Good morning and good evening, everyone. Welcome to Cango Inc.'s Second Quarter 2025 Earnings Conference Call. This call is also being broadcast live on the company's investor relations website and is being recorded. Joining us today are Mr. Paul Yu, Chief Executive Officer, and Mr. Michael Zhang, Chief Financial Officer of the company. Following management's prepared remarks, we will have a question-and-answer session. Before we begin, I would like to point out the safe harbor statement in the company's earnings release, which also applies to this conference call as management will be making forward-looking statements. With that said, I will now turn the call over to Mr. Paul Yu, CEO of Cango. Please go ahead, sir.

Thank you. Good afternoon, and thank you for joining Cango's Second Quarter 2025 Earnings Call. Today marks an important milestone as we report our first quarter following Cango's strategic transformation. This isn't just another quarterly update. It showcases our complete transformation into a leading Bitcoin mining company. In just 9 months, we have scaled to 50 exahash of computing power, placing us firmly among the world's top miners. As part of this transformation, we recently completed a governance and leadership restructuring, onboarding a senior management team with deep expertise across digital asset infrastructure, finance, and energy investments. This leadership team gives us the right mix of skills to hit the ground running and execute our next phase of growth. I'm optimistic about what we can achieve together as we enter a new chapter in the journey. Today, I will walk you through how our strategic execution has fundamentally repositioned us to lead the industry over the long term. Let me start with our financials, where we generated RMB 1 billion in total revenue in the second quarter of 2025, with Bitcoin mining contributing RMB 989.4 million of that amount. However, you can see we incurred a net loss, which reflects two accounting adjustments that temporarily mask our operational strength and should be built as the essential investment in our foundation for the future. First, our clean exit from China. We completed the $352 million divestiture of our legacy China asset in May, which resulted in a one-off loss from discontinued operations. Second, as a part of the acquisition of mining equipment last November, we purchased 18 exahash of mining capacity, so a share-based payment. By the time the equipment was delivered in June, our stock price had nearly doubled, triggering a noncash hedge accounting adjustment in accordance with applicable fair value accounting standards. These were strategic decisions we took into consideration to rapidly build competitive scale and sharpen our focus. The important story is what lies beneath. Excluding these one-off adjustments, the adjusted EBITDA for the quarter was RMB 710.1 million, clear evidence of the underlying strength of our Bitcoin mining business. Now let me move on to the progress we've made in this transformative quarter. We've already achieved one of the industry's largest scales at 50 exahash, representing approximately 6% of the global network's hashrate as of June 30, 2025. July's Bitcoin production reached 650.5 BTC, up 44% or approximately 200 BTC increase from June, primarily driven by the full deployment of the 50 exahash mining equipment since the end of June. In addition, in August, we strategically acquired a 50-megawatt mining site in Georgia, a move that will reduce power costs and enhance operational stability and lays the groundwork for future expansion. We maintain a fortress balance sheet with $118 million in cash and cash equivalents as of June 30, providing ample capital to fund our strategic expansion. Our asset-light strategy provides a distinct advantage. By acquiring Plug & Play mining rigs with minimal upfront capital, we are able to scale more quickly and cost-effectively than vertically integrated competitors. Although this model resulted in higher cash costs per BTC of $83,091 during the quarter, our all-in costs remain competitive at $98,636 per BTC. This is primarily due to significantly reduced depreciation expenses from low equipment acquisition costs, which offset elevated power expenses. As a result, our capital efficiency supports a solid return on capital employed and ensures resilience across market cycles without the burden of heavy equipment financing. Additionally, our geographically diversified footprint across North and South America, the Middle East and Africa helps mitigate regional risks while sustaining industry-leading efficiency. Our roadmap forward is clear and purposeful. In the near term, we will maximize value from our 50 exahash of mining capacity by implementing efficiency upgrades and replicating the low-cost operational model of our Georgia site. Looking to the midterm, we plan to pilot renewable energy storage projects aimed at achieving near-zero-cost mining operations while simultaneously retrofitting select facilities to support HPC applications. Over the long term, we are ultimately building a dynamic computing platform that intelligently balances Bitcoin mining and AI workloads, all powered by our expanding energy expertise. This quarter's results reflect a company making bold and strategic moves. We have accepted temporary accounting adjustments to secure lasting competitive advantages, namely meaningful scale, cost-effective infrastructure, and a focused commitment to pure-play high-value computing. With this strong foundation firmly in place and a clear path to continued value, I have never been more confident in Cango's future. Before I turn the call to Michael Zhang, our CFO, to take you through our financial results for the quarter in more detail, let me quickly review our legacy business. We remain focused on lean asset-light operations for our used car export platform, AutoCango. Since its launch, our platform has attracted over 6 million visits and surpassed 456,000 registered users. It now hosts more than 800,000 vehicle listings with 70,000 different models on offer, connecting China's used car market with international buyers seeking quality inventory. We continue to see steady growth opportunities in this segment in the future. With that, I will turn the call to Michael.

