Yixin Group Limited (2858.HK) Earnings Call Transcript & Summary

Dear investors and analysts, good evening to you all. Welcome to Yixin Group's 2025 Interim Results Conference. Our conference will be held via online audio live stream and conference call. Please allow me to introduce the company's management team attending this results conference. They are Chairman, Mr. Zhang, Xuan; Co-President, Mr. Gao, Zhi; Chief Financial Officer, Mr. Yang, Xiaoguang; Chief Technology Officer, Mr. Jia, Zhifeng. This conference will consist of 2 sessions. First, we will review the company's development through a video, and then, the company's management will elaborate on the 2025 results as well as the Q&A session. Now, let's move on to the first session. We'd like to invite Mr. Zhang, Xuan, Chairman of the company to deliver a speech.

Dear shareholders and investor friends, hello, everyone. I'm Zhang, Xuan, Chairman of Yixin Group. Welcome to Yixin Group's 2025 Half Year Results Conference. On behalf of the company, I'd like to thank you for your concern and support for Yixin. Today's conference will mainly consist of the following parts. First, I will introduce the company's overview and industry development. And then, Mr. Gao, Zhi, our CEO, will introduce the highlights of Yixin's business in the first half of 2025. Next, Mr. Zhifeng, Jia, CTO will introduce the company's core competitiveness and AI programs. And then, our CFO, Yang, Xiaoguang, will introduce the company's financial performance in the first half of 2025 and the development of our overseas business. And then Mr. Zhang, Xuan will introduce our future plans for AI agent. And finally, we will leave some time for the Q&A session. First, let's take a look at the basic situation of Yixin. As an AI-driven fintech platform, Yixin has built a new car plus used car dual well-driven business structure and has always taken technology-driven financial innovation as the main access of strategic development. On the asset side, we have more than 42,000 partners covering all provinces in Mainland China. We have also cooperated with more than 100 financial institutions, including banks, trusts, financial leasing companies on the capital side. We maintain in-depth collaboration with our major shareholder, Tencent, carrying our tech collaboration in cutting-edge flows such as big data, AI and cloud computing. Our industry-leading risk management system ensures the safety of the assets and our management. These comprehensive competitiveness have built a moat for the company's long-term development. It is worth mentioning that with the company's continuous deepening in the auto finance world, by February 2025, the cumulative automobile financing transaction amount has exceeded CNY 400 billion, which demonstrates Yixin's leading strength. Next, let's look at the overall situation of the industry. In the first half of 2025, Chinese passenger car market showed a steady trend and industry competition remains fierce. According to data from the China Association of Automobile Manufacturers, new passenger car sales in the first half this year reached 13.53 million units, a year-on-year increase of 13%. According to statistics from the China Automobile Dealers Association, the transaction volume of used passenger cars in the first half of 2025 reached 7.57 million units, a year-on-year increase of 0.5%. China's automobile market continued to move forward in the new stage of development, driven by intelligence and electrification. According to the China Passenger Car Association, new energy passenger cars maintained rapid growth with a year-on-year increase of about 33.3%. Meanwhile, the penetration rate of new energy terminal retail has exceeded 50% for several consecutive months. In the field of intelligent driving, with the in-depth empowerment of cutting-edge technologies, such as AI, big data and 5G in the autonomous-driven sold coupled with a greater improvement of relevant regulations and cost optimization brought by technological progress. The intelligent driving field ushering a new stage, the penetration rates of high-level autonomous driving continue to rise rapidly. And the outlook for the second half of the year, we can see the intensified industry competition has also attracted attention to anti-involution. With the increase in regulatory policies to rectify the evolution in this sector, the industry is expected to return to rational competition. Enterprises will devote more into tech innovation, improving product quality and optimizing aftersales service and other denial development paths that's promoting the sustainable development of the industry. In addition, the increased intensity of the 2 new policies has laid a foundation for the market to further recover driven by new energy and continue to release growth momentum. The popularization of high-level autonomous driving is accelerating, injecting new momentum into the high-quality development of the industry. Next is the overall situation of the auto finance and fintech market. First, the auto finance industry. According to third-party data, China's auto finance market is about CNY 3 trillion. In terms of policy, the government continues to pay attention to the development of the sold, such as encouraging financial institutions to provide more credit support for consumption through policies. However, compared with developed markets, there is still a gap in the financial penetration rate of the Chinese market. With the deepening of the industry's anti-involution action, the market is accelerating the clearance, and the pattern of the strong getting stronger with resources concentrated in leading enterprises has become more obvious. Yixin will continue to focus on core capacity building, rely on technical advantages and in-depth scenario cultivation and consolidate its leading position in the auto finance world. In terms of fintech, as an important starting point for the high-quality development, fintech can help release consumption potential and empower the development of new productive forces. This field is also supported by policies. The Central Bank has issued a series of policies to promote digital finance, further supporting the deep integration of fintech and traditional finance. This has laid a solid policy foundation for the future growth in the next few years. Third-party data shows that the scale of the fintech market will reach about CNY 800 billion in 2025. Users' demand for convenience and personalized fintech services has promoted financial products towards scenario-based and lightweight creating 2-way value increment for the industry. At present, fintech is in the period of policy, dividend plus technological iteration plus demand exploration. Yixin will rely on its technical accumulation to further explore the application of fintech in vertical field. Next up, please ask our CEO, Mr. Gao, to introduce our business development in the first half of this year.

