Noah Holdings (NYSE:NOAH) released second-quarter financial results and hosted an earnings call on Tuesday. Read the complete transcript below.

This transcript is brought to you by Benzinga APIs. For real-time access to our entire catalog, please visit https://www.benzinga.com/apis/ for a consultation.

Access the full call at https://event.choruscall.com/mediaframe/webcast.html?webcastid=PKGdywVt

Summary

NOAH Holdings Limited reported a 34% year-over-year increase in operating income to RMB 216 million, with a margin of 34.8%, driven by the transition to an AI-enabled operating model.

The AI Wealth Management Department in Singapore achieved monthly profitability, contributing to a shift from traditional relationship manager (RM) growth to AI-driven global expansion, reducing overseas RM headcount by 36.2% while USD-denominated AUM grew by 11.7%.

Future priorities include replicating the AI Wealth Management model in more markets, enhancing AI and data capabilities for investment, and building a global ecosystem partner network to support global expansion.

Full Transcript

OPERATOR

Good day, and welcome to the NOAH Holdings Limited second quarter and half year 2026 earnings conference call. All participants will be in a listen-only mode. Should you need assistance, please signal a conference specialist by pressing the star key followed by zero. After today's presentation, there will be an opportunity to ask questions. To ask a question, you may press star then one on your telephone keypad, and to withdraw your question, please press star then two.

You may also submit questions via the webcast. Please note today's event is being recorded. I would now like to turn the conference over to Doreen Chu with Investor Relations. Please go ahead.

Doreen Chu, Investor Relations

Thank you. Good morning and good evening, everyone. Welcome to NOAH's second quarter 2026 earnings conference call. Joining me on the call today: Ms. Wang, Co‑Founder and Chairlady; Mr. Sander Yin, Co‑Founder, Director and CEO; Mr. Grant Peng, CFO; and Mr. Jason Wu, Deputy CFO. Mr. Yin will begin with an overview of our recent business highlights, followed by Mr. Peng, who will discuss our financial and operational results. They will all be available to take your questions in the Q&A section that follows.

Please note that the discussion today will contain forward‑looking statements that are subject to risks and uncertainties that may cause actual results to vary materially from those in our forward‑looking statements. Potential risks and uncertainties include, but are not limited to, those outlined in our public filings with the SEC and the Hong Kong Stock Exchange. NOAH does not undertake any obligation to update any forward‑looking statements except as required under the applicable law.

With that, I would like to pass the call over to Mr. Yin. Please go ahead.

OPERATOR

Hello, this is the operator. Your line is open. Are you perhaps muted?

Sander Yin, Co‑Founder, Director and CEO

Good morning, everyone. Thank you for joining NOAH Holdings' second quarter 2026 earnings call. As we enter the second quarter, NOAH's transformation reached a new stage. On our first quarter earnings call, we said that Q1 represented the beginning of the validation of NOAH's new operating model. After the second quarter, I believe we can take that conclusion one step further. Revenues associated with our legacy model are being phased out in an orderly manner, while our new operating model is not only being validated, but it's also beginning to generate revenue, asset growth, and profit.

Most importantly, our AI wealth management department and new AI‑enabled front‑office operating model we have been developing over the past year has now established its first meaningful proof point. In Singapore this quarter, we observed three developments that have become increasingly clear. First, Singapore has proved the first meaningful validation of our AI wealth management department model. In approximately 10 months, the business achieved monthly profitability in July.

Second, performance‑based income reached RMB 238 million in the first half of the year, demonstrating how the global investment capabilities we have built over many years are increasingly translating into earnings. Third, while our total employee headcount declined 17% year over year, US dollar‑denominated AUM increased by 11.7%. More importantly, our overseas RM headcount declined by 36.2% year over year while overseas assets continued to grow. This suggests that we are beginning to decouple asset growth from RM headcount growth, a relationship that has historically been highly linear in traditional wealth management. This is also the most important message we would like to communicate to the market today. For NOAH, AI is no longer simply about improving efficiency or reducing costs. It is beginning to change how we organize our front office, how we serve clients, and how we expand globally.

Today I will discuss our financial performance, our Mainland China business, our international business, and our key priorities for the second half of the year. In the second quarter, we generated net revenue of 620 million. Operating income was 216 million renminbi, up 34% year over year, with an operating margin of 34.8%. Non‑GAAP net income attributable to NOAH was 238 million renminbi, up 25.9% year over year and 77.8% quarter over quarter. For the first half of the year, net revenues were renminbi 1.25 billion, broadly flat year over year. Operating income reached renminbi 452 million, up 30.3% year over year, with an operating margin of 36.3%, representing an improvement of 8.4 percentage points from the same period last year. Non‑GAAP net income was 372 million renminbi, up 3.9% year over year. This quarter also marked our 63rd consecutive quarter of non‑GAAP profitability since our IPO.

