Good day, ladies and gentlemen. Thank you for standing by. Welcome to Alibaba Group's June Quarter 2024 Results Conference Call. At this time, all participants are in a listen-only mode. After management's prepared remarks, there will be a Q&A session. I would now like to turn the call over to Rob Lin, Head of Investor Relations of Alibaba Group. Please go ahead.

Good day, everyone, and welcome to Alibaba Group's June quarter '24 results conference call. With us are Joe Tsai, Chairman; Eddie Wu, Chief Executive Officer; Toby Xu, Chief Financial Officer. We have also invited Jiang Fan, Co-Chairman and CEO of Alibaba International Digital Commerce Group to join the call. This call is also being webcasted from the Investor Relations section of our corporate website. A replay of the call will be available on our website later today. Now let me quickly cover the safe harbor. Today's discussion may contain forward-looking statements, including without limitation statements about our strategies and business plans as well as our belief, expectation, and guidance about our business prospects such as the future growth of our business, revenue take rate, profitability, and return on investments and share repurchases. Forward-looking statements involve inherent risks and uncertainties that may cause actual results to differ materially from our expected expectations. For detailed discussion of these risks and uncertainties, please refer to our latest annual report on Form 20-F and other documents filed with the U.S. SEC or announced on the website of the Hong Kong Stock Exchange. Any forward-looking statements that we make on this call are based on assumptions as of today, and we do not undertake any obligation to these statements, except as required under applicable law. Please note that certain financial measures that we use on the call, such as adjusted EBITDA, adjusted EBITDA margin, adjusted EBITA, adjusted EBITA margin, non-GAAP net income, non-GAAP diluted earnings per share or ADS, and free cash flow are expressed on a non-GAAP basis. Our GAAP results and reconciliation of GAAP to non-GAAP measures can be found in our earnings press release. Unless otherwise stated, growth rate of all stated metrics mentioned during this call refers to year-over-year growth versus the same quarter last year. With that, I will turn over to Eddie.