In the first half of 2025, facing the complex external environment, our business scale maintained steady growth. The main operating highlights of our company are as follows: the core main business developed steadily, seizing structural growth opportunities. In the first half this year, the automobile financing transaction reached CNY 32.7 billion, a year-on-year increase of 4%. Among them, used cars grew rapidly on our strategic focus with a year-on-year growth rate of 31%. Our Fintech business achieved explosive growth with the financing scale facilitated in the first half of 2025, reaching CNY 1.53 billion. Next, I will introduce some developments in our business, specifically. First, our core business, automobile financing services. In the first half this year, the total transaction volume reached 364,000 units, a year-on-year increase of 11%. In terms of new cars, the transaction volume in the first half of 2025 was 142,000 units, with a financing amount of CNY 14.5 billion. The slight decline in new cars is mainly because we preferentially captured more profitable financing businesses. Our group continues to deepen cooperation with manufacturers, the number of total-to-total cooperations with automakers and dealers has reached nearly 50. A special team has been set up to expand the collaboration. Meanwhile, production innovation has been promoted. New value products, jointly developed with leading manufacturers, have been piloted. In addition, we have further improved the accuracy and efficiency of services through system upgrades such as optimizing standardized processes and streamlining application sold. In terms of used cars, the transaction volume in the first half of this year was 222,000 units with a financing amount of CNY 1.82 billion, a year-on-year increase of 31%. This growth is mainly due to our active expansion of customer group in the market. We launched the risk pricing product, which directly provides approval plans according to customer qualification. This product has effectively improved the overall approval rate and helped us capture more market share in the used car market. In terms of new energy, the growth of our new energy auto finance business remains significant. The transaction volume of new energy vehicles increased by 33.8% year-on-year to 93,000 units, and the financing amount increased to CNY 9.5 billion. In the first half this year, the financing amount of new energy vehicles accounted for about 52% of the total. At the same time, the rapid growth of the ownership of new energy vehicles has also promoted new energy used cars development. The proportion of used car transactions in Yixin's new energy business has increased to 22%. Next is the Fintech business. Since the official launch of Fintech in 2022, this sector has developed rapidly. It reached CNY 1.53 billion in the first half this year, a year-on-year increase of 58%. Second, revenue achieved a sustained explosive growth, reaching CNY 187 million this year -- in the second half of this year -- in the first half of this year with a year-on-year increase of 124%. Technology empowerment promotes new energy penetration rate. Under the Fintech business model, the penetration rate of new energy in new cars is about 54%. Number four, deepened collaboration to expand business territory, our company's Fintech business has covered multiple institutional partners such as banks, financial leasing companies and automakers with simultaneous expansion of corporation breadth and depth. In the continuous development of fintech, we've always focused on high-quality customers and deepened collaboration, achieving remarkable results. Our core customer group has steadily expanded with the number of core customers increasing from 12 in the first half of 2024 to 15 in the first half of 2025. At the same time, the revenue contribution of core customers has further increased. The average revenue of core customers in the first half of 2025 reached CNY 123 million, accounting for 98% of the total revenue. This data reflects the effectiveness of our customer operation, but also verifies the scale cost advantage of the company expanding service boundaries driven by the data flywheel. Our Fintech business continues to develop in 2 modes: pure technology and technology plus traffic. Pure technology product's output, for example, the risk control model, we provide for well-known automakers. Those only screen high-quality customers, but also reduces fraud risks. In the first half of this year, under the technology mode, the facilitated financing transaction amount reached CNY 503 million, year-on-year increase of 27%. In terms of technology plus traffic services, while maintaining transactions, relying on our channel and data advantages through tech empowerment. We have also enhanced our overall business value, providing product and risk control management support. In the first half of 2025, the financing scale facilitated under this mode reached CNY 14.8 billion, a year-on-year increase of about 60%. In general, the scale effect of Fintech sector is accelerating. In the future, it will rely on cutting-edge technologies to deepen scenario applications and provide comprehensive tech empowerment for institutional partners in the auto finance sold. Next, let's look at value-added services. Relying on cooperation with large domestic insurance institutions, our value-added services provide diversified products such as cars used rights and supplementary insurance around customer needs, aiming to improve user experience and enhance customer loyalty and are important part of our business ecosystem. It has shown steady business growth. The revenue of value-added services in the first half of this year reached CNY 140 million, a slight year-on-year decline, mainly due to the decline in the transaction volume of new car products with higher profitability, along with the overall decline in new car transaction volume. The penetration rates of the core product in value-added services, the rights package continues to increase with the volume reaching 118 -- 189,000 units, a year-on-year increase of 11%. In the first half of this year, the transaction volume of battery car products reached 31,000 units, and they continue to expand in the new energy used car market, further meeting the multilevel market demand. In the future, we will continue to optimize the value-added service product system, deepen cooperation with insurance institutions, expand more high potential value-added product, improve user value and promote the growth of our business. Next, CTO, Mr. Jia, will introduce our core competitiveness and the group's progress in AI.