While total revenue remained broadly stable compared with last year, the composition of our revenue and our operating efficiency have changed meaningfully. I'd like to explain this change clearly. On one hand, revenues associated with the legacy model are being phased out in an orderly manner. Net distribution income declined the first 6% year over year in the third term, including a 53.8% decline in insurance‑related products. This contraction reflects deliberate strategic choices we have made.

I will discuss the Mainland China and the international business separately in more detail later. On the other hand, revenues associated with our investment capabilities are increasingly being realized. Net performance‑based income, or what we call carry, reached 238 million renminbi in the first half, up 364% year over year. Distribution income from investment products increased 13.4%. Investment income was 39.79 million renminbi, compared with a loss in the same period last year.

At the same time, operating efficiency continued to improve. Operating costs and expenses declined 11.6% year over year in the first half, including a 12.7% reduction in personnel costs. These improvements reflect both disciplined cost management and organizational streamlining, but more importantly, we believe that they reflect a structural change in how NOAH is beginning to operate. We are now increasingly operating with a leaner, more digitalized model that relies more on platform capabilities and less on simply adding headcount.

This is one of the structural outcomes we are beginning to see after several quarters of sustained investment in AI. Our view of the earnings structure is clear. During the transformation period, carry realization and cost discipline provide support for profitability, and for recurring revenues to return to sustainable growth, new growth engines must come online. The AI wealth management department is one of the new growth engines we are now actively validating.

This will be one of management's most important priorities in the second half of the year. We maintain the outlook we provided on our first quarter earnings call. We expect our full‑year operating margin to remain at a healthy level, about 30%, while quarterly results may naturally fluctuate depending on product mix and the timing of expenses. Carry is not a one‑time outcome driven by luck. It is the result of a long‑term, systematic investment capability that can continuously generate value across investment cycles.

Our underlying model consists of three layers. The first layer is investing as an LP in leading global funds, which gives us exposure to the most advanced industry insights and investment opportunities. The second layer is using our fund‑of‑funds portfolio to broaden that information network and observe the collective investment decisions of leading global investment institutions. The third layer is co‑investment and direct investment, where we convert the information advantage accumulated through the first two layers into more concentrated sources of potential return and carry.

Together, these three layers form the foundation of our continuously evolving product portfolio. Funds established in different vintages are at different stages of their life cycles. As earlier funds progressively enter the harvesting period, they create the foundation for recurrent carry realization rather than dependence on any single investment or exit. At the same time, AI and data capabilities are further strengthening this investment system.

On the asset side, we positioned ourselves relatively early in several core segments of the global AI value chain, and many of these assets remain in the early to middle stages of value realization. We look through nearly 60 underlying funds and use institutionalized data systems and cross‑validation to understand what leading GPs are investing in together, where they are increasing their exposure, and where capital and industry trends are converging.

On the client side, AI is helping us address another equally important question. It is not only about selecting the right product; it is also about identifying the right client and matching the right product with the right client at the right time. Investment judgment, client understanding, and asset allocation are increasingly being connected through data. The deeper the data becomes, the more precise our judgment can be. When clients achieve strong investment outcomes, the investment base and AUM can increase, which in turn creates the potential for future carry.

This is the long‑term flywheel we are working to build. Our Hong Kong platform has historically distributed US$158 million in carry already, demonstrating a track record of actual realization. Of course, alternative investments are inherently cyclical, and carry will fluctuate from year to year. We will not normalize or make linear assumptions around carry, and we will continue to provide updates each quarter. In terms of transaction values, total fundraising reached renminbi 40.5 billion in the first half, up 22.4% year over year.

US dollar‑denominated product fundraising reached US$2.45 billion, up 8.4% year over year, accounting for 41% of the total. As of June 30, Group AUM returned to its sequential growth and reached renminbi 140.9 billion. US dollar‑denominated AUM reached US$6.5 billion, up 11.7% year over year, while US dollar‑denominated AUA reached US$9.78 billion, up 7.5%. Our balance sheet remains strong. As of June 30, we held approximately renminbi 5 billion in cash, cash equivalents, and short‑term investments, with zero interest‑bearing debt.