Hello, everyone. In the new fiscal year, our strategies of user first and AI-driven are starting to bear fruit. This quarter's results reflect continued steady growth momentum. Taobao and Tmall Group achieved steady year-over-year growth in orders and GMV. Alibaba International Digital Commerce, AIDC, revenue maintained strong growth. Alibaba Cloud revenue, excluding Alibaba consolidated subsidiaries, returned to positive growth, notably driven by core public cloud and AI products. Across other segments, we achieved better operational efficiency and monetization capabilities to varying degrees, which drove significant improvement in some loss-making businesses. Taobao and Tmall Group maintained its user first strategy and continue to invest in key capabilities aimed to deliver quality products and services at attractive prices across diverse shopping needs of consumers. This quarter saw steady growth in GMV and order volume driven by a notable increase in purchase frequency. Internal and external data about e-commerce players' market share dynamics indicates a positive trend in our gradual market share stabilization, further validating the effectiveness of our investments into experience. 88VIP members reached 42 million by quarter end. We will continue to explore and improve program benefits, product offerings, and service experiences for this expanding premium member base. In TTG's operational strategy, we attach great importance to rich and diverse product offerings while focusing on investing and enhancing shopping experiences. We continuously improve the efficiency in matching our products with user traffic and interest stable and sustainable growth. As orders and GMV continue to grow, we're advancing monetization step-by-step, including the launch of our new marketing tool, Quanzhantui. Through algorithm upgrades, we aim to enhance the efficiency of merchants’ monetization while also improving the efficiency of traffic utilization and conversion of user traffic to purchases on our platform. We expect CMR growth to gradually align with GMV growth over the coming quarters. In our Cloud segment, we continue to pursue high-quality revenue and effectively execute our integrated cloud plus AI development strategy. This quarter, Alibaba's overall revenue, excluding Alibaba consolidated subsidiaries, grew 6%, with public cloud revenue maintaining double-digit growth. AI-related product revenues sustained triple-digit growth, continuing to increase its share of public cloud revenue. We're seeing more major customers choosing Alibaba Cloud as their computer infrastructure for AI development. At the same time, Alibaba's proprietary large language models are gaining wider adoption. This year, Alibaba Cloud served as a major cloud service provider for the Olympic Games, providing cloud computing and AI services to Olympic broadcasting services, OBS. At the Paris Olympics, cloud-based live broadcast powered by Alibaba Cloud over satellite signals as the primary means of broadcast for the first time in Olympic history. Two-thirds of broadcasters used live signals transmitted by Alibaba Cloud and real-time around the world, reaching billions of viewers. Additionally, Alibaba Cloud's AI technology saw its first widespread application at the Paris Olympics, including 360-degree replays in real-time, bringing pleasant surprises to the audience. Our long-term strategy for integrated cloud plus AI development comprises three key elements. One, we'll continue to optimize our cloud product offerings, focusing on competitive sustainable growth margin and scalable public cloud products. This forms the foundation for Alibaba Cloud's sustainable, high-quality growth. Two, we'll strengthen synergies between cloud and AI products, not only supporting existing customers and implementing new AI capabilities on Alibaba Cloud, but also enabling AI native enterprises to scale and succeed on our platform. We're committed to capitalizing on both opportunities. Three, we'll continue to invest in R&D and AI CapEx to ensure the growth of our AI-driven cloud business. Looking to the medium and long term, we are confident that Alibaba's overall revenue, excluding Alibaba consolidated subsidiaries, will return to double-digit growth in the second half of the fiscal year with gradual acceleration thereafter. Through intensive R&D investment, we aim to sustain profitable growth while establishing ourselves as a leading cloud service provider for AI with healthy probability and market share leadership. In international e-commerce, AIDC delivered 32% revenue growth this quarter. Our International business segment comprises a diverse business matrix, integrating both cross-border and local offerings as well as B2B and B2C formats. This presents numerous opportunities in terms of both operational models and market expansion potential, and Jiang Fan will provide a more detailed update momentarily. Beyond e-commerce and cloud, we've implemented strategic realignments across our key Internet technology businesses through a thorough evaluation of their product capabilities and market competitiveness while maintaining product competitiveness. Most of these businesses will now place a higher priority on monetization. This quarter, we've already seen these businesses significantly improve in profitability, and we expect this trend to continue in the coming quarters. We expect most of these businesses to break even within one to two years and gradually contribute to profitability at scale. This new fiscal year is a pivotal one for Alibaba as we set our strategic direction and realize the successes of our transformation. We are fully confident in Alibaba's sustained healthy development into the future.

Hi, everybody. It's my pleasure to share with you AIDC's latest business progress. Over the past quarter, AIDC maintained its rapid growth momentum with overall revenue increasing 32% year-on-year, primarily driven by our cross-border business. In terms of the three drivers we consistently focus on. First, ongoing business model and supply chain upgrades, Ali Express maintained rapid year-on-year order growth with the percentage of AE choice orders continuing to be high and gradually stabilizing, enhancing the certainty and consistency of user experience. Ali Express and China continue to jointly optimize logistics variance and significantly reduced average delivery time quarter-over-quarter. We further optimized the organic balance of user experience, product richness, and efficiency across our marketplace, semi-consignment, and full consignment models. The unit economics of the Choice business improved by nearly 20% quarter-over-quarter. Second, product and technology innovation. We advanced our AI and intelligent technologies across scenarios like AI customer service, cross-platform product placement, product description optimization, multilingual search, and price recommendations, which boosted efficiency and user experience. These innovations were deployed across our platforms and have served around 500,000 merchants and covered 100 million SKUs. Third, sustained growth in key markets. Ongoing improvement in user experience bolsters AE's confidence to step up investments in key markets to expand user base and maintain leadership as the exclusive e-commerce partner of the European Cup 2024. We effectively boosted AE and Trendyol's brand awareness. We're also exploring diverse localized business models. In Turkey, Trendyol increased monetization and profitability, solidifying its e-commerce leadership and also enriched product offerings and marketing efficiency in the Gulf region while continuing in user growth. In Southeast Asia, Lazada achieved single-month EBITDA profitability for the first time in July. This milestone reinforces our confidence to efficiently invest in the Southeast Asia market to consolidate our market share. So, in conclusion, our strategic focus is twofold: enhancing operational efficiency across all business segments, while on the other hand, actively investing in key markets to drive quality growth and scale with the goal of achieving more substantial profitability at scale in the future. Thank you.

