Under the new lending assistance regulations, performance has shrunk significantly! Can publicly listed fintech companies break through using AI and overseas expansion?

Ask AI · How do new regulations on loan assistance promote high-quality transformation in the industry?

Affected by the new loan assistance regulations, fintech companies in the loan assistance sector generally saw their performance shrink in the fourth quarter of last year, but overall annual performance remained steady. Among them, artificial intelligence is still an important lever for loan assistance fintech companies to reduce costs and improve efficiency and performance.

With the new loan assistance regulations (《Notice on Strengthening the Management of Commercial Banks’ Internet Loan Assistance Business and Improving the Quality and Effectiveness of Financial Services》) officially implemented, the industry’s overall financing costs have been steadily declining, entry thresholds and compliance regulatory standards continue to rise, and loan assistance fintech companies face mounting performance pressure.

Based on annual report data disclosed in recent days, Qifu Technology, Xinye Technology, Lexin, Xiaoying Technology, JY (JiaYin) Technology, and Yiren Zhike all saw a significant contraction in their fourth-quarter performance last year, with double-digit year-over-year declines in both operating revenue and net profit. Looking at the full year of 2025, overall performance remained steady, but some institutions saw their net profits drop sharply, and the “Matthew effect” in the industry has been reinforced.

In the past few years, performance benchmarks for loan assistance fintech companies have been anchored in outstanding loan balances, the scale of facilitated disbursements, and user growth. However, after the implementation of the new loan assistance regulations, loan assistance fintech companies have strengthened AI (artificial intelligence) empowerment, stressing growth with quality, and seeking room for growth in overseas markets and niche segments.

In the view of industry experts, under the new loan assistance regulations, the old benchmarks based on scale, outstanding loan balances, and disbursement volume are no longer sustainable and may even imply risks. The shift toward growth with quality, overseas income, and scenario GMV (gross merchandise value) not only aligns with a compliance-driven direction, weakening dependence on scale and strengthening profitability and risk indicators; it also guides market expectations, conveys the results of the transformation, and stabilizes valuations and confidence—directly reflecting the industry’s transition from scale-driven races to quality-focused games.

Significant performance contraction in the fourth quarter of last year

Affected by the loan assistance regulations, which took effect on October 1, 2025, loan assistance fintech companies generally saw their fourth-quarter 2025 performance contract.

Specifically, in Q4 2025, Qifu Technology’s revenue was 40.93 billion yuan, down 8.68% year over year; net profit was 10.16 billion yuan, down 46.89%. In Q4 2025, among Qifu Technology’s two major business segments, credit-driven services saw some growth in net revenue, but platform services net revenue contracted sharply. This quarter, platform services net revenue was only 6.61 billion yuan, versus 15.93 billion yuan in the same period of 2024 and 13.37 billion yuan in the previous quarter.

Due to the contraction of domestic business and the increase in risk-control provisions, in the fourth quarter of 2025 Xinye Technology’s revenue was 30.24 billion yuan, down 12.5% year over year; net profit was 4.16 billion yuan, down 39.0%. After adjusting to exclude non-recurring items, operating profit was 5.20 billion yuan. Transaction volume in the China market was 387 billion yuan, down 28.3% year over year. By the end of last year, domestic loan balances were 683 billion yuan, which also declined.

Looking only at last year’s fourth quarter, Lexin’s performance also contracted. In Q4 2025, Lexin’s operating revenue was 30.43 billion yuan, down 16.8% year over year; net profit was 2.14 billion yuan, down 41.0%. In Q4 2025, Lexin’s credit service revenue was 24.85 billion yuan, down 8.4% year over year; technology-empowered service revenue was 1.70 billion yuan, down 71.7%; the total loan origination amount was 500 billion yuan, down 3.8%.

JY (JiaYin) Technology and Xiaoying Technology also faced clear pressure. In Q4 2025, JY (JiaYin) Technology’s revenue was 10.90 billion yuan, down 22.4% year over year; net profit attributable to ordinary shareholders of the company was 1.01 billion yuan, compared with 2.76 billion yuan in the same period last year. In Q4 2025, Xiaoying Technology’s revenue was 14.68 billion yuan, down 14.1% year over year; net profit was 57.20 million yuan, with a decline of 85.2%. Among this, loan facilitation service fees were 4.41 billion yuan, down 49.8% year over year; the loan disbursed amount was 227.68 billion yuan, down 29.5%.

In Q4 last year, Yiren Zhike’s performance deterioration was even more pronounced. Net profit fell into a loss of 8.82 billion yuan, whereas it recorded a profit of 3.31 billion yuan in the same period of 2024. According to the financial report, in Q4 2025 Yiren Zhike’s loan issuance amount was 120 billion yuan, down 40% quarter over quarter; the ending outstanding loan balance fell to 286 billion yuan, down 17% quarter over quarter from 342 billion yuan at the end of the previous quarter.

Overall steady performance for the full year

Although performance in Q4 last year declined across the board, looking at the full year, the overall performance of loan assistance fintech companies remained stable.