We also made important progress this quarter in resolving the legacy counseling matter. During the first half, we completed the issuance of shares to clients who had previously entered into the settlement agreement. In the second quarter, we introduced a new settlement proposal for clients who had not yet settled, and the number of investors accepting the proposal continues to increase. Given the adequate proposition we made previously recorded, we recognize a partial reversal during the second quarter.

The uncertainty associated with this last matter has now been meaningfully reduced. Our mainland China business continues in the direction we established in the third quarter, returning to the fundamentals of investment and asset allocation with greater focus on standardized assets that offer sustainable long-term allocation value. In the second quarter, mainland China generated net revenues of Renminbi 384 million. For the first half, net revenues totaled Renminbi 776 million, up approximately 20.7% year over year. NOAH Upright raised Renminbi 8.27 billion in the first half and generated net revenues of Renminbi 414 million, up 59.8% year over year.

For Upright, our operating philosophy is becoming increasingly simple. The most important measure of success is whether our clients make profit. We focus on client profitability, client retention and weighted investment returns. Our current product shelf is primarily focused on defensive strategies such as market-neutral quantitative strategies and CTAs. In the current market environment, we are not pushing scale for the sake of scale. Instead, we are placing greater emphasis on asset quality, client investment outcomes and long-term trust.

Gopher generated net revenues of Renminbi 341 million in the first half, broadly flat year over year. Our mainland China insurance business generated net revenues of Renminbi 3.382 million in the first half. We have proactively reduced and gradually exited the traditional high-commission protection product model, shifting instead toward comprehensive services such as family succession and inheritance planning. This adjustment began before the relevant regulatory requirements and reflected our own long-term assessment of client value.

The future positioning of our mainland China business is becoming increasingly clear. We will focus on secondary market investments, serve clients through professional investment and asset-allocation capabilities, and use AI to improve client engagement and service efficiency. Let me first review the numbers for our international business. Net revenues from the international segment were Renminbi 469 million in the first half, accounting for 37.7% of group net revenues and declining 21.9% year over year.

This decline needs to be understood in context. Nearly 90% of the decline came from the deliberate contraction of our insurance business and the exit from legacy referral channels. Excluding those two factors, revenue from US dollar-denominated investment products was broadly flat year over year while our client base and assets under control continued to grow. US dollar-denominated AUM increased 11.7% year over year. As of 9th June, registered overseas clients reached 21,059, up 11% year over year.

Overseas Diamond and Black Card clients reached 1,791, up 8.9%. Active overseas clients reached 3,494 in the second quarter, up 8.5% sequentially. Our assessment of the current stage of our international business therefore remains unchanged. We are not losing clients or assets. The legacy engines are simply being phased out faster than the new engines are coming online. This is consistent with the revenue mix transition we have discussed with the market over the past several quarters.

However, one development in the second quarter is particularly important. As of the end of the second quarter, overseas RM headcount was down 36.2% year over year while US dollar-denominated AUM increased by 11.7%. Under the traditional wealth management operating model, these two outcomes would rarely occur at the same time. Historically, the industry's growth formula has been straightforward: more RMs lead to more clients, more clients lead to more AUM and more revenue.

However, the past year we have been working to change that equation. On our first quarter earnings call, we introduced the three front office engines that we believe will define NOAH's future operating model: first, AI-empowered RMs; second, the AI Wealth Management Department; and third, AI ecosystem expansion. After the second quarter, we are beginning to see the second front office engine, the AI Wealth Management Department, move from an organizational concept into a real operating model.

So what is the AI Wealth Management Department? It is not simply about giving traditional RMs a few additional AI tools. At its core, it represents a new way of organizing the front office of a wealth management business. Historically, one client was typically associated with one RM. The client experience, product understanding, frequency of engagement and often a significant amount of client information were highly dependent on that individual RM.

This makes the traditional wealth management industry inherently difficult to scale. The AI Wealth Management Department seeks to redesign this process. AI and a centralized wealth management team handle a significant portion of high-frequency, standardized and digitalized client engagement and daily services. Licensed professionals are responsible for the critical stages requiring judgment, compliance and professional accountability. Ecosystem partners then expand our client reach.

Under this model, a client no longer belongs simply to an individual RM. The client is served by the combined capabilities of the entire NOAH platform. This is where we believe AI can fundamentally change wealth management. It is not simply about helping one RM prepare materials faster. It is about giving wealth management the opportunity to move from an individual productivity model toward an institutionalized productivity model. Singapore is the first market where we have fully tested this model.

From its launch in the fourth quarter of last year, Singapore AUM grew from less than US$100 million to more than US$400 million by the second quarter and the business achieved monthly profitability in July. What's even more important is how that growth was achieved. Today, 92% of clients are covered by our AI-enabled service model for day-to-day engagement while licensed professionals remain responsible for regulated activities, professional judgment and compliance delivery.

