18 months. Operationally, model drift and adverse selection (more marginal borrowers enter as cost falls) are multi-quarter hazards that would raise loss rates with a 3–9 month lag.

Consensus appears to treat the AI rollout as binary upside; it underweights the asymmetry that rapid AI adoption creates between front-office growth and back-end loss recognition. The pragmatic trade is to capture upside from efficiency gains while protecting against regulatory/credit shocks: structure directional exposure with time buffers (9–18 months) and explicit hedges to limit a one-off regulatory derating. Relative positions versus other Chinese fintechs will exploit idiosyncratic execution on AI and balance-sheet resilience rather than macro beta alone.

Key Decisions for Investors

• Long LX equity (6–12 month horizon) paired with short QFIN (360 DigiTech) to isolate idiosyncratic AI execution: target a 60/40 notional split. Rationale: capture Lexin’s faster AI-driven unit-cost improvement while hedging sector-wide regulatory/credit risk. Stop-loss: 25% on net position; target 2x upside if AI reduces servicing costs materially within 12 months.
• Buy a 9–12 month call spread on LX to cap cost (long nearer-ATM call, short 150–200% OTM call) to express asymmetric upside from operational leverage while limiting premium outlay. Risk/reward: limited downside to premium vs. 2–4x upside if market re-rates ROE on improved loss-adjusted margins.
• If macro/FX or headlines suggest regulatory tightening (trigger event window: next 30–90 days), tactically flip to a short LX/long large-cap Chinese bank pair for 3–6 months: banks provide deposit stability and lower cyclical credit exposure. Risk: pair may underperform in rapid macro-improvement scenarios; cap exposure to 3–5% portfolio weight.
• Monitor cloud/AI infra suppliers (e.g., Alibaba Cloud) and collection-outsourcers for a 2–4 quarter supply-chain play: initiate small long exposures on evidence of rising vendor spend by Lexin (quarterly procurement cadence). These are lower-beta ways to play technology-driven margin expansion.