This past quarter's financial performance continues to demonstrate our strategy to revitalize growth in our domestic commerce segments while reinforcing our leadership position. At the same time, we are improving the monetization and/or efficiency of our loss-making businesses with the goal of achieving sustainable business growth and profitability. Taobao and Tmall Group is winning the mind share of our consumers. This past quarter, we achieved high single-digit online GMV growth and double-digit order growth. AIDC continues to achieve robust revenue growth driven by rapid order growth in cross-border businesses, especially from Ali Express Choice. Alibaba Cloud's revenue quality is improving back to growth. Overall revenue, excluding Alibaba consolidated subsidiaries, grew over 6%, driven by double-digit public growth and increasing adoption of AI-related products. Our loss-making businesses are improving their monetization and operating efficiency significantly. For example, the local service group and Lazada significantly narrowed their losses during the quarter. During this quarter, we repurchased a total of 630 million ordinary shares or 77 million ADS for a total of $5.8 billion. This includes the concurrent buyback of approximately 14.8 million ADS for $1.2 billion that was associated with our new CV issuance. As of June 30, 2024, we had 90 billion ordinary shares or approximately $2.4 billion ADSs outstanding. Compared to March 31, 2024, there was a 2.3% net reduction in our outstanding shares after accounting for shares issued in our ESOP. As of June 30, 2024, we still have $26.1 billion remaining in our share repurchase program. Additionally, as part of our plan to minimize annual dilution and better utilize the cash generated by our domestic businesses, we have started to replace a portion of Alibaba Group's ESOP incentives with long-term cash incentives for our employees starting last quarter. These cash incentives will be reflected as costs in our adjusted financial metrics, including segment EBITDA. While this change in compensation structure will lower adjusted EBITDA, it will also result in less ESOP dilution in the future. Total revenue. On a consolidated basis, total revenue was RMB243.2 billion, an increase of 4%. Adjusted EBITDA decreased 1% year-over-year to RMB45 billion. Excluding the effects of long-term cash incentives, adjusted EBITDA growth would have turned positive on a like-for-like basis. Our non-GAAP net income was RMB40.7 billion, a decrease of RMB4.2 billion or 9%. Our GAAP net income was RMB24 billion, a decline of RMB9 billion or 27% primarily due to a decrease in income from operations and the increase in impairment of some investments, partly offset by the mark-to-market changes from our equity investments. As of June 30, 2024, we continued to maintain a strong net cash position of RMB405.7 billion or $55.8 billion. Free cash flow decreased by RMB21.7 billion to RMB17.4 billion. This year-over-year decrease mainly reflected the increase in expenditure related to our investments in Alibaba cloud infrastructure and other working capital changes related to factors including our planned reduction of direct sales businesses. Now let's look at cost trends as a percentage of revenue, excluding SBC during this quarter. Cost of revenue ratio decreased 1.1 percentage points. Product development expenses ratio remained relatively stable. Sales and marketing expenses ratio increased 1.7 percentage points year-over-year, primarily due to our increased investments in e-commerce businesses. G&A expenses ratio increased 1.4 percentage points. Excluding the provision of RMB3.1 billion from a one-time shareholder class action lawsuit, our G&A expenses ratio would have remained relatively stable. Now let's look at the segment results, starting with Taobao and Tmall Group. Revenue from Taobao and Tmall Group was RMB113.4 billion, a decrease of 1%. Revenue from our China commerce retail businesses was RMB107.4 billion, a decrease of 2% compared to RMB109.8 billion in the same quarter of 2023. Customer management revenue increased by 1% year-over-year, primarily due to a high single-digit year-over-year growth in online GMV, partly offset by a decline in take rate. The year-over-year take rate decrease was primarily due to an increasing proportion of GMV generated from new models that currently have lower monetization rates. Direct sales and other revenue on the China commerce retail business was RMB27.3 billion, a decrease of 9%. As mentioned, we are proactively scaling down certain direct sales businesses, including those in Taobao and Tmall Group. The China