This allows our professionals to spend significantly less time on repetitive administrative work, information organization and standardized service and more time on the things that truly require human capability: understanding clients, building trust, identifying needs and communicating around important decisions. At the same time, external ecosystem partners contributed 42% of our new AUM. Singapore raised US$150 million in the first half, up 126% year over year, and this growth was achieved without relying on a large expansion in RM headcount.

For us, the significance of these numbers goes well beyond the growth of the Singapore business itself. For the first time, they demonstrate that NOAH may be able to gradually shift wealth management from a growth model highly dependent on recruiting more RMs toward a model driven by an AI platform, licensed professionals plus ecosystem partners. If this model continues to be validated, it has the potential to change our unit economics, management span and ability to replicate our business globally.

This is why we view the AI Wealth Management Department as a potentially important new growth curve for NOAH. Looking at our international business by segment, NOAH Wealth generated net revenues of Renminbi 193 million in the first half, including Renminbi 88.65 million in the second quarter. Gopher Asset Management generated net revenues of Renminbi 198 million in the first half. US dollar-denominated private equity fundraising reached US$410 million, up 13.4% year over year.

US dollar-denominated structured products and hedge fundraising reached US$590 million, up 33.2%. Gopher generated net revenues of Renminbi 78.524 million in the first half while its independent broker network expanded to 238 professionals. Our international strategy has not changed. We remain focused on serving Chinese high net worth families around the world. Their assets, families, residency application and next-generation planning are becoming increasingly global.

Historically, serving these clients simultaneously across Hong Kong, Singapore, Japan, Canada, Australia, the UK, Europe and the United States would have required a large local RM and operating organization in every market. That made economically efficient global coverage extremely difficult. Yet the AI Wealth Management Department changes the core structure of group markets that previously could not be economically covered because of insufficient client density and may increasingly become addressable through a combination of the AI Wealth Management Department, local licensed professionals and ecosystem partners.

This is why we often say AI is making it possible for the first time for NOAH to serve Chinese high net worth families around the world and at scale. Coming to the last part of this presentation would be the second half priorities: from one proof point to a replicable system. In the first quarter we introduced three front office engines that are beginning to shape a new operating model: AI-empowered RMs, the AI Wealth Management Department and AI ecosystem expansion.

If the first quarter was about introducing and initially validating this model, then the most important progress in the second quarter was that the AI Wealth Management Department produced its first operating proof point. Our priority for the second half is to move from one proof point toward the system that can be replicated across markets. We will focus on four major areas. First, replicate the Singapore AI Wealth Management Department model across more markets.

The Singapore model has completed its first stage of validation from launch to AUM growth to monthly profitability. We are now working to replicate this model in Hong Kong and Japan, with plans to gradually expand into Canada, Australia, the UK and Europe. We do not intend to replicate the traditional wealth management model of building a heavy physical footprint in every market. Our basic architecture will be centralized AI wealth management capabilities plus local licensed professionals plus local ecosystem partners.

AI empowers the service capability of the platform's licensed professionals provide professional judgment and compliance delivery. Ecosystem partners allow us to reach more clients with a lower fixed cost base. If this model can be replicated successfully, we believe it could meaningfully improve the future unit economics of our international business. Yeah, Sanjay. Second, use AI to strengthen investment capabilities and the sustainable realization of carry. On the asset side, we look through Lilly's 60 underlying funds and continuously analyze the common investment and follow-on investment decisions of leading robo GPs. On the client side, we use AI and data capabilities to improve the quality of product-to-client matching. This is not simply product recommendation. Our objective is to understand the client's real needs, risk tolerance, existing asset allocation and historical investment behavior, and then we can identify the assets that are most appropriate for our clients.

When clients achieve strong investment outcomes, satisfaction and reinvestment rates would be improved. Reinvestment drives AUM growth, and when high-quality assets ultimately realize the value they generate, the carry flywheel has already begun to turn. In the second half, we intend to further deepen the data foundation and improve the efficiency of this system. Third, we will continue to strengthen our international operating infrastructure. During the quarter, the Group continued to strengthen the operating infrastructure supporting our international business development.

The Group established a partnership with U.S.-licensed banking institution Column Bank, or Column National Association, to enhance account opening, multi-currency settlement and payment processing capabilities for our international clients. This partnership is part of the international middle- and back-office infrastructure we have been building since 2024 and is designed to improve client service efficiency and operating scalability across our licensed entities in Hong Kong, Singapore, the United States and Japan.

