New regulations for loan assistance implemented, fintech companies' three strategies for 2025: comply with regulations and uphold the bottom line, expand overseas to capture growth, AI becomes a moat

Ask AI · How do the new regulations on loan assistance promote the fintech industry’s shift toward quality improvement?

Our newspaper (chinatimes.net.cn) reporter Fu Le reports from Beijing

As of April 1, major listed fintech companies have completed disclosures of their Q4 2025 and full-year results. With the implementation of the new regulations on loan assistance overlapping with uncertainty in the macro environment, the fintech industry has generally been under pressure. Many platforms have proactively tightened risk control and shifted toward prudent operations. At the same time, technological innovation continues to empower core businesses. Overseas markets have become an important growth engine, and the industry is moving from scale expansion to quality improvement, efficiency gains, and global deployment.

On April 2, Wang Pengbo, Chief Analyst at Broadcom Consulting, told reporters from Huaxia Times that in Q4, fintech companies were mainly still affected by the concentrated rollout of regulatory policies. This includes regulations on interest-and-fee structure, funding contribution ratios, and data compliance. As a result, most companies proactively adjusted their business cadence, causing both scale and profits to face pressure in the quarter. Overall, it shows a phased characteristic of “growing revenue without growing profits,” which also reflects that the industry is shifting from extensive expansion to a development path that places greater emphasis on quality.

Risk control continues to tighten

Lixin’s financial report shows that in Q4, the company’s revenue was RMB 3 billion, and profit (Non-GAAP EBIT) was RMB 299 million. For the full year of 2025, the company’s profit (Non-GAAP EBIT) was RMB 2.216 billion.

In 2025, Xinyeshike achieved annual operating revenue of RMB 13.57 billion and net profit of RMB 2.55 billion. In terms of operating indicators, the total transaction volume facilitated reached RMB 200.3 billion, and the loan balance was RMB 70.9 billion. In Q4 alone, Xinyeshike’s revenue was RMB 3.02 billion, net profit was RMB 420 million, and the transaction volume facilitated in that quarter was RMB 42.8 billion.

For the full year of 2025, Qifu Technology achieved revenue of RMB 19.2 billion, and Non-GAAP net profit reached RMB 6.35 billion, down 1.0% year over year.

Jiayin Technology’s total loan-matching facilitation transaction volume for full-year 2025 was approximately RMB 129 billion, realizing operating revenue of about RMB 6.22 billion. Of that, in Q4, the facilitated loan-matching transaction volume was RMB 24.2 billion, with operating revenue of about RMB 1.09 billion.

During the period, Weixin Jinke recorded total revenue of RMB 3.871 billion, and its full-year loan facilitation scale in Mainland China was RMB 58.45 billion.

Yiren Zhike’s full-year revenue was RMB 5.72 billion, down 1% year over year, and net profit was RMB 40.5 million.

Overall, the industry generally shows a situation of “growing revenue without growing profits.” The revenue and transaction growth rates of leading platforms have clearly slowed. The extensive expansion model in the past is no longer sustainable. As a result, many institutions have proactively reduced lending scale, raised credit approval thresholds, lowered product pricing, and adopted prudent operating strategies. They also strengthen risk control by focusing on high-quality customer segments, implementing refined management, and controlling costs—keeping risks within a controllable range.

In the opening year of the “15th Five-Year Plan,” the fintech industry is entering a critical transition period for quality improvement. The new regulations on loan assistance will take effect starting October 1, 2025. It is clearly stated that the credit enhancement service fee must be included in borrowers’ comprehensive financing costs; the range of comprehensive financing costs is clearly specified. It also clarifies that platform operating entities may not, in any form, charge borrowers interest and fees, and credit enhancement service entities may not indirectly increase credit enhancement service fee rates in forms such as consultation fees or advisory fees.

“Under the constraints of the new regulations, the industry generally faces multiple challenges, including a narrowing interest-and-fee space, rigid funding costs, and rising compliance investment,” Wang Pengbo said. He noted that to address risks moving upward, platforms have been tightening entry thresholds, optimizing credit granting strategies, iterating product formats one after another, and strengthening collections management to stabilize asset quality.

This trend of industry adjustment is also reflected at the level of institutional cooperation. As cooperation access standards related to the industry are gradually advanced, resources are accelerating toward major players. After reviewing the lists of cooperative institutions publicly released by several small and medium-sized banks recently, the reporter found that leading institutions have become the primary choice for bank cooperation layouts. Small institutions are excluded from the cooperation scope because they lack sufficient resource reserves and have incomplete risk-control systems, making them unable to meet bank cooperation entry standards.

The continuous improvement of policies has set clear development boundaries for the industry, and it also forces institutions to shift from scale expansion to refined operations.

Lixin CEO Xiao Wenjie said that in Q4, the new regulations were officially implemented. Lixin adheres to compliance as guidance to drive business development, optimizes its product matrix and personalized service experience centered on users, and effectively enhances sustainable operations and risk-resilience capabilities.

Zheng Yan, Chief Risk Officer of Qifu Technology, said that due to the impact of the external environment, the level of asset risk rose in Q4. The company has tightened risk control, and the risk profile of newly issued loans in recent periods has shown marginal improvement.