commerce wholesale business revenue increased 16% to RMB6 billion, primarily due to an increase in revenue from value-added services provided to paying members. Taobao and Tmall Group adjusted EBITDA decreased by 1% to RMB48.8 billion, primarily due to the increase in investments in user experience and technology infrastructure, partly offset by the narrowing losses in certain businesses. But this increasing investment led to better consumer retention, increased purchase frequency, and positive feedback regarding the overall shopping experience. Revenue from Cloud Intelligence Group was RMB26.5 billion, an increase of 6%. Overall revenue, excluding Alibaba consolidated subsidiaries, grew over 6% year-over-year, driven by double-digit public cloud growth and increasing adoption of AI-related products. AI-related product revenue continued to grow at triple digits year-over-year. Cloud's adjusted EBITDA increased by 155% to RMB2.3 billion, the increase was primarily due to improving product mix through our focus on public cloud adoption and operating efficiency, partly offset by increasing investments in customers and technology. We're observing strong and sustained demand for AI products and solutions from our customers and are making significant investments to address this demand effectively. We believe our investments in AI capabilities and infrastructure will help strengthen our market leadership. We are confident in our ability to balance these investments with steady profitability improvements, ensuring sustainable growth and value creation. Revenue from international commerce retail business increased 38% to RMB23.7 billion, primarily driven by order growth from Ali Express' Choice as well as improvements in monetization. Revenue from our international commerce wholesale business increased by 12% to RMB5.6 billion, primarily due to an increase in revenue generated by cross-border related value-added services. AIDC's adjusted EBITDA was a loss of RMB3.7 billion, an increase of RMB3.3 billion compared to a loss of RMB420 million in the same quarter last year, primarily due to the increase in investments in Ali Express and Trendyol's cross-border business, partly offset by Lazada's significant reduction in operating loss from the improvements in its monetization and operating efficiency. Revenue from Cainiao Smart Logistics Network Limited was RMB26.8 billion and increased 16%, primarily driven by the increase in revenue from cross-border fulfillment solutions. Cainiao's EBITDA decreased by 30% to RMB618 million, primarily due to increased investments in cross-border fulfillment solutions, partly offset by improved operating efficiency. Revenue from local service group was RMB16.2 billion, an increase of 12%, driven by the order growth of both Amap and Ele.me and revenue growth from marketing services. Adjusted EBITDA was a loss of RMB386 million compared to a loss of RMB2 billion in the same quarter last year, primarily due to improving operating efficiency and increasing scale. Revenue from our Digital Media & Entertainment Group was RMB5.6 billion, an increase of 4%. Adjusted EBITDA was a loss of RMB103 million compared to a profit of RMB63 million in the same quarter last year. Revenue from all other segments increased 3% to RMB47 billion, primarily due to the increase in revenue from Freshippo, Alibaba Health, and Intelligent Information Platform, partly offset by the decrease in revenue from Lingxi Games and Sun Art. Adjusted EBITDA from all other segments was a loss of RMB1.3 billion compared to a loss of RMB1.7 billion in the same quarter of 2023, primarily due to improved operating results from Sun Art, Freshippo, Alibaba Health, and Lingxi Games, partly offset by the increased investment in technology businesses. To wrap up, as Eddie and Jiang Fan mentioned, for our Taobao and Tmall Group, as order and GMV growth, we are advancing monetization efforts step by step. We expect CMR growth to gradually align with GMV growth over the coming quarters. For Alibaba Cloud, we are pursuing high-quality revenue and effectively executing our integrated cloud plus AI development strategy. We are confident that revenue from external customers will return to double-digit growth in the second half of the fiscal year with gradual acceleration thereafter. For AIDC, we focus on proactive investment to drive high-quality growth enhancing operating efficiency across all business lines. Our loss-making businesses are improving their monetization and operating efficiency. We expect most of these businesses to breakeven within one to two years and gradually contribute to the profitability scale.