The relevant services are provided only to non-mainland Chinese resident clients who meet applicable law and regulatory requirements. Why does this infrastructure matter? Because if the AI Wealth Management department is ultimately going to scale across markets, AI in the front office is only one part of the equation. Account opening, transactions, payments, compliance and asset execution must also operate through an integrated and efficient infrastructure.

Front-office AI transformation and back-end personalization must happen together for global expansion to generate true economies of scale. Last but not least, build a global ecosystem partner network powered by AI. Our enterprise-focused platform has now accumulated more than US$33 million in AUM. Our business partner program has been launched with the first group of partner institutions registered in Hong Kong and Singapore. All regulated activities are conducted by licensed professionals.

We are beginning to see our third front-office engine, AI ecosystem expansion, connect with the AI Wealth Management department. In the future, an ecosystem partner should not need to rebuild an entire infrastructure covering product research, technology, account and product compliance. Once NOAH has built this infrastructure, our partners can leverage these capabilities within the appropriate regulatory framework to better serve their own clients.

We build the financial infrastructure once and the global network can reuse it rapidly. This is the model we are looking forward to, and we are working toward expanding our service reach through platform capabilities rather than expanding our organizational footprints through fixed cost. We remain in the middle of transformation. The near-term pressure points are transparent and we will not avoid them. But compared with several quarters ago, our confidence in the future operating model is increasingly supported by real operating data.

In the first quarter, the new operating model began to show initial signs of validation. By the second quarter we are already seeing profitability. AUM growth improves, the positivity, a working Singapore proof point and a model that's beginning to demonstrate weak capability. Most importantly, the implementation of the AI Wealth Management department has given us a much clearer view of something we believe is fundamental. AI is not simply helping NOAH do what we already do faster and at a lower cost.

It may allow us to do things that were not economically possible under the traditional wealth management model. Under the traditional model, global expansion meant continuously hiring more RMs, opening more offices and adding more fixed costs. Yet under the new model, we aim to serve Chinese high-net-worth families across different markets through a combination of an AI-powered platform plus licensed professionals and global financial infrastructures and an ecosystem partner network.

If this model continues to be validated, NOAH's future growth will become progressively less dependent on how many people we add and increasingly depend on, first, how many clients our platform can serve; second, how deeply our AI can understand those clients; third, how much value our investment platform can create; and how many professional partners our ecosystem can connect. This, in our view, is the fundamental transformation of a wealth management company in the AI era.

NOAH is evolving from a traditional wealth management institution into an AI-driven global wealth management platform serving Chinese high-net-worth families around the world. This transformation will not happen overnight, but beginning this quarter it is no longer simply efficient, it is becoming a real operating model with clients, AUM, revenue, profitability, and increasingly evidence of the repatriation across markets. Wealth management is a lifetime commitment.

Thank you for your time and I will now turn the call over to our CFO, Grant, who will walk you through our financial results in greater detail.

UNKNOWN, CFO

Thank you, Doreen, and thank you, Sander, and good day to everyone joining us. As Sander laid out in great detail just now, we're at the turning point of upgrading our business. Our second quarter was about the quality of our profitability rather than pure growth in revenue. Our operating profit was RMB 216 million in the second quarter, up 34% year over year, with operating margin at 34.8%. Non-GAAP net income was RMB 238 million, up 25.9% year over year and 77.8% sequentially.

For the first half of 2026, operating profit was RMB 452 million, up 30.3%, with a record half-year margin of 36.3%. This was also our third consecutive quarter of year-over-year operating profit growth. The cost optimization behind the margin expansion is structural rather than cyclical. Operating costs fell 13.7% year over year, and total headcount streamlined about 17%. For the first half, costs were down 11.6% flat on flat growth in revenue. Delivering the same revenue from a materially smaller cost base is what produced margin and efficiency.

Importantly, this is increasingly about more than simply reducing overheads. It reflects a leaner operating model enabled by AI and process redesign. As our CEO discussed, US dollar AUM grew 11.7% year over year even as overseas IAM headcount declined 36.2%. This is early financial evidence that we're beginning to support a larger asset base with a more efficient organization. Carry, or performance-based income, was RMB 138 million in the quarter and RMB 238 million in the first half.

Carry is not a one-off event in our model. We have recognized performance income in every year we have reported and grew 78% in 2025. In Hong Kong alone we have distributed US$158 million of carry to date. Supporting the story is a US dollar asset base that keeps growing: US dollar AUM up 11.7% and US dollar AUA up 7.5% year over year. Of course it's challenging to forecast carry income because realization depends on market conditions, exit opportunities, and the timing of underlying portfolio realizations.