Thank you. The company's core competitiveness lies in building an integrated risk control management system covering all scenarios and the entire chain and using AI technology throughout the pre- and post-financing links to achieve both precise control and improve the business efficiency. In the customer acquisition and application submission links, we've realized dual engine customer acquisition alliance offline conducted intention certification through AI, selected users from diversified channels and implemented front-end dynamic monitoring through the preaudit link to effectively identify and intercept high-risk customers, controlling business quality from the source. In the credit evaluation mix, we rely on the self-developed credit evaluation system. We integrate traditional machine learning and multimodal large model technology to build accurate algorithm models. Combined with the company's huge data assets, we focus on deepening our analysis of customers' credit history and the risk characteristics to achieve precise matching of financial plans under the evaluation of people plus cars. In the post-loan asset management link, our asset center in Shanghai implements full life cycle management users reasoning large models to build an AI asset management brand, breaking the traditional aging management model and formulating the most appropriate response plans for assets in different stages and states. Meanwhile, we've established an asset auction platform to optimize asset value recovery to the greatest extent. In February 2025, the company took the lead in completing the local deployment of DeepSeek technology, becoming the first enterprise in the auto finance world to fully implement this business technology. Joining on the training experience and learning strategies of DeepSeek R1, we've transformed the self-developed vertical domain launch model and reduce the model operation cost through the MOE architecture and reasoning acceleration technology. The transformation has achieved a remarkable result. Important decision-making scenarios, such as risk assessment and asset management, 100% of customers are matched with the optimal financial plan, and both approval efficiency and accuracy have increased by more than 50%. Through the system, we've insured asset safety and also laid a solid foundation for the healthy development of the business. On the basis of continuous optimization of the full chain risk control management system, we've always taken technology plus scenario as a core driving force and made every effort to promote the transformation of our capabilities into actual productivity. In the first half of this year, Yixin began to try the application of Agentic AI in finance. In practice, we found that technology has reached a 12 years' breakthrough point, which may bring a new rounds of updates to our business paradigms. First, the reasoning models are close to the level of human experts in decision-making and judgment. Second, multimodal large models are more efficient than ordinary humans at a perception level. The significance can be explained through a specific scenario application. In the customer acquisition link, while AI process multimodal conversations with customers in real-time, it cannot only adjust the conversation according to the context to complete the information collection, but also link the product module to recommend the optimal plan. Meanwhile, it can collaborate with the risk control anti-fraud module and intimate and initiate risk inquiries in the customer acquisition stage. It can realize the prepositioning of anti-fraud and also synchronizes the full process compensation information to the risk control module so that risk judgment is no longer limited to a structure filled with credit reports and applications. Signals such as tone, speech, expression and language fluency that are missing in traditional models can be integrated into our model through raw data, improving the accuracy of judgment. We believe that such a model can help build the development direction of the financial credit fraud in the future. In the past, a single AI model could only cover the capabilities of a single link and needed to be connected through complex systems. Now, multimodal agents can independently and flexibly core various capabilities to provide end-to-end overall value. So through interaction with customers and environment, it can realize dynamic decision-making and process judgment and rely on AI's autonomous learning ability to promote the improvement of efficiency, scale and business form, ultimately improving industry efficiency. In the future strategy session, I will continue to elaborate on our plans and goals for the business implementation of Agentic AI. So we plan to officially release the first Agentic large model in the auto finance world in the second half of 2025, integrating reasoning, multimodal voice, domain-specific model metrics and tools that are supporting the models to realize dynamic decision-making and multi-agent collaboration. Achieved significant improvements in customer experience, thereby increasing the penetration rates of our AI in the entire industry. Next, the CFO, Mr. Yang, will introduce our financial performance and the development of our overseas business.