Hi, everyone, for today's call, you are welcome to ask questions in Chinese or English. A third-party translator will provide consecutive interpretation for the Q&A session. Please note that the translation is for convenience purposes only. In case of any discrepancy, our management statement in their language will prevail. If you are unable to hear the Chinese translation, bilingual transcripts of this call will be available on our website within one week after the meeting. Operator, we can now connect to the speaker and SI conference lines, start the Q&A session. Thank you.

Thank you. We will now begin the question-and-answer session. Your first question is from Ronald Keung from Goldman Sachs.

And I want to ask a question about our take rates that with the high single-digit GMV growth and 1% CMR growth, it does imply a wider gap this quarter. But the management, you just mentioned about the narrowing gap that should come. So how should we think about the take rate trajectory, the time needed for CMR to align with GMV? And can you talk a bit on what are the actual measures that we're doing? Is it software service fee or ad tech upgrades? How would these new products translate to Taobao and Tmall's respective take rates into the next few quarters?

This is Eddie, and I'll take that first before handing over to Toby. In the Taobao and Tmall group, the priority at present very much has been on enhancing the user shopping experience so as to drive heightened purchase frequency and GMV. Now as market share stabilizes, which is what we certainly seen this quarter, we can start to accelerate the pace of our monetization efforts. But there are a few factors that you have to look at. First of all, the various new products we've launched, including live streaming and the RMB10 billion subsidy program have resulted in high user return and return to platforms and high user repurchase rates. But it's going to take a bit of time to roll out the new monetization approaches and products for them to gain traction. But that's certainly what we're looking at doing over the coming few quarters. The second thing to look at is our new advertising product, Quanzhantui, that we're introducing. Of course, in order for this to achieve solid performance, you need more than just that product in place. You need effective coordination and synergy across three different pieces. The first being user traffic on the Taobao platform. The second being having a sufficient number of advertisers taking part across a sufficient number of different sectors with sufficient coverage of different kinds of products. And then beyond that, it will take, say, six or more months to fully optimize the algorithms and also to optimize the user data. So, starting from the rollout of Quanzhantui back in April, it's likely to take something like 6 to 12 months of work to bring all of those different things together to achieve really solid progress and growth. So, in short, for those three different things to come together and for this all to really be working well, likely looking at something like 12 months from the initial launch date. And then to that, I'll hand over to Toby to also say a few words about technology service charge.

Yes, this is Toby. Just to add regarding the technology service charge that we started talking about charging as of September on Taobao and on Idle Fish at a rate of 0.6%, and that is to be calculated on the basis of actually received GMV, actually received by the merchants. We've come to this decision in consideration of both prevailing market practice and the opinion of our merchants, and feedback from our merchants. Of course, we will have special measures to take good care of our SME merchants, in particular, those whose actual annual GMV is below a certain threshold either have that refunded or find other ways to provide assistance for them to expand their CASA base and grow their GMV to that threshold. So, I would expect that with this starting as of September, you can expect to see this making a difference in the financials over, say, the next seven months of this present fiscal year.

I have a question on your AIDC business. How does management think about the overall competitive landscape now? And how will that evolve over time? What are the competitive advantages of Ali Express for both the merchants and consumers that will help you defend your market position in some of the important countries? And then a follow-up on that is with Lazada seeming to be turning profitable starting in July. Is that a timeframe that you target to achieve profitability for Ali Express as well?