In 2025, Qifu Technology’s net revenue was 192.05 billion yuan, up 11.88%; net profit was 59.76 billion yuan, down 4.35% year over year. For the full year, Qifu Technology facilitated disbursements of 3,270.7 billion yuan, up 1.6%, and the re-borrowing rate reached 93.3%. As of December 31, 2025, Qifu Technology’s outstanding loan balance was 1,260.1 billion yuan, down 8.8% from the end of the previous quarter and down 8% year over year. The cumulative number of credit-granting users was 63.60 million, up 11.8%; the cumulative number of borrowing users reached 38.90 million, up 13.0%.

In 2025, Xinye Technology achieved annual operating revenue of 135.69 billion yuan, up 3.85%; net profit was 25.45 billion yuan, up 6.57%. In terms of operating indicators, Xinye Technology’s annual facilitated transaction amount reached 2,003 billion yuan, and its outstanding loan balance was 709 billion yuan. Full-year domestic transaction volume was 1,863 billion yuan, down 5% year over year. As of December 31, 2025, the total number of registered users in the China market was 187.4 million, up 8.6%; the cumulative number of borrowers was 29.0 million, up 8.2%.

Unlike Qifu Technology’s “increase in revenue with a decline in profit,” Lexin achieved “a decrease in revenue with an increase in profit.” In 2025, Lexin’s total revenue was 131.52 billion yuan, down 7.4% year over year; profit attributable to the parent company increased 52.4% year over year to 16.77 billion yuan. Adjusted net profit attributable to ordinary shareholders increased 49.2% year over year to 17.95 billion yuan. Based on the financial report, Lexin’s 2025 revenue declined while net profit rose significantly. On the one hand, this was due to optimization of the revenue structure; on the other hand, it was driven by lower costs and improved asset quality. In 2025, Lexin’s funding costs were 2.29 billion yuan, down 30% year over year; provision for contingent guarantee obligations was 31.75 billion yuan, down 13.2% year over year. In addition, changes in the fair value of financial guarantee derivatives shifted from a loss of 9.79 billion yuan in the same period last year to a profit of 5.08 billion yuan.

Xiaoying Technology and JY (JiaYin) Technology also delivered relatively steady full-year performance. Among them, Xiaoying Technology’s full-year net revenue in 2025 was 76.39 billion yuan, up 30.1%. Net profit was 14.65 billion yuan, down 4.9% year over year. JY (JiaYin) Technology’s revenue last year was 62.22 billion yuan, up 7.3%; net profit attributable to ordinary shareholders of the company was 15.36 billion yuan, up 45.36%; and the full-year facilitated loan matchmaking transaction volume was approximately 1,290 billion yuan, up 28%.

By contrast, Yiren Zhike’s performance saw larger fluctuations. In 2025, Yiren Zhike’s total net revenue was 5.719 billion yuan, a slight decline of 1% year over year. Full-year net profit was only 40.50 million yuan, down 97.4% year over year compared with 15.82 billion yuan in 2024. The financial report disclosed that in 2025 Yiren Zhike’s annual provision for contingent liabilities was 23.66 billion yuan, up 172% year over year. Even considering only the single quarter of last year’s fourth quarter, the provision reached 11.10 billion yuan. Coupled with the downtrend in loan assistance fee rates, an increase in asset impairment provisions, fluctuations in the fair value of crypto assets, and other factors, profitability ultimately dropped significantly.

Strengthening AI empowerment

From the financial reports, under the impact of the new loan assistance regulations, artificial intelligence (AI) remains an important lever for loan assistance fintech companies to reduce costs and improve efficiency and performance.

It is reported that Qifu Technology’s AI approval officer, leveraging multimodal technology, can automatically identify credit review materials submitted by customer managers, issue real-time reminders for corrections and optimizations, and significantly improve the first-pass rate of materials. In the credit review and approval process, the accuracy of identifying substantive risks has reached more than 90%. On the business side, Qifu’s FocusPro solution, oriented toward serving banks’ independent risk control and helping banks build core capabilities, helped partner banks increase loan disbursement volume by approximately 448% year over year in 2025, with an end-of-year balance of 11.8 billion yuan.

In 2025, Xinye Technology’s R&D expenditure increased by 8% year over year, and in the fourth quarter it introduced AI intelligent agents into the service application process. The agent is directly embedded into the application flow, breaking complex processes into clear logical steps and providing real-time guidance. Xinye Technology also launched a large-model-native multi-agent (Multi-agent) dialogue engine, promoting intelligent upgrades of customer service systems and seat-assistance. On the coordinated underlying data architecture, it successfully deployed the domestic standard data middle platform into overseas new markets; more than 70 tasks are now live.

According to the financial report, in the fourth quarter of last year, Lexin’s AI large model and intelligent agent technologies have evolved from auxiliary tools into core capabilities, deeply embedded in key financial scenarios such as risk control, customer service, and operations. Among them, intelligent upgrades in the risk control domain were the focus of Lexin’s technology investment in the fourth quarter. In the fourth quarter of last year, Lexin integrated large-model cognition, reasoning, judgment, and intelligent agent execution capabilities, consolidating AI core capabilities on the user side. This drove the user experience from passive response to active forward-looking prediction, improving user experience. Large models and AI intelligent agents have been deployed across multiple risk-control stages, enhancing both risk-control efficiency and identification accuracy.