Thanks, Paul. Hello, everyone, and welcome to our second quarter 2025 earnings call. Before I start to review our financials, please note that unless otherwise stated, all numbers are in RMB terms. Total revenues in the second quarter of 2025 were RMB 1 billion. Revenue from Bitcoin mining business in the second quarter 2025 was RMB 989.4 million with a total of 1,404.4 bitcoins mined in the second quarter of 2025. The average cost from mined Bitcoins excluding the depreciation of mining machines was USD 83,091 per coin with all-in costs at USD 98,636 per coin during the quarter. Revenue from automobile trading income was RMB 12.4 million in the second quarter of 2025. Now let's move on to our costs and expenses during the quarter. Cost of revenue exclusive of depreciation and amortization in the second quarter of 2025 was RMB 836.9 million, depreciation and amortization in the second quarter of 2025 was RMB 156.4 million, general and administrative expenses in the second quarter were RMB 21.7 million. Due to one-off loss from discontinued operations and noncash impairment loss, we recorded an operating loss of RMB 1.3 billion and a net loss of RMB 2.1 billion in the second quarter of 2025, respectively. Excluding the impairment loss and one-off loss from discontinued operations, we recorded adjusted EBITDA of RMB 710.1 million in the second quarter 2025 compared with RMB 5.4 million in the same period of last year. Moving on to our balance sheet. As of June 30, 2025, we had cash and cash equivalents of RMB 843.8 million. Starting with our second quarter 2025 results, we intend to change the reporting currency of our consolidated financial statement from RMB to U.S. dollars, reflecting the profile of our revenue and profit after divestiture of our China asset in May 2025. The change is expected to be effective from the company's results for the third quarter of 2025, which will be reported in U.S. dollars. This concludes our prepared remarks. Operator, we are now ready to take questions.

And today's first question comes from Emerson Zhao with Goldman Sachs.

This is Emerson from Goldman Sachs. I have 2 questions. Number one, could you outline your roadmap for computing power over the next 12 months as well as the capital expenditure plans? For example, will you continue to acquire mining sites or order new miners? Number two, we understand the previous announcements mentioned your strategic direction of green energy plus storage. Could you update us on when new progress is expected?

Thank you, Emerson, for attending the conference call. I will take the first question, and our CEO, Paul, will take the second one. For the first question, for computing power, our goal for the second half of the year is to fully unlock the value of the current 50 exahash computing power. This will be achieved by improving operational efficiency, upgrading machines and selectively acquiring mining sites with low electricity costs. One example of such a site is our newly acquired mining facility in the state of Georgia. Of course, if suitable opportunities come up, we will be open to expanding computing power through M&A as well. As for capital allocation, we will continue to maintain strict capital expenditure discipline. We are also evaluating opportunities in areas, including computing power expansion, green energy storage, AI HPC centers, and more for long-term growth. For the second half of the year, in particular, the focus will be on selective mining site acquisition. Specifically, we are looking at sites that significantly reduce electricity costs, enhance energy security, and support the stability of our overall operations. More importantly, by operating this infrastructure, we will get critical hands-on management expertise. We believe this will provide a solid foundation and strategic flexibility for future business expansion into energy plus HPC sector. Thank you.