Thank you. This is Jiang Fan. Well, the first thing I would say is that AIDC, unlike the other businesses within the Alibaba Group, comprises many different types of businesses. It's not a single business. We are present in many different local markets with local brands in those markets, with local teams operating them. Overall, AIDC is really a combination of both cross-border and local business. So, something that we're working on in that broader context of being in different markets with different local brands, local experience, local teams operating them is we're integrating and sharpening our advantage around product supply out of China. On the consumer side, I think certainly one advantage we have is very good consumer brands with high brand recognition, and we are working very hard to enhance the user experience on AE. In the past, AE was perhaps a low-efficiency platform that too often offered consumers a relatively poor user experience, but we've been working very hard on improving that. As for Lazada, yes, as you noted, we were EBITDA positive for the first time for a single month in July of this year. We'll continue the adjustments as we have there while preserving our market share, finding ways to enhance efficiency and increase our profitability. AE Choice is an important model that we've been investing in, and it's taken time to get to where it is. UE continues to optimize. In respect of AE Choice, over the next few quarters to come, we'll continue to make those efforts to drive higher efficiency and achieve high-quality growth.

My first question is about Stock Connect. Can management comment about our thoughts about the status of a primary listing and the triaging going into Stock Connect? And my second question is about our cloud business. Given that we are expecting the external cloud revenue back to double-digit growth in the second half of the fiscal year and accelerate going forward, I just want to get some color with regard to our AI revenue contribution. What's the goal that we are looking for in the long run coming from AI?

Okay. Thomas, thank you for your question. I will take the first one, and then Eddie will answer the second question. Regarding your question about the Stock Connect, I guess your question is about the primary conversion at our end. Currently, we are actively pursuing a Hong Kong primary listing. We will have an AGM on the 22nd of August. So, we have a proposal that's subject to the shareholders' approval. Then, we are aiming to complete the conversion process after we obtain approval from the shareholders and get it completed before the end of August. Then you join the Stock Connect, it was still subject to different exchange procedures. So that's the update on the status.

This is Eddie. I'll take your second question about cloud. There's very, very robust demand among our customers for AI and AI relevant products. If you look at the pipeline, you can see that this demand is still far from being satisfied. So, we think it's a very clear-cut trend that revenue from external customers will achieve double-digit growth in the second half of the fiscal year. In terms of the breakdown of that AI product revenue. Most of that growth will be driven by AI products. If you look at the industry as a whole, demand for CPU-based traditional cloud computing is relatively limited, where most of the growth is now focused on GPU-based AI product development. So, I would say something like more than half of that expected growth will be driven by AI products.

Just first of all, a quick follow-up on the prior cloud question. Obviously, we are already showing some early indicators of the positive momentum, and the management shared the strong demand. Given the current kind of macro positions and softened enterprise demand into the second half, just can you comment on kind of the mix impact, especially in the next several quarters? And then also, the second question is on the Taobao and Tmall GMV reacceleration. So, we're encouraged to see our overall GMV momentum catching up. And actually, for the second quarter, we already caught up to industry pace reported by MBS. Monitoring, can management comment on the key drivers behind the GMV reacceleration? And anything you can really talk on the return rate that we've seen across the ecosystem?

This is Eddie. I'll take the first question that has to do with cloud. I think your question had to do with how macroeconomic conditions are resulting in softening enterprise demand. But that's not at all what we're seeing. What we see certainly among our own cloud customers is that their AI budgets for this year are higher, significantly higher than what they were last year. Any enterprise that is digitalized and relies on digitalization must be investing in AI. All the industries now that are using AI are leveraging it to enhance their competitiveness and efficiency. So, any enterprise that is in any way digitally reliant will certainly not scale down its investment in AI because to do so would impair their own competitiveness and efficiency. So that's a clear trend that we're seeing, and we're not seeing any kind of slowdown or softening. As to your second question on the recovery in GMV growth rate at TTG. I think really this comes down to our overall strategy for the Taobao and Tmall group, where the most important thing we're investing in is the underlying capabilities that allow us to deliver quality products on quality services at attractive prices. Those are the most basic parts of the user experience when it comes to e-commerce. Now if you look at TTG, we have an extremely rich and diverse product assortment and also a very complex, multifaceted user base. So, it's important for us to segment those consumers to make available different kinds of product supply and different kinds of service experience to cater to that diverse range of needs of those different segments. That allows us to develop different kinds of supply capabilities for those different segments. Underlying all of that is our ongoing investment in optimizing supply chain and supply chain efficiency to ensure we have quality products and quality services at attractive prices for those different consumer tiers.