So with that, let me take you through the details. Net revenue: second quarter net revenue was RMB 620 million, down 1.5% year over year and 0.9% sequentially. First half net revenue was RMB 1.25 billion, in line with last year. One-time commissions were RMB 87 million, down 44.1% year over year. The decline is mainly attributed to insurance income, where commissions fell 58.2% year over year. This also reflects competition intensified in this market and our own position to sort of walk away from business that does not meet our margin and suitability standards.

Recurring management fees were RMB 360 million, down 10.8% year over year and 5% sequentially as legacy RMB private equity assets run off. A moment more on carry, because it's often read as a windfall—it's not. The reason is structure. Let me explain why it recurs. First, our investment work is institutionalized across 67 private equity funds built over more than a decade. We've looked through to holdings across more than 50 sub-funds. Our multiple funds have invested across various vintages since early financing rounds of notable holdings such as ByteDance and Anthropic.

When realization of carry depends on the timing and form of exits, this diversified portfolio across vintages provides a broad underlying base from which future performance-based income may be realized over time. To caveat on that, carry is realized-driven and will not be smooth or linear from period to period and should not be annualized off any single quarter. Accrued amounts move with valuations in both directions. We do not accrue carry or forecast carry.

We record it on cash basis. We do not budget on peak carry, and one-off gains do not enter our fixed cost base. Turning to the drivers, transaction value of distributions were RMB 17.2 billion in the quarter, at 1.1% year over year and down 26.4% sequentially. First half volume was RMB 40.5 billion, up 22.4%. US dollar products grew 8.3% year over year and were the engine in the quarter. Two things drove the sequential decline. The first quarter was an exceptionally strong RMB fundraising quarter, and with plans in June to position ahead of the A-share adjustment that did come in in July.

We track this line closely because volume drives commissions today and builds the asset base that pays management fees later. Our US dollar asset base grew. US dollar AUM reached US$6.5 billion, up 11.7% year over year. AUA reached US$9.8 billion, up 7.5%. International registered clients were up 11% year over year. That is the clearest evidence our mix is moving toward investment-related business. More importantly, this asset growth was achieved when overseas IAM headcount declined 36.2% year over year.

We view this as early decoupling between asset growth and IAM headcount growth as an important indicator of the operating leverage we're seeking to build through AI-enabled servicing and our evolving front office model. As Sander just laid out, as highlighted earlier, total operating cost and expense continues to drop. Looking at the breakdown, total compensation and benefits for the second quarter fell 13.1% year over year to RMB 260 million; for the first half, compensation was down 12.7% to RMB 527 million.

This continues efficiency work we have discussed for several quarters, with AI process redesign allowing for a smaller organization to carry the same service coverage. As Andrew discussed, the AI Wealth Management Department represents a further evolution of this model, using AI and centralized service capabilities for high-frequency and standardized client engagement. While licensed professionals remain responsible for regulated activities, professional judgment, and compliance delivery, the early financial evidence is encouraging.

Overseas IAM dropped year over year about 36%, while US dollar AUM increased 11.7%. In Singapore, where we're testing this model end to end, AUM has grown from less than US$100 million to more than US$400 million, and the business achieved monthly profitability in July. Obviously it's still early, but if this model proves replicable, future asset growth can become progressively less dependent on proportional increases in IAM headcount and fixed cost.

On the non-personnel side, second quarter selling expenses dropped 10% year over year, while for the first half they were down 18.6% to RMB 92 million. Moving on, operating profit and non-GAAP net income: operating profit for the second quarter was RMB 216 million, up 34% year over year, delivering an operating margin of 34.8%, up 9.2 percentage points from 25.6% a year ago. On a sequential basis, operating profit is down 8.7%, tracking lower quarterly net revenues.

For the first half, operating profit reached RMB 452 million, up 30.3% year over year, with a record margin of 36.3%. The key takeaway is conversion efficiency. Even on flat revenue, our cost discipline allowed a significantly higher portion of revenue to flow straight to operating income. On the bottom line, second quarter non-GAAP net income reached RMB 238 million, up about 26% year over year and 77.8% sequentially, with non-GAAP net margin expanding to 38.4%.

For the first half, non-GAAP net income reached RMB 372 million, up about 4% year over year. On a GAAP basis, second quarter net income attributable to shareholders was RMB 232 million, up 30% year over year and 86.2% sequentially. Record items remain modest, with share-based compensation falling 44% year over year to RMB 7.2 million in the quarter. Two non-cash items below the operating line also contributed to that result, which might naturally raise questions, so I want to clarify here.