JY (JiaYin) Technology achieved phased results in 2025 across directions including multimodal anti-fraud, AI seat assistance, and data intelligence. In 2026, JY (JiaYin) Technology will continue to push key upgrades around the “4+2” strategy. It plans to restructure the existing four product matrices into two main lines: “production lines” and “non-production lines.” The production lines focus on creating core business value, covering three major areas: intelligent customer acquisition, intelligent risk control, and intelligent marketing. It will continue to explore identification and acquisition of high-quality customer segments driven by AI, deepen applications of multimodal technologies such as voiceprints and knowledge graphs in anti-fraud scenarios, and enable intelligent generation of marketing content and compliance review.

Tan Ning, Chairman of the Board and CEO of Yiren Zhike, said: “2025 is Yiren Zhike’s 10th anniversary since listing and a key year for the implementation of the company’s AI strategy. We place technology for good at the core and drive AI to evolve from an efficiency tool into a business collaborator with warmth. Through the full filing of the ZhiYu (Smart Language) large model and the iterative upgrade of the Magic Cube platform 2.0, we have achieved intelligent upgrades across the full chain of customer acquisition, risk control, and operations. In the future, we will continue to deepen AI services, empower industry development with technological innovation, and create more valuable digital intelligent service experiences for users.”

Wang Pengbo, a senior analyst at Broadcom Consulting who analyzed for reporters, said that multi-traffic channels for customer acquisition, emerging technology applications, and other factors have become important levers for future development of fintech companies. The enablement of artificial intelligence on performance has been fully implemented: it improves efficiency across the full business chain, including risk control, marketing, and customer service. Leading companies have achieved large-scale, scalable applications of AI by optimizing technology to reduce operating costs and improve service efficiency.

“Going global” to find growth opportunities

Against the backdrop of continuously rising domestic entry thresholds and compliance regulatory standards, loan assistance fintech companies further increase their overseas market layout to find new growth space.

With domestic business contracting, overseas businesses become a new growth engine. In 2025, Xinye Technology’s full-year international transaction volume was 14 billion yuan, up 38.6%. In the fourth quarter of 2025, international borrowers reached 3.8 million, up 133.8%; new borrowers were 1.6 million, up 117.3%. The international market transaction volume was 4.1 billion yuan, up 41.4%. The outstanding international loan balance was 2.6 billion yuan, up 52.9%. International business revenue was 950.9 million yuan, up 28.6%, accounting for 31.4% of total revenue, hitting a historical high.

The financial report shows that Lexin and JY (JiaYin) Technology have both opened up overseas markets. In Lexin’s case, the strategic layout in Mexico and Indonesia—two key core markets—became evident last quarter. Scale, profitability, and asset quality were optimized in parallel, making them a new driving force for company development.

In 2025, JY (JiaYin) Technology’s business scale in Indonesia grew by approximately 187% year over year, while registered user numbers grew by approximately 119% year over year, gradually revealing the scale effect. Mexico’s business accelerated significantly in the fourth quarter of last year. The total disbursed loan amount grew by approximately 105% year over year, and the number of registered users grew by approximately 110%. The business model completed a stage of validation. In the future, it will continue to explore other market opportunities. Through cross-region and cross-cycle layout, it will steadily advance the expansion of its overseas business footprint.

In addition, Qifu Technology is also increasing resource investment in overseas markets. Qifu Technology CEO Wu Haisheng said: “While making necessary adjustments to our domestic business, we continue to strengthen our overseas expansion efforts. In the fourth quarter of 2025, we increased investment in overseas resources, secured the necessary qualifications in local markets, and built local teams, expanding our business footprint to multiple developing-country markets. In the long run, we believe these overseas markets will not only bring substantial growth opportunities, but also help us diversify the geographic structure of our business.”

Wang Pengbo said that in some overseas regions, credit penetration is relatively low while digital financial demand is rising rapidly, which can provide institutions with incremental markets and new growth curves. Leading institutions can export domestically mature risk control technologies, credit models, and operational capabilities to form differentiated competitive advantages. Meanwhile, international expansion helps diversify the risks associated with a single market.

Wang Pengbo believes that overseas markets will remain an important incremental direction for listed fintech companies. He expects that they will further increase their level of deployment, focusing on overseas regions with development potential. At the same time, they need to respond to risks such as changes in local regulatory policies and intensifying market competition. The challenges stem from the fact that financial regulatory systems, licensing and permit qualifications, credit environments, and user habits vary significantly across countries and regions. Compliance entry and localized operating costs are also high. Risk-control models need to be re-adapted to local credit data and default characteristics, and companies also face competition from local financial institutions, international payment and credit platforms. Cross-border capital management, building talent teams, and policy changes will all test business stability.

Reporter Yu Jichao

Text editor Yao Hui