Regarding the green energy plus storage, it is one of our most important strategic objectives. We are advancing through two parallel paths. First, we are actively seeking M&A targets globally for rapid deployment. And second, we will develop critical hands-on management experience by investing in pilot projects that are developed with experienced partners. Thank you.

And our next question today comes from Pingyue Wu with Citic Securities.

This is Pingyue from Citic. And I have also 2 questions. The first is the company previously mentioned to follow a light asset model, and we also acquired mining sites. So does it mean that the company is gradually shifting towards integrated operation? And my second question is, is there a plan to acquire more low electricity cost mining sites in the next phases? Are low-cost regions, such as Latin America and the Middle East prioritized? What are the screening criteria?

Thank you for your question. Acquiring those mining sites isn't solely about reducing costs. Beyond cost reduction, there are three strategic benefits. First, a stable energy supply supporting our existing large-scale computing power requires reliable and consistent energy sources; second, infrastructure and operational expertise. We gain critical operational experience that lays a solid foundation for upgrading our capacities. Third, creating a foundation for strategic transformation. By securing low-cost power and scalable sites, we are building the infrastructure needed for future transformation into AI data centers. With these benefits in mind, our criteria for selecting mining sites also fall under three pillars: low-cost electricity to maintain business competitiveness, sufficient capacity and power redundancy to support future upgrading, and stronger regional grid stability to handle high-load demand after transformation. Our asset-light strategy avoids the risk of CapEx, heavy mining models, and by managing energy sources and billable sites, we have achieved due strategic positioning for both current business support and future AI transformation. In our view, selectively acquiring mining sites doesn't change our capital-light strategy in the mining business. Instead, it enhances our operational efficiency and gradually creates a pathway towards green energy and AI computing centers. Regarding the plan to acquire more mining facilities, in general, we will continue to monitor M&A opportunities for energy projects in line with the company's strategic transformation needs. In mining site selection, we conduct in-depth evaluations based on actual operational performance, detailed cost-benefit calculations, and potential strategic synergies. Regionally, the U.S. is our current priority. Most of our miners and hosting sites are already located there, and we want to strengthen our local operations and marketing position. The Middle East, a region with attractive energy prices and policy stability, is also within our scope of consideration. For specific targets, we will focus on factors such as local climate, energy prices, and policy stability. We tend to prioritize acquiring sites where we already have long-term stable collaborative relationships, good operating conditions, and sustained low power cost advantages. Thank you.

Our next question today comes from Joey Chee with Securities.

This is Joey from Securities. I have two questions. First, we've observed that leading mining companies like Mara and CleanSpark have reached 50 EH. How does Cango plan to maintain its market share in computing power amid a shortage of miners or challenges in energy efficiency? Second, when promoting infrastructure investments in the U.S., are you facing any restrictive policy risks?

Thank you for your question. I will take the first one. As one of the largest miners with the current scale of 50 exahash per second, maintaining competitiveness isn't just about increasing the total volume of computation power. The critical factor, I think, is the deep optimization of computing power efficiency. We have established a unique asset-light model, focusing on the strategic acquisition of secondhand on-rack miners to achieve rapid low-cost mining capacity expansion. This strategy will continue to be a core advantage. Meanwhile, we will not cling to inefficient capacity. We closely monitor miner performance and economic metrics. We dynamically phase out inefficient capacity and upgrade to more energy-efficient miners. Miner supply shortage will not be a bottleneck. I think we've got the expertise and the relationships to acquire cost-effective power to grow our business and drive our growth. Furthermore, we remain open to computing power M&A opportunities that align with our strategies. Thank you.

Regarding U.S. infrastructure investment, we continuously monitor changes in the policy environment. We have developed a local compliance team and legal advisers to evaluate potential restrictive policies and mitigate related risks. Currently, most computing power-friendly states in the U.S. have no restrictive policies on power access or land use for data centers. In fact, they have introduced computing power infrastructure investment services. Only a few states have strict approval process requirements for converting critical computing power to general-purpose computing power. We have established regular communications with local energy regulators to ensure compliance and efficiency during project implementation. Thank you.

