In August 2026, Ping An Insurance (Group) Company of China, Ltd. delivered a double-digit first-half performance, with operating revenue reaching 575.138 billion yuan, up 15.0% year-on-year, while net profit attributable to shareholders of the parent company hit 92.585 billion yuan, a substantial 36.1% increase.
Two figures from the interim report deserve particular attention. The first is 6.61 trillion. As of the first half of 2026, Ping An's insurance funds investment portfolio reached 6.61 trillion yuan, up 1.9% from the start of the year. Looking at a longer timeframe, the average net investment yield over the past decade stands at 4.8%, with an average comprehensive investment yield of 4.9%. Managing this massive capital pool represents Ping An's foundational discipline and the bedrock of its ability to navigate insurance capital through economic cycles.
The second figure is 210 billion. In the interim report, Ping An disclosed its Token consumption data for the first time. Daily average usage surged from 30 billion at the end of 2025 to over 120 billion by June 2026, quadrupling within six months. By August, it had surpassed 210 billion. For Ping An, tokens initially represented computing power costs. But as daily consumption jumped from 30 billion to 210 billion, the definition of cost shifted, now exchanged for data asset accumulation and iterative improvements in model capabilities. With each spin of the flywheel, the commercial value generated per unit of Token increases, representing Ping An's new growth momentum.
In a recent dialogue with media, Ping An General Manager and Co-CEO Xie Yonglin framed liabilities and assets as the "old coursework," while positioning AI as the critical "multiplier effect" driving new growth. He revealed that Ping An's billion-yuan investment in technology has already formed a flywheel effect of "factor supply - business reshaping - closed-loop feedback." Although senior management acknowledges that Ping An's value has yet to be fully reflected in the market, and its stock price awaits validation, the narrative of Ping An's new growth engines has already begun.
Xie Yonglin joined Ping An in 1994 and has risen from the grassroots to General Manager over 32 years, spanning the insurance, banking, and investment sectors. In 2005, he served as Deputy Director of the Group's Development and Reform Center. From the end of 2013, he took full responsibility for Ping An Securities' operations, and in 2016, he returned to Ping An Bank as Chairman. On December 24, 2019, as Group Co-CEO, he assumed the role of General Manager of Ping An.
A whole life insurance policy often represents the most extensive financial commitment spanning an individual's lifetime, constituting one of the most representative forms of "long-term capital" in China's financial system. "Insurance funds have a unique nature. They represent people's retirement savings, even their 'lifeblood money'. Therefore, prudent investing with sustainable returns is paramount," Xie Yonglin emphasized. Insurance capital investment must adopt a long-term perspective, abandoning short-term speculative trading mentalities, and instead smoothing volatility through diversified asset allocation across stocks, bonds, and alternative investments.
With over 30 years of financial industry experience and having navigated multiple insurance cycles, Xie Yonglin maintains a clear-eyed assessment of industry transformation. As the market debates participating insurance, hard technology, and AI, this veteran's strategy of "defense and offense" is precise: protection-type products should occupy the high ground, hard technology investments must respect the three red lines of solvency, liquidity, and interest spread loss, and AI serves as a critical tool to empower core operations. Discussing life insurance strategy, he acknowledged that in the current low-interest-rate environment, participating and universal insurance products hold an elevated share. However, protection-type products form the essential foundation for long-term value and risk protection functions in life insurance, and this fundamental purpose cannot be abandoned. He further stated that Ping An will not simply pursue the scale or proportion of any single product category but will instead drive deep integration between protection and wealth management scenarios.
Protection must remain the core mission
In 2026, China's financial market is undergoing a profound logic reshaping. With interest rate centers continuing to decline, "risk-free returns" have become scarce, positioning participating insurance products, with their "protection plus floating surplus" advantages, as key players in household asset allocation. Within the insurance industry, participating products now represent over one-third of the market, reshaping the competitive landscape. According to Shenwan Hongyuan Securities data, among the top five main products disclosed by 54 life insurance companies in Q1 2026, participating insurance premiums reached 213.8 billion yuan, accounting for 48%, surpassing traditional products to rank first.
