Detailed narrative
Credit-cycle inflection and proactive tightening
Management declared the quarter an inflection point after a year of regulatory tightening and industry-wide credit normalization. Leading indicators turned first — first-payment defaults roughly halved from a year ago, every delinquency bucket improved from its Q4 2025 peak, and the asset recovery rate rose for the first time in five quarters — which management reads as proof that proactive credit tightening and AI-driven risk management are working.
The improvement created a healthier operating environment that supported margin expansion, and crucially continued through April and May, after quarter-end, underpinning the expectation of lower provisioning ahead.
Revenue dynamics and the deferred-revenue buffer
Total net revenue of RMB 915.1 million fell 41% year-over-year but only 4% sequentially (from RMB 957.6 million in Q4 FY25), which management framed as increased stabilization. The buffer is the deferred revenue recognition feature of what William called the 'risk-free model' — though he also described the same legacy assets as 'risk-taking' assets moments later, an internal inconsistency (likely ASR garbling of the model name) worth flagging.
Continued recognition of deferred revenue from legacy assets built up over past quarters partially offset lower revenue from new loan facilitations, as loan origination volumes declined and the portfolio was deliberately resized. All figures on the call were stated in RMB despite the company's USD reporting metadata.
All-in-AI: the three-pillar ecosystem strategy
The CEO devoted most of his remarks to positioning Yiren as an AI-native, multi-industry operating platform built on three pillars: (1) the established fintech platform (lending, insurance) providing recurring cash flow, application scenarios and proprietary data; (2) AI infrastructure, where the company is evaluating consolidating its existing computing resources to support internal AI initiatives and potentially serve enterprise customers over time⏳ — explicitly flagged as early-stage evaluation with no investment decision made; and (3) AI applications, spanning incubated agents in financial services (intelligent credit management, insurance assessment) plus education, professional development and entertainment.
Management argues the combination — fintech cash flows, proprietary AI stack, and incubated applications — is difficult to replicate and lets Yiren capture value across every layer of the AI economy.
Magicube 2.0 and internal AI adoption
AI is now integrated into every major business function — marketing, customer acquisition, underwriting, risk management, collection and customer service — and management says it has moved from productivity tool to core operating fabric. The newly launched Magicube 2.0 multi-agent platform (built on the proprietary Zhiyu LLM) adds enterprise governance via the ZhiNao orchestration agent and autonomous workflow execution via the XuanJi agent, positioning it as the core AI platform for both internal operations and the long-term ecosystem strategy.
Tangible proof points cited this quarter include the >50% YoY reduction in customer acquisition cost as a share of revenue and the sharp drop in marketing intensity.
Incubation portfolio and warrant structure
Over the past three years Yiren has invested in or incubated more than nine startups, and has now entered warrant agreements with four of them granting staged options — with no obligation — to increase ownership up to a controlling interest at pre-arranged exercise prices, contingent on operational and strategic milestones.
Management stressed these rights do not constitute current control or consolidation, and that carrying values (historical cost or equity method) may understate operational progress. Two portfolio companies were showcased: an AI-native education platform and an AI-native entertainment/IP company, both framed as future pillars of the ecosystem rather than short-term financial investments.
Insurance repositioning toward digital distribution
Despite continued industry-wide pressure on traditional brokerage commissions following regulatory reforms six quarters ago, the strategic pivot to Internet insurance carried the segment back to growth on both a sequential and year-over-year basis for the first time since the reforms.
The migration of consumers toward online, on-demand insurance purchasing, combined with AI-powered customer acquisition, is expected to make the business an increasingly meaningful revenue contributor. The digital-channel transition is also driving cost optimization — origination, servicing and other operating costs fell to RMB 197.6 million from RMB 250.9 million sequentially, and G&A declined 26% year-over-year, aided by AI-enabled automation across customer service, operations and collections.
Balance sheet, capital allocation and outlook
The balance sheet remains strong, giving room to fund innovation while preserving flexibility. Capital allocation priorities are unchanged: prudent investment in technologies and businesses that strengthen competitive advantages, disciplined risk management, and financial flexibility — a framework spanning the core business, internal AI development (R&D up 27% year-over-year by design) and selective external AI investments.
Management is cautiously optimistic💬 for the remainder of 2026 across its three outlook planks (credit, growth, AI), while explicitly monitoring the macroeconomic and regulatory environment. Notably, no quantified financial guidance was issued and no Q&A session was held.