KE Holdings Inc. (BEKE) Earnings Call Transcript & Summary

Hello, ladies and gentlemen. Thank you for standing by for KE Holdings Second Quarter 2026 Earnings Conference Call. I am Siting, IR Director of KE Holdings. Please note that today's call, including the management's prepared remarks and Q&A session with all being Chinese. Simultaneous interpretation in English will be available on a separate line for the duration of the call. To access the gold in Chinese, you will need to dial in to the Chinese language line. [Operator Instructions] Today's conference call is being recorded. The company's financial and operating results were published in the press release earlier today and are posted on the company's IR website. On today's call, we have Mr. Stanley Peng, our Co-Founder, Chairman and Chief Executive Officer; and Mr. Tao Xu, our Executive Director and CFO; Mr. Xu will provide an overview of our business update and financial performance. Then Mr. Pang will share more on the progress of our strategic transformation. Before we continue, I refer you to our safe harbor statement in our earnings press release, which applies to this call as we will make forward-looking statements. Please note that KE's earnings press release and this conference call include discussions of unaudited GAAP financial information as well as unaudited non-GAAP financial measures. Please refer to the company's press release, which contains a reconciliation of the unaudited non-GAAP measures to comparable GAAP measures. Lastly, unless otherwise stated, all figures mentioned during this call are all in RMB. Certain statistical and other information relating to the industry in which the company is engaged to be mentioned in this call has been obtained from various public available official or unofficial sources. Neither the company nor any of these representatives has independently verified such data, which may involve a number of assumptions and limitations, and you are cautioned not to give undue weigh to such information and assets. For today's call, management will review the Chinese as the main language. Please note that English translation is for convenience purposes only. In case of any discrepancy, management statements in their original language will prevail. We now I will turn the call over to our CFO, Mr. Tao Xu.

[Interpreted] Thank you, Siting. Hello, everyone. Welcome to our Q2 2026 earnings call. Let me begin with the key financial takeaways. Our total GDV returned to growth. Despite a modest year-over-year revenue decline, profits increased significantly, materially outperforming both GTV and revenue. In Q2, increased 6.3% over year-over-year, while revenue decreased to 5.7% year-over-year. This revenue decline stemmed primarily from adjustments in our home renovation and furnishing business and revenue recognition impacts from iterative product modeling, home rent services. Non-GAAP Net income grew 74.9% year-over-year to CNY 3.185 billion. Non-GAAP net margin reached 13%, up 6 percentage points year-over-year, a 3-year high. Profit improvements were driven by a healthier cost structure, strict financial discipline and a higher operating efficiency. Contribution margins across all core business lines improved year-over-year and quarter-over-quarter, driving the group's gross margin up 6.7 percentage points year-over-year to 28.6%. Simultaneously, GAAP operating expenses fell 14.1% year-over-year. This combination of gross margin expansion and the lower operating expenses fueled our profit growth. Next, I'll review our segment financial performance. First, existing home transaction services. Q2 scale return to growth and profitability improved significantly. GTV reached RMB 629.89 billion, up 8% year-over-year and 17.9% quarter-over-quarter. Revenue was RMB 7.02 billion, up 4.5% year-over-year and 14.5% quarter-over-quarter. GTV outpaced revenue growth year-over-year, primarily because of non-Lianjia GTV, where platform service fees are recognized on a net base accounted for a larger ship. This quarter, non-Lianjia platform service revenue increased to 27.8% year-over-year and 29.8% quarter-over-quarter. With a stable network scale, we advanced refine our operations to boost per store output, helping connected to stores outperformed the market and enhancing overall platform efficiency. Q2 contribution margin reached 46.1%, up 6.1 percentage points year-over-year, driven by a lower fixed labor costs and a structural shift toward a higher-margin platform service revenue. It also rose 4.8 percentage points quarter-over-quarter, benefiting from operating leverage and net revenue recovery and further business mix improvements. Second, the new home business. Q2 scale remained stable year-over-year, while profitability continued to improve. GTV who reached RMB 258.39 billion, up 1.2% year-over-year and 77.1% quarter-on-quarter. Revenue reached RMB 8.95 billion, up 3.8% year-over-year and 75.9% quarter-over-quarter. Despite a pressured market, we maintained stable scale by collaborating on high-quality projects, improving customer conversion and optimizing costs. Q2 contribution margin reached 28.8%, up 4.4 percentage points year-over-year, driven by cost structure optimization from refined operations, it also rose 3.1 percentage points quarter-over-quarter