However, an excessively high proportion of participating insurance cuts both ways, potentially lowering the value margin of life insurance business and creating over-reliance on interest spreads for underwriting profitability. In response, Xie Yonglin stated that developing participating insurance products in a low-rate environment represents a key direction for product transformation, effectively meeting customers' long-term savings and wealth management needs. Yet the company will not simply pursue the scale or share of any single product type. Protection products remain the crucial foundation for life insurance's long-term value and risk protection capabilities, and the company plans to promote integration between protection and wealth management scenarios. Having experienced multiple insurance cycles, Xie Yonglin has direct insight into the impact of interest rate fluctuations. He acknowledged that with the current high proportion of participating and universal products under low interest rates, the industry should gradually increase the share of critical illness and other protection-type products, allowing protection business to reclaim its high ground.
He disclosed that in the second half of the year, the company will increase promotion of longer-payment-period products and optimize premium payment structures to enhance value margins. In May, a "pension annuity plus nursing insurance" combination aligned with retirement scenarios was launched. In August, a new critical illness product hit the market, using a "participating whole life with advance payment critical illness benefit" approach to achieve "dividend-enhanced coverage," thereby boosting critical illness protection. Participating critical illness products await regulatory guidelines before rapid development and launch.
From core life insurance metrics, Ping An's life and health insurance business operating profit reached 55.872 billion yuan in H1 2026, up 2.3% year-on-year, while new business value amounted to 24.847 billion yuan, up 11.2%. In the property and casualty segment, new energy vehicle insurance has become a focal point. Xie Yonglin stated directly at the 2026 interim results conference, "One in every four new energy vehicles in China is underwritten by Ping An, and the comprehensive cost ratio for NEV insurance business is controlled at a certain level." Interim data shows Ping An underwrote 6.78 million new energy vehicles in H1 2026, up 27.8% year-on-year on a comparable basis, with NEV insurance original premium income reaching 26.415 billion yuan, up 21.5% and outpacing market growth.
In this highly competitive NEV insurance sector, traditional leading insurers face competition from automaker-affiliated insurance companies that are also aggressively expanding. When asked about this competitive dynamic, Xie Yonglin offered a measured analysis of "two endowments, two approaches." Automaker insurance companies hold advantages in "scene proximity, data proximity, and customer proximity," accessing extensive data on vehicle configuration, driving behavior, maintenance, and software updates while directly reaching customers at purchase, usage, and repair touchpoints, thereby reducing certain distribution costs inherent in traditional insurance models. Insurers, by contrast, leverage advantages in "larger risk pools, longer data history, deeper experience, and broader networks." Major insurers have accumulated vast historical data across different vehicle models, regions, drivers, and accident types, combined with long-standing underwriting data, actuarial capabilities, claims networks, and risk management expertise, providing advantages in complex risk pricing and large-scale operations.
Investment cannot adopt a gambling mentality
Currently, trillion-yuan insurance capital is actively deploying in capital markets. Financial Regulatory Administration disclosures show that as of the end of Q1 this year, insurance funds in use reached approximately 39.44 trillion yuan, surpassing public funds' 37.53 trillion yuan for the second time since the end of 2025. Among this, equity holdings reached 3.84 trillion yuan, approaching the 3.99 trillion yuan scale of public actively-managed equity funds. Ping An manages 6.61 trillion yuan in insurance funds, making its allocation logic a bellwether for the industry. Xie Yonglin advocates constructing a "barbell-shaped" balanced structure to achieve organic balance between stability and flexibility.
One end of the barbell consists of "ballast stones": low-valuation, high-dividend assets as the base, focusing on high-quality blue chips in energy, finance, and infrastructure. Through OCI or equity method accounting, these high-dividend assets consistently contribute stable income while shielding the income statement from short-term volatility. The other end comprises "growth engines": targeting emerging industries such as AI, advanced manufacturing, digital economy, and green low-carbon sectors, positioning in new productive force growth leaders to capture long-term excess returns and capital appreciation opportunities during economic transformation.