Investment income was a positive RMB 42 million in the quarter against a loss of RMB 14 million a year ago, and RMB 40 million in the first half against a loss of RMB 8 million. That is our own portfolio at work. The income from equity affiliates was a gain of RMB 55 million in the quarter, after a charge in the first quarter, leaving a loss of RMB 10 million for the half against a gain of RMB 36 million in the first half of last year. While these are non-cash items, there is a broader point.

Product access and distribution are commoditizing; what's durable in this business is the ability to make clients money and to be paid only when they do. That is how we're built. We co-invest alongside our funds and clients, carry pays only after they make a profit, and fees persist only if they stay. Our investment portfolio and carry are two views of the same capability. On the legacy Kensing matter, we made real progress this quarter. We accelerated the related share insurance, which removes a significant piece of uncertainty, and we launched a new settlement plan for the remaining colonies.

Settlements are concluding on average below our current provisional level. Contingent liabilities were RMB 455 million at June 30, down from RMB 505 million on March 31. Moving on to balance sheet and shareholders' return, we ended the quarter with RMB 5.0 billion in cash and short-term investments, no interest-bearing debt, and a current ratio of 4.3 times. Shareholders' equity was RMB 9.8 billion. We're trading at roughly half of book value, a valuation that fails to reflect our true intrinsic value and underlying earnings capabilities, as demonstrated by our second quarter annualized non-GAAP return on equity of 9.7% and first half annualized ROE of 7.6%. To deliver sustainable returns, we completed our 2025 dividend distribution in July 2026, making our third consecutive year maintaining a 100% net income payout ratio and bringing accumulated dividends from 2022 to 2025 to roughly RMB 2.4 billion. In parallel, under our share repurchase program launched in 2024, cumulative execution reached over 3.2 million ADS shares for more than US$34 million. These capital deployment actions highlight our commitment to enhancing shareholder value and our confidence in NOAH's long-term earnings potential.

To close, I want to leave you with three key takeaways. First, what changed? We demonstrated the upgraded earning power of our platform, delivering a 34% year-over-year increase in operating profit, mid-30s operating margins, and third straight quarter of profit growth, driven by disciplined cost management and an increasingly efficient operating model. We're also beginning to see early financial evidence of the new operating model. As Sander just pointed out, USD AUM continued to grow, up 11.7% year over year, while overseas IAM headcount dropped.

Singapore has provided the first meaningful proof point that the AI-enabled wealth management model can support asset growth without proportional increases in headcount. Secondly, what will need work? Core commissions and management fees both fell as year-to-date return growth and regulatory headwinds tightened noticeably. Third, what underpins our foundation? Our international business continues to expand in both asset and client count, our investment franchise maintains an unbroken annual track record of generating carry, and our balance sheet remains debt-free with strong liquidity fully backing our high payout commitments.

The financial objective behind our transformation is straightforward: to build a business where AUM, volumes, and revenue continue to grow faster than fixed costs and headcounts. So thank you for your continued trust, time, and partnership. We're now happy to take your questions.

Doreen Chu, Investor Relations

Hi, Doreen here. Since we have the webcast, we received a few questions. Let me read through the first one and let management do the answers. First question is from Kevin from Citibank, and he's asking on contingency expenses and the related liabilities. We noticed there was a reversal of contingent litigation expenses since second quarter. Can management give more color on the recent development on the litigation and can we expect more reversal in the next few quarters?

Thank you.

Jason Wu (Deputy CFO)

Okay, I will take that question. So as the CEO just mentioned, in the second quarter we introduced a new settlement plan relating to Kempsing. As of today, more than 80% of our affected clients have accepted the settlement plan, so our legacy risk exposure has declined substantially. As we have made provisions in prior years for the litigation risk associated with all unsettled clients, we are now making adjustments in the provision balance on a quarterly basis in line with actual settlement progress.

We are continuing to engage with those remaining unsettled clients, but unfortunately we cannot make any prediction on the future provision reversal because we are still tracking the settlement progress, and we'll keep focusing on the core operating profit.

Doreen Chu, Investor Relations

Thank you, Jason. Maybe we should open the line to see if there's anyone using the phone with questions.

OPERATOR

As a question, please press star then one. Our first question today comes from Peter Zhang at JPMorgan. Please go ahead.