benefiting from the same factors plus operating leverage from revenue growth. Third, home renovation and furnishing. Q2 revenue was RMB 3.19 billion, down 30.1% year-over-year and up 36.4% quarter-over-quarter. The year-over-year decline reflects our proactive adjustments of inefficient customer acquisition channels and exit from cities with a weak unit economics. New home market pressures also dampened renovation demand. The quarter-over-quarter revenue increase reflects seasonal business recovery. Q2 contribution margin was 39.6%, up 7.5 percentage points year-over-year and 3.4 percentage points quarter-over-quarter, driven by lower material costs through centralized procurement and refined cost management. Fourth, home rental services. Q2 revenue was RMB 4.83 billion, down 14.8% year-over-year and 3.6% quarter-over-quarter. This stemmed from transitioning Carefree Rent to a lighter low-risk product model utilizing net basis revenue recognition. While this reduces reported accounting revenue, managed rental units continued rapid growth. By end of Q2, managed units exceeded 790,000, up approximately 34% year-over-year with a net base product comprising over 50%. Q2 contribution margin reached 15.3%, up 6.9 percentage points year-over-year. This reflects a favorable product mix shift in operating improvement from lower labor, installation and post lease costs. Quarter-over-quarter contribution margin rose 0.5 percentage points, driven by a continued increase in net based products. Fifth, emerging and other businesses. Q2 revenue reached RMB 550 million, up 26.4% year-over-year and 7% quarter-over-quarter. Next, turning to cost expenses and the profits. Q2 store-related costs were RMB 560 million, down 25.9% year-over-year and broadly stable quarter-over-quarter. The year-over-year decline reflects Lianjia's rent cost optimization and network adjustments. Total Q2 gap operating expenses were RMB 3.989 billion, down 14.1% year-over-year, driven by improved organizational efficiency, optimized marketing spend and continue to financial discipline. Operating expenses rose 21.3% quarter-over-quarter due to higher selling expenses from the home renovation seasonal recovery and bad debt provisions in new home business. Specifically, G&A expenses were RMB 2.04 billion, down 2.1% year-over-year. The 18.9% quarter-over-quarter increase resulted from a full bad debt provision of around RMB 280 million following the prudent assessment of Seneca-related receivables and collateral. So the marketing expenses were RMB 1.4 billion, down 26.1% year-over-year due to optimized sales, personnel costs and refining marketing spend, but rose 29.6% quarter-over-quarter from seasonally higher rental -- home renovation selling expenses. R&D expenses were RMB 550 million, down 13.4% year-over-year due to lower labor and technical service costs, but up 11.4% quarter-over-quarter due to increased technical service fees. On the bottom line, Q2 GAAP operating profit reached RMB 3.026 billion, up 185.6% year-over-year. Non-GAAP operating profit was RMB 3.59 billion, up 123.6% in the year. GAAP operating profit rose 137.8% quarter-over-quarter with a 12.3% margin, up 8.3 percentage points year-over-year and 5.6 percentage points quarter-on-quarter. Non-GAAP operating profit grew 115.7% quarter-over-quarter with a 14.6% margin, up 8.5 percentage points year-over-year and 5.8 percentage points quarter-on-quarter. This year-on-year and quarter-over-quarter margin expense was driven mainly by higher gross margins and lower operating expenses for issues. Q2 GAAP net income was RMB 2.624 billion, up 100.8% year-over-year and 109.1% quarter-over-quarter. Non-GAAP net income was RMB 3.185 billion, up 74.9% year-over-year, 97.6% quarter per quarter. Finally, turning to cash flow, balance sheet and shareholder returns. Our Q2 net operating cash inflow was RMB 6.61 billion, net new home account receivable tenor was around 39 days, down around 12 days year-over-year, reflecting the effective risk management. Excluding customer deposits, our end of Q2 broad cash balance remain at around RMB 67.3 billion. This ample liquidity strengthened our risk and resilience whilst supporting business development and shareholder returns. In Q2, we spent around USD 250 million on share repurchases, including our first buyback in the Hong Kong market. In first half, we spent around USD 450 million on repurchases, up around 14% year-over-year, representing around 2.4% of our year-end 2025 outstanding shares. Since launch of this share repurchase program in September 2022 through Q2 2026, we have repurchased around USD 2.99 billion in shares, representing around 14.8% of outstanding shares prior to the program start. In summary, Q2 profitability improvements reflect combined cost optimization, operating enhancements and a favorable business mix. Looking ahead, maintaining a solid balance sheet and ample liquidity will anchor our long-term growth. Across all core new and technical investments, we will enforce strict ROI display and take customer value operating efficiency and sustainable returns as our priority or key metrics. Ultimately, we will balance business development with the shareholder returns to consistently create long-term value. Next, I'll turn the call over to our Chairman and CEO, Mr. Stanley Peng.