"These two asset classes are actually highly complementary," Xie Yonglin noted. High-dividend assets provide stability and cash flow, helping reduce portfolio volatility and enhance defensive capabilities, while growth assets offer long-term appreciation potential, helping the portfolio capture long-term returns from economic transformation. Regarding the hard technology direction within the "growth engine," Xie Yonglin remains cautious: "Under the trend of declining interest rates, hard technology represents an inevitable choice for insurance capital asset-liability management. However, insurance capital investment cannot engage in gambling behavior. Deployment must strictly adhere to solvency and regulatory ratio red lines, maintain diversified holdings and multi-industry allocation, and strictly control concentration in any single track or project." Citing companies like Unitree Robotics as examples, he emphasized that the focus extends beyond the popular hard technology track to these enterprises' technological competitiveness, industrialization progress, and long-term market potential. "Regarding entry timing, we will not simply chase market hotspots. We will comprehensively consider the enterprise's development stage, commercialization progress, valuation levels, and the overall risk-return characteristics of the insurance portfolio, planning investments with proper pacing and discipline."
Investing in technological innovation presents challenges including delayed returns, high failure rates, and high volatility, raising questions about the boundaries of insurance capital "patience." Xie Yonglin responded that the patience boundary consists of a risk framework defined by "solvency, liquidity, and interest spread loss"—constrained by solvency, safeguarded by liquidity management, and targeted at avoiding interest spread losses. In other words, within these boundaries, there is greater patience for short-term fluctuations. However, once fundamentals, asset-liability matching, or risk-adjusted returns undergo substantive changes, investment discipline must prevail, with financial risk prevention remaining the top priority.
How AI creates the "multiplier effect"
Having secured the liability side and stabilized assets, where does Ping An's next offensive focus lie? Xie Yonglin's answer: AI. Speaking about AI, his tone carries added conviction: "When the tide of the times surges forward, for Ping An, AI is not an option to choose or avoid, but a necessity. For an enterprise to maintain sustained competitiveness, increasing investment in AI is imperative at this stage."
Is AI for Ping An merely "icing on the cake" or a decisive "game-changer"? Xie Yonglin's positioning is clear and pragmatic: AI's core mission is to empower the main business. As the financial industry enters an era of homogenized, zero-sum competition, Ping An aims to use AI to build a "service differentiation" moat. The pace of implementation demonstrates the weight of this commitment. In the 2026 interim report, Ping An disclosed Token consumption data for the first time: daily average usage jumped from 30 billion at end-2025 to over 120 billion by June 2026, quadrupling in six months. By August, the figure had exceeded 210 billion.
With Token consumption soaring, how does Ping An balance computing power investment against business returns? Xie Yonglin's answer is "value orientation." Project initiation undergoes quantitative evaluation across dimensions including revenue growth, cost reduction, loss mitigation, and service upgrades, prioritizing projects with optimal input-output ratios to ensure every investment has a clear return logic. He further revealed that in computing power infrastructure construction, Ping An has focused on efficiency improvements and upgrade initiatives: while daily Token consumption quadrupled, high-end computing power increased only 50%, with unit Token intelligent computing costs declining 75%. The billion-yuan technology investment has begun demonstrating flywheel effects. Currently, Ping An's cumulative technology investment exceeds 10 billion yuan, having formed a flywheel loop of "factor supply - business reshaping - closed-loop feedback." Building on this foundation, Ping An is exploring along the path of "frontier exploration - scenario breakthrough - service innovation."
Can AI transform from strategic investment into genuine internal efficiency gains? Xie Yonglin believes artificial intelligence has not yet crossed the "singularity." At the current stage, Ping An enhances enterprise efficiency through three major AI initiatives. First, AI coding: implementing "AI-native working methods," Ping An's self-developed coding tool "Ping An AiCode" has achieved 77% AI-generated new code, with the OPC model proving "one person as a team" automated delivery. Second, AI operations: conducting end-to-end process redesign across the group and five business units in seven benchmark positions, advancing from "tool digitalization" to "organizational intelligence," making AI genuine enterprise productivity. Third, AI office empowerment: serving over 260,000 employees across the group, covering 80% of routine needs, transitioning from "people working" to "AI assisting people in collaborative work," embedding intelligence into the organization's capillaries. Results are supported by data: for business operations positions, the seven benchmark roles undergoing AI process redesign have achieved 21% efficiency improvement, with an annual target of 30%.
Looking ahead, Xie Yonglin stated that Ping An will connect diverse scenarios across protection, wealth, retirement, and health, transforming customers from purchasers of single financial products into users of comprehensive services, thereby enhancing customer loyalty, reducing acquisition costs, and increasing per-customer value. "Particularly, we must leverage AI well, positioning it as the 'multiplier effect' for new growth momentum."