And our next question today comes from William Gregozeski with Greenridge Global.

Congratulations on all the progress you guys have made over the last 10 months on this change. I also have 2 questions. The cost per BTC has increased, and you mentioned you have the new mining equipment, the Georgia acquisition, and then you're making optimizations. Is there a number we should be looking at for what you think the cost is going to be as you exit the year on that with your current exahash portfolio? And then the second question is, given how undervalued the stock is, do you plan to do any repurchases or just prioritize the cash for operational expansion?

Thank you, William. I think I will take most of your questions. For the first one, as we just disclosed in our second quarter results, our cash cost per BTC was approximately $83,000 with an all-in cost of around $98,000. The deployment of our new equipment in July, bringing our operational hash rate to 50 exahash, is a key step towards improving both metrics. The increased scale and efficiency of these new fleets are expected to increase our absolute Bitcoin production and improve our cost profile on a per-coin basis. However, we are also seeing continuous upward pressure on mining costs across the sector, driven primarily by the rapid increase in global network hash rate. Therefore, while our new risks will gain us further economies of scale, we anticipate these industry-wide headwinds will also be reflected in our cost structure during the third quarter. And as for your second question, we absolutely agree that our current stock price doesn't reflect the intrinsic value of our business. We firmly believe that the most sustainable way to achieve a fair value is through the continued development of our core business and enhancement of our profitability. Therefore, our primary focus is to strategically deploy our cash and liquidity to fund high-return operational expansion and our business transformation. This includes investing in new high-value areas such as AI data centers, which we believe will drive future growth and profitability. Simultaneously, we are deeply committed to delivering returns to our investors. We will continue to take a balanced approach, carefully evaluating our business development needs along with capital market conditions to manage our liquidity prudently. This means we will consider all tools at our disposal, including our ongoing share repurchase program, to ensure we are creating long-term value for our shareholders. Thank you.

And our next question today comes from Kevin Dede at H.C. Wainright.

Paul and Michael, you've addressed this question a number of times, but it's still a little unclear to me. The August hash rate was almost 44 exahash of the 50 exahash that you have deployed. I don't understand how improvements and optimization will get you that next 6 exahash. Could you maybe explain that? And how should we look at that going forward?

Thank you, Kevin, for your questions. I think first, we'd like to highlight that the near 44 exahash disclosed in our August production report represents our effective operational hash rate, although it's reached about 87% effective rate but it still has a difference on the gap compared to the top miner, which is above 90%. So we think we still have space to improve our efficiency. I think that's our main way to improve our efficiency is to maybe upgrade our inefficient miners and also to develop our operational team. I think we can further improve the efficiency of the miners. Paul, do you have anything to add?

Yes. I think it takes time to implement our rigs on track after acquiring all the mining machines. And also, there are a lot of curtailments in the U.S. during the summertime. I think we will improve our operational efficiency in the future. Thank you.

Thank you, Paul. I had suspected that curtailment was certainly a factor. Okay. Just from a high level, what do you think is the most important aspect of Cango's June report? What would you really like to have resonate for investors?

This quarter was truly a milestone for us. First, we completed the divestiture of our China operations and the acquisition of 18 exahash in assets. We have finalized the change in ownership and established a new management structure and team. We are now fully prepared and ready to move forward. From a business perspective, having 50 exahash means we have officially entered the first tier of industry players. At the same time, our Bitcoin holdings continue to grow, now exceeding 5,000 BTC, and this has solidified our mine-and-hold strategy. On the financial side, it's especially worth highlighting that while maintaining scale and validating the effectiveness of our business model drove solid financials, we also see continued opportunities for cost optimization, which will remain a focus in our next phase.

Perfect, Paul. It was a very comprehensive review, and I appreciate you discussing it with me. Congratulations on all the progress you've made.

Thank you.

And that concludes the question-and-answer session. Thank you once again for joining Cango's Second Quarter 2025 Earnings Conference Call today. Have a great day.