Yan Dinggui, Chairman of Jiayin Technology, believes that given that uncertainty in the macro environment still exists, the company maintains a prudent stance. It will continue to inject innovation into development momentum and build operational resilience to withstand cyclical volatility.

Open up overseas growth space

As the domestic industry continues to adjust, more fintech enterprises are going overseas to optimize business structures and tap new growth opportunities. In 2025, companies such as Lixin, Xinyeshike, and Qifu Technology all accelerated their overseas deployment. In markets such as Mexico and Indonesia, both scale and profitability improved in tandem, and some enterprises further expanded into developed markets such as Australia.

In terms of overseas business, during the quarter, Lixin’s strategic deployment in Mexico and Indonesia—its two major core markets—showed results. Scale, profitability, and asset quality were improved simultaneously, becoming a new driver for the company’s development.

Xinyeshike’s internationalized business has also maintained strong momentum. In Q4, international revenue accounted for 31.4% of total revenue. In 2025, Xinyeshike’s active users in the Indonesia and Philippines markets doubled to reach 5.9 million. By launching diversified customized products tailored to local consumption preferences, such as “buy now, pay later,” Xinyeshike further deepened its presence in overseas markets. In Q4, Xinyeshike officially rolled out its Australia market strategy, taking a key step toward developed markets.

In 2026, Qifu Technology will begin exploration and trials in multiple overseas markets in parallel, accelerating its overseas market deployment. Wu Haisheng, CEO of Qifu Technology, believes that macro uncertainty in Q4 led to tighter industry liquidity and rising risks. The company responds proactively by enhancing risk control and adjusting its business and cost structures, and it will continue to develop overseas markets, promoting diversification of its business structure.

In terms of overseas business, in 2025, Jiayin Technology’s business scale in Indonesia grew by about 187% year over year, and the number of registered users grew by about 119% year over year. Its Mexico business accelerated significantly in Q4. For the full year, total loan disbursements grew by about 105% year over year, completing a phased validation of its commercial model.

At present, fintech’s overseas expansion is no longer just sporadic trial runs. It has entered a new stage of large-scale, refined operations. Overseas business is becoming an important support for enterprises to hedge industry cycles and achieve steady development.

Wang Pengbo said that in 2026, the industry’s overall scale may continue to trend toward a mild contraction. However, there are still structural opportunities in sub-segments. Some platforms explore new growth paths through overseas deployment, scenario integration, or technology services. To achieve stability in performance and a recovery, the key is to balance the relationship among compliance, risk, and returns. On one hand, deepen cooperation with licensed institutions to ensure funding stability; on the other hand, enhance user value through refined operations and continuously optimize asset structure to improve long-term profitability.

Intelligent-agent scale deployment

Since 2025, leading platforms have accelerated the deployment of self-developed large models and intelligent agent platforms. They have applied AI technology comprehensively to core scenarios such as risk control, underwriting/approval, and customer service, improving decision-making efficiency and the precision of risk control.

According to financial reports, in Q4, Lixin’s AI large-model and intelligent-agent technologies have moved from auxiliary tools to core capabilities, deeply embedded in key financial scenarios such as risk control, customer service, and operations. Among them, the intelligent upgrade in the risk-control area was the key focus of Lixin’s technology investment in Q4. In Q4, Lixin integrated the large model’s capabilities in cognition, reasoning, and judgment with the execution capabilities of intelligent agents, consolidating AI core capabilities to the user side. This drives user interaction from passive responses to proactive forward-looking anticipation and improves user experience. During the quarter, large models and AI intelligent agents were implemented across multiple risk-control steps, improving risk-control efficiency and recognition accuracy.

In 2025, Xinyeshike’s R&D expenditure increased by 8% year over year. In Q4, Xinyeshike introduced AI intelligent agents into the service application process. These intelligent agents are directly embedded in the application flow, breaking complex procedures into clear logical steps and providing real-time guidance. Through this experience upgrade, both the user application completion rate and conversion rate have been substantially improved.

Qifu Technology has developed two core intelligent agents that have been deployed and applied in multiple scenarios. Qifu Technology’s AI approval officer leverages multimodal technology to automatically recognize the credit-review materials submitted by customer managers, provide real-time reminders for corrections and optimization, and improve the first-pass rate of submitted materials. The AI loan officer focuses on the marketing side, providing frontline customer managers with end-to-end marketing intelligent-agent support. On the business side, Qifu Tech’s FocusPro solution helped partner banks increase loan disbursement volumes by about 448% year over year in 2025, and end-of-year balances reached RMB 11.8 billion.

Jiayin Technology uses its intelligent-agent platform and machine-learning platform as two major technology foundations. It explores the identification and acquisition of high-quality customer groups driven by AI, deepens multimodal technology applications such as voiceprint and knowledge graphs in anti-fraud scenarios, and empowers intelligent generation of marketing content as well as compliant review.

In 2025, Yiren Zhike’s Zhiyu large model and Magic Cube AI intelligent agent platform-based technical system were fully rolled out. Through the comprehensive filing/registration of the Zhiyu large model and the iterative upgrade of Magic Cube platform 2.0, Yiren Zhike achieved intelligent upgrades across the full chain of customer acquisition, risk control, and operations.

Byline: Feng Yingzi | Editor-in-Chief: Zhang Zhiwei