The main financial highlights of our company during the reporting period include the following aspects. Operating income maintained rapid growth, reaching CNY 5.45 billion in the first half of 2025, a year-on-year increase of 22%. Operational efficiency continued to improve with the ratio of sales management, R&D and other expenses to income dropping to 20%. Profitability continued to strengthen with net profit reaching CNY 550 million in the first half of this year, a year-on-year increase of 34%. Asset management scale grew steadily, reaching CNY 112.1 billion as of the end of June, an increase of 12% compared with the same period last year. Next, I will further introduce some key financial indicators. So first, let's break down our revenue structure, and you can see a fast improvement in our management and operation. And now, let's break down the revenue structure in more details. First, as we mentioned, that our operating income maintained rapid growth, reaching CNY 5.45 billion, a year-on-year increase of 22%. So in terms of our revenue breakdown, you can see our platform income, it increased by 24% year-on-year, accounting for 79%, among which the fintech business continued to grow at a high speed with its revenue proportion rising to 34%, becoming the largest business segment. And our self-operated income increased by 60% year-on-year, mainly due to the growth of the scale of financial leasing assets during the period. Next, let's look at the growth of the company's gross profit and its driving factors. In the first half of this year, the net interest margin of leased assets increased significantly year-on-year, mainly because the capital cost rate continued to decline. The average capital cost decreased by 80%, opening up space for the net interest margin. The yield on the asset side is relatively stable. The adjusted yield is yield before deducting the channel fee. In terms of platform business, in the first half of this year, the net service rate of platform financing further increased, mainly driven by 2 aspects. The main reason similar to the self-operated business is the benefit of the decline in the implied settlement interest rate of cooperative fund base. In addition, it is due to the successful switching of our business structure. The proportion of used cars in platform financing services further increased to 56%. In particular, cooperation with financial leasing companies has led to a further increase in the proportion of our long-tail high-priced used car customers. And next are some changes in terms of expenses. The total sales management and R&D expenses of our company increased by about 12%, mainly due to our investment in talent, software and hardware in AI, overseas and other facts. However, due to the further release of [indiscernible], the overall operating expense ratio continued to decrease. It has dropped to about 20% in the first half of this year. The ratio of credit impairment loss to income has also increased. There are several reasons behind. And later, we will go through our vintage performance. Also, in terms of the macro uncertainties, we have taken a prudent judgment on the external environment, which is also reflected in the provision coverage later. In our ECL measurement model, the relevant parameters for potential external adverse factors in the future are set relatively conservatively. And let's introduce our profit situation. Our profitability continues to strengthen with continuous rapid growth in recent reporting periods. The IFRS net profit in the first half of this year reached CNY 550 million. If considering adjustments for some noncash expenses, the non-IFRS net profit reached CNY 650 million. In general, we achieved the expected results, and we'll continue to work hard in the second half of the year to strive for a good performance for the whole year. Now let's look at the asset liability situation. On the capital side, the company emphasizes the safety and stability of capital sources and uses various means to promote the continuous decline of financing costs. As of the end of June this year, more than 100 financial institutions have been cooperated with and the number of bond and asset securitization investors has exceeded 60, among which foreign institution investors has exceeded 10, including Sumitomo Mitsui, JPMorgan Chase and et cetera. And in the future, in terms of financing, standardized products continue to be welcomed by the market, and our high-quality automobile assets are still relatively scarce in the market. The comprehensive interest rate cost of some recent structured products has come to less than 3%. And in terms of issuing credit bond varieties, the company has continued to promote the development of diversified financing channels, so you can see our structure and our costs have been further optimized compared with last year. In the future, we will continue to explore and develop to maintain the strong support of the liability side for the business. Next, on the asset side, you can see the company's asset management scale has grown rapidly in recent years. And in terms of asset quality, we have also shown steady progress as well. And our provincial -- our provision coverage ratio reached 207% at the end of the period, showing reasonable coverage of nonperforming assets. And next, I would like to share with you our liquidity situation. Yixin has always had high requirements for liquidity management. And our asset liability ratio has been maintained within a relatively healthy range. And finally, I would like to report on our asset vintage performance. In 2020 and 2021, we already had a good performance, but you can see our vintage delinquency rates of new car assets have remained at a desirable level, significantly better than those generated during the period such as 2020 and 2021. So now, in the past 3 years, our performance has been significantly better than 2020 and 2021. And we have maintained a good predictability in terms of used cars as well. And at present, the curves of assets in recent years have risen somewhat, but this is mainly because after the epidemic, we moderately expanded our risk appetite in the used cars sold. And now, we have more long-tail customers, and this upward trend has not actually exceeded the past range. We will continue to optimize the migration rate at all stages and reduce the life cycle loss in the subsequent asset recovery link. And finally, I would like to briefly introduce the company's overseas progress. So Singapore has been our first lending country. It is the first market that we've launched our business. It saw its revenue increase by 400% year-on-year in the first half of 2025 and achieved profitability. Its financial services cover nearly half of local used car dealers, ranking among the top 3 nonbank service providers. The launched export dealer SaaS system helps car dealers realize inventory digitalization and intelligent connection of financial channels, becoming an important carrier of technology export. With Singapore as a great platform as a fulcrum, we will accelerate global resource allocation and risk management driven by finance plus technology. In the first half of 2025, by focusing on segmented groups, we entered other markets and launched a differentiated auto finance products. So in the second half of this year, we plan to replicate this model to expand into other Southeast Asian markets. So in overseas practice, we have summarized some of our key capabilities that can be replicated to other markets. Next, let's hand over the time to Michael to introduce our group's AI strategic planning.