[Interpreted] Thank you. Investors and analysts, good evening. So last quarter, we discussed our shift towards consumer-centric transformation this quarter, I will talk about how the changes translate into our operations. In Q2, I observed 2 trends of operation foundation stabilized and our organization truly organized. So this foundation enables the long-term change. I will address 5 key questions. The first one, what changes as transformation enter failure operations; second, that in consumer-centric meaning bypassing agents; thirdly, if AI advances will, even become obsolete; fourthly, how is AI plant in our business and what is result; fifthly, how will we know we are on the right track moving forward. [So for the first question, what's changed as transformation enter failure operation. In this quarter, I spent a lot of time on the front line [indiscernible] stores, properties, construction sites and discussing issues with clients, agents and store owners that changes or down to 3 areas. First, refined operation we are shifting from the 1 side fit oil approach to the district specific and the projects, specific strategies. So rather than checking a single city-wide metric, we analyze specific districts or projects to [indiscernible] solutions and what is the solution for each [indiscernible]. For example, in the high-end community where clients feel property across districts, our legacy geographic bound model field, and we regrouped operational units based on extra clients viewing [indiscernible] assigning project efforts while professional presentations and client experts to address specific family needs. So with 600 projects driving half the cities volume, standardizing these professional judgment into a clear division of labor allow us to replicate this model and other cities have become similar operations -- exploration. Second shifted metrics, scale and market share still matter, but now we focus more on consistent agent transactions, writing agent efficiency and income healthy store profitability and stable service quality. So leasing LSG plays perfectly. In 2025, we have at most 700 agents for leasing at the peak and that the average agent efficiency fell below 2 transactions. Instead of adding headcount, we divided it into smaller blocks, rematching properties, clients and agents based on familiarity and the capabilities. So from April to July, average agent efficiency jumped from 3 to 5.6 transactions and zero transaction ratio dropped from nearly 25% to under 10%. So I think what matters is but the effective organization matters more than near headcount. Thirdly, mobilized people. Managers have elected meeting rooms for front line. This quarter, managers personally sold [indiscernible] listings, revisited debt leads and accompanied agents to signing centers. And my only requirement for manager is to presence. You cannot learn to swim without getting into the water. So in short, operationalizing transformation wins refined operations, shifted metrics and mobilized people. So this stems from a sing approach, solving real consumer and frontline problem first, than reorganizing our people resources in the platform. So we are moving towards changes and that they are now being seen in operational units. The second question is does being consumer-centric meaning bypassing agents? So this assumes that if the platform moves closer to the consumer, it must take from the agents. Historically, we only split a single transaction commission which is a zero-sum game. So -- and this is what we did in the past. But to break this equation, we must create more light, create more high-value tasks not just redivided the same money. Consumers are changing. Good is to be aesthetic property [indiscernible] today by I think good means a proper match [indiscernible] determine goods expanded from 1 to 3, the property, the family situation and also the service provider. So the service provider is now a vital variable, not just a conduit. So if decision becomes harder, tasks must be segmented. There are 3 reasons. First, the required knowledge. It feeds one's personal capacity. For example, we needed to know the properties, clients, circumstances, mortgage and renovations and the furnishing business. So this exceeds one's capacity. Second, building expertise required a mutually exclusive path. You must either deeply route yourself in one project or follow a group of clients. So you cannot do both simultaneously. So that is the second reason. The third one is the most valuable action has shifted from providing options to confidently eliminating them. I think we are not only offering more choices to the consumers, and in start, we needed to help them to filter. However, filtering does not mean transaction. As long as income relies solely on closings, true professionalism won't develop. I think professionalism must be financially viable. So therefore, we are [indiscernible] income from closed deals, lining them entirely with the buyers or seller. This assisted [indiscernible] is the client manager. So previously, platform, in fact, stopped once a lead reach agent. The client manager insures continuity. AI organizes data and while human services [indiscernible] stage and needs. The agent received fully provided clients and because the client managers are not paid per transaction and that they remain purely objective. So as I have mentioned, managers are not paid per transaction. So from May to July, this will handle over 50,000 [indiscernible], achieving a 7.4% related to showing conversion rates