This is Toby, and I'll take that other question about return rates. Purchase return rates are increasing across the industry as a general trend. But in the case of Alibaba, our return rate is slightly lower than the industry average. The return experience is a very important part of customer experience overall. When the consumer has a good experience around returns, that's going to be favorable for increasing retention and purchase frequency. On our platform this year, if you look at our Net Promoter Score, NPS, especially among mid- and high-tier consumers, our NPS has improved considerably. Our return rate is slightly lower than the overall market average, and I don't think that's going to discourage merchants from investing and doing business on our platform.

My question is regarding management's comment about some of these loss-making businesses aiming to reach breakeven in the next one to two years. Could management share more detail in terms of individual segments including local services, AIDC as well as most of the retail business under the all-other segment? How should we look at the timing of kind of breakeven target in the next one to two years individually?

Thank you, Gary. I'll address that question. What Eddie and I were conveying is that aside from Taobao and Tmall, which are our core businesses, the other ventures, some of which are currently operating at a loss, are working on improving their efficiency and monetization. We expect that in the next one or two years, they will significantly reduce their losses and reach breakeven. After that, they should begin to generate scalable profitability. To achieve this, it is essential that these businesses enhance their efficiency and focus on monetization. It's crucial for them to improve their monetization rate through their business models. With increased revenue and better spending efficiency, they are on the path to breakeven and profitability. For instance, I gave two examples: local services and Lazada. In local services, we need to grow the scale and improve unit economics. For instance, in the case of earnest delivery, continuous improvement in unit economics is necessary. In Amap, particularly with the ride-hailing sector, enhancing both scale and efficiency will help in significantly reducing losses and moving toward profitability.

My question is about the disparity between CMR and GMV. It’s impressive that you achieved high single-digit growth for GMV this quarter, but the gap between the two seems to be high single-digit as well. Is that the peak we can expect moving forward, particularly after the recent launch of Transat and with the upcoming implementation of the technology service fee starting in September? I would appreciate any insights on this. Additionally, I think Toby mentioned that the technology fees amount to 0.6% of GMV, but some smaller merchants may have that waived. I'm curious if that figure reflects a net percentage of GMV and how you foresee the technology service fees impacting CMR. Any clarification would be helpful.

Okay. Thank you for the question. I'll take that. I think as previously, Eddie and I mentioned in my prepared remarks, the gap between the GMV, online G&A, and CMR reflects the drop in the take rate. Why the take rate dropped? Because currently, some of the new models with a very high growth in GMV currently still have relatively low monetization, which will be ramped up in the next few quarters. If you like, that's why there's still a gap. As we also said, in the next few quarters, CMR growth will be at the same pace as GMV, which means the take rate will start to get relatively stabilized. That's what we're expecting. Eddie explained why GMVs currently grow faster than the take rate. The reason is that by executing our strategy, we first need to invest in consumers, get more purchase frequency from consumers, then get the GMV growth. As soon as we start to observe GMV back to growth and then the market share stabilizing, we will gradually ramp up the monetization pace. A few things we're doing. Eddie explained transient, which is currently being rolled out. The penetration into traffic will take a number of quarters, so that will certainly improve the take rate, which means we accelerated the CMR growth. Secondly, with this software service fee charge, it will also contribute to the increase in the take rate. A few other measures as well. These are all the things we've been doing, and we started to do it because we see the growth back on the GMV, and we see the signs of stabilizing the market share. Your question about the 0.6 charge is on the completed transaction, completed GMV. So, it's charged on the completed GMV. However, we will refund back to some of the small and medium merchants if their annual GMV is below a certain threshold.

I have one question maybe on the financial number. One is on the tax rate. I noticed that the after-tax rate is quite high for the current quarter. Can you maybe explain what's the rationale behind? And how should we model the expense going forward? The other thing is about the decline, 56% decline in our free cash flow. I understand that it is related to our investment in AI cloud as well as some one-off expenses on free cash flow related to the planned reduction of direct sales business. Should we expect that to gradually normalize in the next one or two quarters, especially for the direct sales business part?