Next, I will introduce the strategic planning of Agentic AI for the full chain of auto finance. First, regarding the prefinancing link, we plan to fully implement agentization in the prefinancing link in the second half of the year. By deploying preaudit agents and telemarketing assistant agents, we will realize the automation of pre-loan processes. The initial verification of core results shows that customer communication time can be shortened by 50% and fraud identification coverage could also be greatly improved. Manual intervention in a single link could be reduced by half, and the conversion rate is expected to increase by 10%. Then, regarding the end-to-end AI risk control in the financing process, we will achieve initial implementation in 2026. The core breakthroughs are in 2 aspects. First, upgrade the existing risk control system with full model end-to-end AI effectively, reducing information loss in intermediate links of traditional path and improving the generalization ability of risk control models so as to identify and respond to more long-tail risk cases. The full model coverage will exceed 50% of customer groups. Second, the scope of differential pricing will continue to expand, covering 80% of customer groups and improving customer acceptance through dynamic plan adjustments. The short-term goal is to increase the approval pass rate of risk controllable customer groups by 20%. Primarily focusing on the post-financial link, we will fully implement agentization in 2026. We'll extend the asset management brain to the interested collection scenario, reduce the amount of interested collection for risk customers, predicted to be 5 minus. And the MI collection rate is expected to increase by 15%. This has been experimentally verified and will increase the collection amount. From the perspective of the full chain, it's expected that the pre-loan AI penetration rate will reach 95% in the short term. The post-loan rate will reach 60% with a long-term goal of over 90% intelligence in the entire business chain. This will reduce the cost per transaction and greatly improve scale efficiency. In the third of technical IP, we continue to apply for more new AI invention patents. 18 new patents were applied for in the first half of this year, and it is planned to open source the general basic model of pre-loan agents around Q3 2025, promoting partners to realize model capability sharing through training data upload and building a core operative ecosystem. Thank you very much for listening.