outperforming the broader market 5%. So the platforms mentioned is evolving from leading commission to building a structure where every specialized scale is independently verified and compensated. ECM is shifting from a single listing workflow to a modular ecosystem, which includes consulting, showing, contracting, reporting, marketing materials, renovation and leasing and so on. So anyone creating incremental value is a service provider. And this is our definition, which is expanded. So the main goal is enabling professional service providers to win in the long term. So being consumer centric means transforming the single agent into a group of independently valuable specialized roles. And now we have the help of AI, which gives us more impetus. So the question, as AI advances, will agents become obsolete? So this assumes agents only sales static information easily fetched by AI. However, technology reshuffles value. And we are -- some things depreciate while others become scarce. So we should ask what is depreciating and what is becoming more scarce? So for the scarce part, what kind of progress that platform managed service provider can make? So what is depreciating? Static information, [indiscernible] revenue, price and year build and also the layout of the house. So I think this kind of information cannot support the decision-making, and it is very easy to get. So if we only transmit -- we only transport information, we may have no more opportunities going forward. So what is scarce dynamic, deep inspiring in fact and they cannot be fabricated, for example, the reason of selling, renovation potential or local market assessment from seasoned managers, and how is the situation in the communities by the managers and what is the closings and how was the deal last time. And we -- this information lives in people's minds and the industry likes the pipeline to capture and reuse age. And fundamentally, AI does not be the consequence of poor decisions. And AI may not take any accountabilities. So as the cost of housing mistake rises, consumer needed to reduce uncertainty growth. Therefore, 3 things will happen. Firstly, the industry becomes more valuable by mitigating uncertainty, segment creating valuate hard, requiring deep data and deeper surveys. The first thing is those who transform in the direction become more valuable, including platforms and managers. So we do not need information with players we need professional who dare to make judgment and take responsibility. So the previous question is about the industry and service provider. And if we look around and if we look inward, then what goes -- comes to the question for how is AI applied in our business and with what results? Actually, the business itself is a reduction function, what is our input and what is outgrow, and there is human capital labor capital and technology in the function. So in today's AI, we should know the situation in AI in the industry. So a subitem or direct variables. So it's the supply units and efficiency tool or if it is a direct variable, it requires attrit. So we needed to change attitudes in the first, we now also open some of the foundational pages, and we are lowering the threshold. So we are worried about whether there will be disruption. And we are thinking about how AI can be a new production factor rather than an opponent enables innovation. So I think the consumers finally [indiscernible] value. I think the consumers need a better experience, and we need to solve the problems of consumers. The second is it changes management. In the recent 200 years, we have improvement in the finance and management, and we need quantifiable data in the management. And I think we all benefit from this methodology. In KE Holdings and also Lianjia, we need standard, and we also need tools for the improvement. However, for the unquantifiable, they cannot be measured. This is also a big problem. But sometimes, we may only focus on that number, and we find that -- sometimes we find that the numbers are too abstract and the consumers now become the numbers and also become the #1 in the standard. However, with the help of AI, AI brings the unstructured data and [indiscernible] language and the numbers are totally different information and signals and the granularity shifts from the managing average to managing individual properties, clients and agents. Previously, we managed the average, but now we have the computation power and the knowledge, and we can have the tailored solution for each individual. And this third party is about AI changes division of labor. We talk about the segmentation of the task or in the company by AI. Now we have -- these scenarios, which include financial human resource product technology and also from the stage by stage and the computation power. But now, we have are breaking down the threshold and all of them are in the computation power of AI. And previously -- so the old divisions merge the new ones emerge. In our Lianjia new home business, we shifted a labor between humans and AI. AI helps saves compare proposals using a dynamic and knowledge base allowing agents to focus on understanding clients. So the agents could fine-tune their understanding of the client. So this produces both close deals and also useful organizational capabilities. So these only come from the front line. So this disruption reshapes the organization. So it concerns on 4 things. First is cost. AI lower fixed costs and increases variable costs, enabling rapid iteration. So whoever iterates fast, who creates