Peter Zhang, Analyst at JPMorgan

Thanks for giving me the opportunity to ask questions. This is Peter Zhang from JPMorgan. Congratulations on the very strong second quarter result, and I have two questions. First, it was mentioned that we are currently in a transition period and the AI strategy has gained very strong momentum. I wish to understand how long this transition period could take and what metrics will you recommend investors to monitor to track the progress of this transition, and do we have any longer term target for this AI strategy, say any target in the next three to five years.

My second question is also on AI wealth management. I wish to understand, for the KPIs you give to your relationship managers, how the new KPI under the new AI wealth management model looks and how these compare to the KPIs in the old models, and how NOAH is growing your client base under the new AI wealth management model. And can you also give any examples on how AI has improved our—say, how AI has improved the productivity for the RMs, for example the number of clients an RM can provide service to and the amount of AUM an RM can serve.

Thank you.

Doreen Chu, Investor Relations

Let me do the translation. Thanks, Peter, for the questions. What Charlie was trying to explain—what we've been elaborating the whole morning—is the new system NOAH is trying to build up, which no longer just relies on RMs and how many clients one RM can serve. What we've been repetitively talking about is how we've been using three platforms, which are the AI and cloud RMs. We still value humans, and AI can be the tool to empower performance, but at the same time it's more about how we have built up the AI wealth management department, and also we have this platform called AI-plus ecosystem expansions.

To give you some examples: in Singapore, with only six people, we've been able to cover 500 clients, which in the past was basically impossible. With the new system the company is building, we found that it enhances more than just efficiency and more than just the number of clients per RM. Also, our experience with the AI ecosystem platform is that we've been able to cooperate with experts from different industries which may have clients that have wealth management needs but don't have the license.

By cooperating with us, we provide fees to them, and at the same time they can better serve the client as well, and they can refer the clients to us for wealth management needs. So it's no longer just a very traditional way to look at the wealth management business under this model. A 3–5 year target is probably not the best, because under the AI era changes could be very fast. Taking Singapore as an example, we only started the business in September last year with AUM of around 86 million, but now it's already over 400 million—with those AUM being really high-margin, high-quality AUMs.

What we've been trying to emphasize here is that under the AI-enabled company we are no longer the traditional model. It's not about hiring more RMs and hopefully the RMs can get clients; ultimately it's about AUM. We've been trying to expand our global footprint, and in Sander's speech we've already mentioned we are now replicating the AI model from Singapore to Hong Kong, and we will further expand the offices into different cities as well. At the same time, the cost will be under control, and it's no longer the traditional wealth management company model.

We believe that it's going to have a drastic change in a relatively much shorter time compared to the old model. We tried in Hong Kong when research came here, but it doesn't seem to be a very successful experience, and that's why in Singapore we found the path to success, and we are going to repeat that in many different cities as well.

OPERATOR

There are no further audio questions at this time. I hand the call back over to the company.

Doreen Chu, Investor Relations

Thank you. We have another question here, which is the second question. It's from Ms. Wang, and she's asking: would you be able to tell us more on the dividend expectation for the current fiscal year and periods ahead? Following the presentation on the AI development strategy, may we ask for your view on the anticipated effects on both revenue and profitability in the medium to long term? I think we have basically answered the second part of the question about how AI development is going to help the revenue and profitability in the future, so I would like management to address our dividend policies.

UNKNOWN, CFO

So as we have just mentioned, for the past three, four years, since 2022, we have cumulatively given out about 2.2 billion RMB. Given our strong position in the balance sheet and also highly managed liquidity, we expect to continue to distribute a significant portion of the income to our shareholders to maintain a high return. But we have not decided on exactly the portion or ratio of the future dividend. We believe we will remain consistent on the asset allocation and also shareholder return policy to our shareholders.

Secondly, we believe that, as Charity Lee and the CEO have mentioned, when AI continues to push our profitability and earning power and upgrade the business model, we will be able to sustain at least a comparable level of shareholder returns in the future.

Doreen Chu, Investor Relations

Yes, thank you, Grant. Can I check if there's any questions from the phone?

OPERATOR

There are no further phone questions at this time.

Doreen Chu, Investor Relations

If that's the case, I think we will wrap up the presentations today, and thank you everyone for joining us this morning. I understand that there may still be further questions. Please contact the IR team at your convenience, and I look forward to talking to you in more detail in the near future. Thank you very much.

OPERATOR

Thank you. This does conclude our conference for today. We thank you all for attending today's presentation. You may now disconnect your lines.

Disclaimer: This transcript is provided for informational purposes only. While we strive for accuracy, there may be errors or omissions in this automated transcription. For official company statements and financial information, please refer to the company's SEC filings and official press releases. Corporate participants' and analysts' statements reflect their views as of the date of this call and are subject to change without notice.