Thank you, Kenneth. I'll take this question. Your first question is about the tax rate. If you're using our P&L's tax divided by profit, you normally cannot reach a sort of like the real effective tax rate because we do have a lot of, if you like, permanent differences and one-off items that do not have tax implications. We are monitoring on an adjusted basis on the tax rate, which is relatively stable in this quarter compared with last quarter. So, in terms of your model, I guess, our team can sort of look into Tmall and sort of communicate offline. Your second question is about free cash flow. As I explained, there is quite a significant drop in free cash flow. One of the reasons is the significant increase in expenditure on the AI infrastructure investments. The other reason is related to working capital changes in relation to planned scale-downs. For these scale-downs, many of them, if you like, a majority of them are related to planned scale down in some of the direct sales businesses within Taobao and Tmall. The reason is we believe that in some of these business categories, the 1P format probably is not a more efficient format. So that's sort of practice. You'll also have some businesses that were not affected because we proactively made that decision, but some were affected by the overall market situation. For example, as I said, revenue dropped from businesses like Sanna and Intime, which are in the 1P format. The drop in their scale impacts working capital. Is this a permanent impact? To me, I think that this is relatively temporary. During the process of reducing scale, we will see outflows. But when the business size becomes stable, we will no longer see these types of outflows.

Great. My question is related to AI, a more big-picture question for you guys. Can you talk about maybe the implication of open-source large language models, how that impacts the adoption of AI in China? Also, how should we think about large language models as they transition into agents? How do we think the use cases may change for your merchants and customers?

Thank you. On your first question. In this recent wave of generative AI development, different companies have made different choices around whether to develop open or closed source models. Alibaba Cloud is itself a cloud service provider. The clear choice for us has been to develop open-source large language models. That's very helpful, of course, for developers. Developers want more control. They want more scope to optimize. We feel that providing open-source large language models to developers is advantageous for that. That fits entirely into our own strategy, as we are a cloud service provider. It's doubly advantageous. They can develop their offerings using our open-source models, and then they're very likely to continue to select Alibaba to deploy those offerings at scale. As for the idea of agents, the models are becoming larger, more multimodal, with more and more parameters becoming stronger and more capable. There will be a need for agents, I think, to deal with some of the more complex needs around inferencing, especially agents that can call on different underlying large models. I know lots of developers are working on it, but I think going into the future, this kind of development will become more automated.

Let's open the last question.

The last question is from Youssef Squali from Truist.

It's Youssef Squali from Truist. So just on the follow-up on the AI question. One, if I look at your CapEx, it's more than doubled year-on-year in Q1. Number one, is that a good run rate to work off of for the rest of the year? Number two, as you get the return on those investments, can you help us just understand your framework of reference? One of the key questions we often get here at least in the U.S. is the ROI associated with the AI and kind of the timeline or the time or framework on which you kind of expect these investments to start bearing fruit, and maybe give us examples of where you're seeing early green shoots?

Okay. I will take the first question, and Eddie will answer the second question. I think with respect to CapEx, just as you have observed from our reporting, it increased very significantly. The reason is that we see clearly very strong demand and the backlog. That's why we are investing. We would expect a similar level of investments in CapEx going forward in the next few quarters.

This is Eddie. To continue with the second part of this answer. Certainly, as Toby said, over the next several quarters, we expect to continue to invest in AI CapEx at that kind of pace. It's simply because we see a lot of demand, a lot of unmet demand from many clients. You look at the pipeline those clients have, and you know there will be ongoing demand. Whether it's for training or inference or API calls, when we make these kinds of CapEx investments, as soon as we get a server up, a server is instantly running at full capacity. There's that kind of demand. We can expect to see a very high ROI over these next quarters because we're building compute power to meet existing demand, and that new compute power coming online is getting instantly taken up and running at full capacity on day one.

Well, thank you, everyone, for joining. We will see you next quarter.

Thank you. That does conclude our conference for today. Thank you for participating. You may now disconnect.