more value. And the next is a trial and error. So in the past, it takes a lot of efforts. Right now, throughout the -- it takes a long path to evaluate, test, validate a proposal. So the bigger the organization, I mean, the longer the chain is. So many people just hesitate. So right now, AI shifts innovation from heavy slow investments into a high frequency and low cost profitability gains. So this allows us to trial and test the multiple models at the same time, and we have a higher probability of winning out the game. Next is the front line and to the mid-office. So the front line workers [indiscernible] can rapidly build and test the solutions, so the mid-office can then scale them. Last, but not least, managers. So in the past, the bigger the organization, so the lower the efficiency is. Right now, I actually talked to a lot of the managers. We don't feel like a lot of percent of value. Right now, AI [indiscernible] organization is changing the role, handling the reporting. We're forcing managers to stop being megaphone and start creating real business value. So they are not just simply presenting the numbers, they are actually creating real genuine value from the frontline. Because they are in the process of creating the value. Last, finally, the bottlenecks shifts to humans. Look at KE. We have a long industrial process. AI can perfect a lot of the workflows. And those where the human invention becomes the bottleneck, so there is this human and human miner action that AI can have to replace. So whether we can unite people together and provided them with the training, allow them to work efficiently with AI, so one is culture to the other is evolution. So this is essentially a change that we are talking about towards the whole industry. Now back to the very first question, whether AI is a direct variable? Because it changes who we serve our judgments, our process and our organization. So this is a direct of variable. That means we're not simply installing AI into the company, we are regrowing the company with AI. So looking into the next phase, how we will know we're on the red track moving forward then we're most separate 2 things, where we need to place heavy bets from [indiscernible] seek answers? I think there are 3 areas we are placing heavy bet, deep service, deep data and a platform ecosystem. So as information democratic [indiscernible], deep data becomes scarce and the harder the decision-making becomes and the deeper service becomes more valuable. As labor specializes, a platform is needed to orchestrate it. So while we are seeking answers, AI's final form and the ultimate structure of the management and expertise remain uncertain. Directional matters require [indiscernible] bets. So how do we capture users' evolving need? So management, of course, carries this value. So more for logical matters require small investments, rapid testing and cutting losses early. So why do we need to separate [indiscernible] by certainty? Because, again, we have already proven that directive matters requires unwavering bets, whereas the morphological matters require more investment in rapid testing. So looking back at the past 2 quarters, we have approved in some areas that keeping investment in areas with low marginal returns is meaningless. The purely skill-driven model is debt. We should stop those meaningless investments. Most of all, we have to validate 4 things. First, professionals. Facing AI, whether they can use it directly or indirectly to create value? Do they have a new definitions for what is professionalism and whether they are committed to this concept. And second, for managers, whether they can return to the front line and produce high-quality judge ment to recreate this sense of value. The third is the processes and judgments. With the deeper services, can they earn the trust from their customers, whether they can earn a broader recognition -- a better recognition or trust. Number four, organizational capabilities, can we turn a single success into a replicable capability? So in such a discontinuous transformation, so for me, industries, they are pretty much faced with the same challenge. The way I see it, human conviction is the leading indicator, numbers are the lagging indicator. So many of the management tend to hide their expertise within themselves. So without the open sharing, we can also make that into replicable successful model. So our core test is whether we can consistently execute consumer centricity and enable professionalism to win. This must be embedded in our culture workflows. So we'll mainly success across 4 pillars, customer, service provider, operations and replicability [indiscernible]. So if you look at these 5 things, so we have to brief define our playbook. So consumers are facing harder decisions to make. So that is driving deeper specialization. So the AI is depreciating role [indiscernible] into while elevating true expertise and reorganizational internal work. So our direction is certain, deep service, deep data and a platform ecosystem. So Q2 is not the conclusion, it is just the beginning. Thank you. I'll now turn the call to the analysts for Q&A.

Thank you, Stanley. [Operator Instructions] The first question comes from Timothy Zhao from Goldman Sachs.

Congratulations on the strong Q2 results. My question is on the overall property market. It saw a diverging trend in volume and price in Q2, with some fluctuations and momentum in Q3. Given the uncertainty ahead, what controllable levers does the company have for Q3 and the full year?