Now we hand it over to the moderator. We can now move on to the Q&A session.

[Operator Instructions] Now let's take the first question. It's from Chinese investor from Guotai Haitong Securities.

I'm analyst from Guotai Securities. Congratulations on the good performance. I'd like to ask you the question. So you have seen a faster progress, and the transformation has been very successful. So I'd like to know how do you reach this transformation? What's the core reason behind? And what's your plan for the future, whether you would increase the proportion of certain business segments?

So let me answer your question. From our perspective, in Yixin Group, we discussed with our investors many times. We hope to transform our model -- our business model to a more lightweight one so that our value can be better fulfilled so that our quality can be further improved. So that's our main reason. And in terms of external development, so we can notice that on the side of the financial institutions, there is more momentum to acquire the automobile assets. We can see the banks and financial institutions have collaborated with us for many years, so they have good operation foundation. They've also accumulated data along the process. So they see the ability in our tech performance and asset management so that they are willing to pilot the new project with us together. And also, on the side of banks and financial institutions, there's also pressure from the regulation today, banks and financial institutions want to have more independent development. So we cannot -- so we do not provide the full -- we do -- so in terms of the external environment and our own capabilities, we think we have the advantages. We are at the right position for the transformation.

So let me try to add one point. So in terms of our strategic position, so we have Yixin 1.0, 2.0; 2.0 from heavyweight -- heavy assets to self-operated to lightweight models. So we've been making this transformation at our interim result conference and our annual results conference. So we've been updating our strategies, and we've been introducing AI capabilities. So in the near future, we are going to further empower our business with AI capabilities to have more module-based AI capabilities. So we have had much experience, and we have our private data. The quality and quantity are both very high. So our model has also been widely accepted by our partners. Therefore, we hope in the near future, as we continue to move on, while we collaborate with our partners, we can provide more AI module services. So we hope we can have more collaboration in this regard. And also, we hope that we can continue to deliver our products in this way. So we hope that our agentic models can be provided as a product for partners to experience, try and deploy. So it would become part of our revenue since next year. So AI capabilities would be very important in our revenue structure in the near future. So we are going to continue this transformation, so whether it's fee-based, whether it's other forms of revenue. So again, we are making the transformation always. We hope to be a more technology-driven company and output our capabilities. So AI is not just used to reduce our cost and improve our efficiency, it has already been perfected in our business. We hope we can further deliver output capabilities to the whole industry. Right now, we can see in the industry, there is great expectations for our business, for our capabilities in the whole sector.

And now we give the floor to the second question from a Chinese investor from Citi, Vicky.