Thank you, Timothy. In the first half, the existing home market showed a structural recovery in transactions with prices bottoming. In Q2, this recovery became more evident to pay parity across cities and price segments. By [indiscernible], transaction volumes recorded faster in Tier 1 cities, where the first half prices also showed a greater sequential resilience. In Q2, year-over-year growth in registered existing home transactions in Tier 1 cities outpaced other cities. According to Baker Research Institute, in the first half, Tier 1 existing home prices rose cumulatively by 3.6% quarter-over-quarter, while national prices remained broadly stable year-over-year. Prices across all tiers remain in an adjustment based. For our platform, volume for lower-priced homes grew faster than mid- to high-priced homes. However, the transaction mix across unit sciences remains stable, indicating housing demand hasn't broadly downgraded to smaller homes. Instead, this reflects a downward shift in transaction price bands as prices adjusted. Meanwhile, higher-priced homes saw smaller year-over-year price declines, showing resilience in core upgrade-oriented and high-quality residences. In the new home market, overall Q2 volume remained under pressure. Projects in core cities where the strong product offerings showed better support. Structurally, existing homes accounted for over 80% of the total national residential transaction area in the first half becoming the market [indiscernible] for housing demand. Overall, we see a structural transaction recovery while prices continue to bottom, core cities and high-quality supply and more resilient, but the market remains pelorized. With more property choices, customers are deciding cautiously, valuing professional judgment and transaction certainty. They need professional decision support, not just transaction matching or facilitation. This highlights our platform's accumulated service capabilities. Based on this, we will focus on 3 areas. First, capturing structural market opportunities to strengthen revenue resuming. We will allocate resources based on market performance across cities, customer groups and property tax reinforcing coverage in higher-tier cities. Meanwhile, centered around the content treatment engagement, precise matching and professional execution will help customers make better decisions and convert genuine demand into transactions. Second, we'll continue to reinforce financial discipline and flexible resource allocation. Our leaner cost structure improves our ability to [indiscernible] market volatility. If pressure persists, we will dynamically allocate resources prioritizing our core professional service provider network over short-term profit. Even if the market improves, we will not return to extensive expansion. New investments [indiscernible] ROI and service validation before scaling, ensuring transactions translates efficiently into profits and cash flows. Third, we'll also prioritize cash flow and a solid balance sheet. We'll strictly manage receivables and collections, control risk exposure and limit nonessential investments to preserve flexibility. Therefore, our second half operations will not rely on market events. On the revenue side, better decision support will help us win more customers. On the financial side, our healthier cost structure will protect cash flow and core capability in weak markets and release greater operating leverage on market improved. Thank you.

Our next question comes from John Lam from UBS.

Thank you, Tao, for your answering. So my question effect in Q2, the profit of pace revenue growth significantly. So could the management break down the impact of business performance, operating efficiency, expense baseline? And if there is any one-off factors? And for those improvements, how sustainable they in the long run?

Thank you for your question. In Q2, the profit improvement was mainly driven by higher contribution margins across the core business and the lower operating expenses. For the core business contribution margins, they improved year-on-year and quarter-on-quarter, driving the group's gross margin up 6.7 percentage points year-on-year to 28.6%. At the same time, GAAP operating expenses fell 14.1% year-on-year. There are 3 drivers: first, lower cost and expenses baseline over the past year were optimized Lianjia [indiscernible] and agent structure by expanding management expense, consolidating resources and reducing low productivity investment. And this lower fixed labor cost and our breakeven point, so we also have a persistent baseline. Second, increased operating efficiency in housing transaction in new homes generating coverage or high quality projects and improving customer conversion enhanced transaction resilience. We also have stable monetization and a better channel efficiency drove profit growth. And for the existing homes, focusing on the priority, listings and refine operational support for connected stores, significantly [indiscernible] connected store revenue and profit contribution, effectively increased the unit economics and the business mix in new business. So we help centralize procurement and refined cost management, lower the material cost ratio in home renovation. In rental services contribution margin improved due to a mix shift toward a net basis revenue product alongside the January operating improvements in labor, installation and post lease costs. Looking ahead to the next 2 quarters, under a mutual market assumption, the lower cost baseline will contribute to support profit driver marketing channel incentives and certain frontline sales cost may fluctuate quarter-on-quarter due to revenue scale, mix and also seasonality. We will not simply extrapolate a single quarter's profit, but focus on achieving balance revenue and profit growth. So if the market improves, incremental revenue growth relate stronger operating leverage from the lower baseline creating greater profit upside. If pressure continues over healthier construct reduces profit sensitivity to market volatility and simply put our current structure increases both upside potential and downside protection. But in the long run, this optimization builds a healthy operating foundation. This step -- this is Step 1 out of 1 of our strategic transformation, optimizing resources allocation for current market, and this is how we can place the uncertainty. Step 2 is directing limited resources toward initiatives that create customer funding rather than just cutting cost. And ultimately to work through workflows, evaluations inventive and the platform tools, we will impact efficient resource allocation into our dealer organizational capacities to support a sustainable growth.

Thank you, Mr. Tao. The next question comes from Xiaodan Zhang from CICC.