SOBKY (SOBKY) Q2 2027 Earnings Report, Transcript and Summary

Thank you very much for your waiting. We will now begin SoftBank Corporation's financial results briefing for the second quarter for fiscal year ending March 31, 2027. We would like to introduce today's speakers. President and CEO, Miyakawa; Senior Vice President and CFO, Head of Finance Unit, Akiyama. Today's briefing is also being streamed live over the Internet. Without further ado, I would now like to invite President and CEO, Miyakawa, to present an overview of SoftBank Corporation's consolidated financial results and business performance.

I'm Miyakawa of SoftBank. First of all, I would like to offer my sincere apologies. On October 7, unauthorized access caused a service disruption at our subsidiary, IDC Frontier. We deeply regret the concern and inconvenience caused to our customers and all other stakeholders. IDC Frontier has 4 regions. In one of them, East Japan Region 1, a vulnerability in open source software used for authentication was exploited. Then in 4 of the 6 zones, some customer servers were shut down and encrypted by a third-party. The incident affected 495 corporate and local government customers. We conducted a thorough investigation. We have confirmed that there was no exfiltration of customer server data. You may ask, had we applied the patches for applications developed in-house, we had conducted assessments and completed patching. However, we have not assessed the open source software whose vulnerability was exploited. This is deeply regrettable. We take this failure very seriously. To prevent this from happening again, we conducted thorough inspections across our group companies. We urgently checked the systems designed to prevent unauthorized external access as well as the software in use. Attacks exploiting vulnerabilities in open source software, including PHP and Java are increasing for the past month. So we have urgently rechecked possible entry points and implemented countermeasures. New AI models are being released on a weekly basis and new vulnerabilities are continually being discovered. We'll take every possible measure to address these threats. Now let me turn to our financial results for the second quarter. Revenue reached JPY 3,798.7 billion, up 12% year-on-year. Progress towards our full year forecast is 51%. Looking at the breakdown, revenue increased in all segments. 3 businesses: enterprise, distribution and financial, achieved double-digit revenue growth. Operating income was JPY 642.1 billion, up 2% year-on-year. Progress towards full year forecast was 58%, indicating steady performance. Last year, we recorded a remeasurement gain of nearly JPY 60 billion. Despite that high comparison base, we achieved operating income growth in the first half. Excluding onetime factors, operating income grew by 13%. By segment, operating income increased in 3 businesses: enterprise, distribution and financial. Consumer operating income declined 2%. However, from the beginning of the fiscal year, we had anticipated a heavy first half burden from the amortization of customer acquisition costs incurred in prior periods. Performance is better than we expected. We will keep taking steps for sustainable growth. Net income reached JPY 364.5 billion, up 5% year-on-year, progress was 65%. Here is a summary of our progress towards the full year forecast. Revenue, operating income and net income all reached record high. Looking at individual segments, all are progressing steadily against their full year forecast. In light of this strong progress, we have decided to revise our full year forecast upward. Reflecting business performance, we are raising our full year forecast for revenue to JPY 7.9 trillion, for operating income to JPY 1.16 trillion and for net income to JPY 610 billion. All these figures represent double-digit growth compared with the previous fiscal year. We have set a net income milestone of JPY 600 billion for FY '28, but we now expect to reach that milestone 2 years ahead of schedule. Next year, we will update the forecast for the next fiscal year and onward. Let me now review performance by business segment. First, the enterprise business. Revenue was JPY 538.9 billion, up 12% year-on-year. Within that, cloud and AI revenue increased by 37%. We identified this area as a growth driver in our midterm management plan, and we are now seeing that growth in the numbers. AI is driving our growth. Operating income was JPY 124.8 billion, was up 20% year-on-year. Next, an update on Patching as a Service or PaaS. By the end of October, the number of companies signed up had exceeded 7,000. Almost every day, we hear reports of unauthorized access and data breaches. We are working at full speed to find vulnerabilities in our customer systems and fix them as quickly as possible. At the same time, attacks powered by advanced AI are evolving every day. Each time a new AI model is released, previously unknown vulnerabilities can be discovered and exploited. As attackers become more sophisticated with AI, companies must also use AI to strengthen their defenses. Otherwise, they simply cannot keep up. We spent most of the time on addressing those cybersecurity issues. Again, we, as a defense team needs to make sure that we have AI as a tool to address attackers. We must repeat the cycle frequently; step 1, vulnerability assessment; step 2, remediation patch deployment; and step 3, verification. This will become extremely important. We have not yet fully automated the verification process in step 3. New AI models are now being released every few weeks. We are now at the point where the cycle needs to be repeated every week. We will continue development towards full automation. Next, let me update you on our Hokkaido Tomakomai AI data center. As you can see, preparations are progressing steadily toward opening by the end of this fiscal year. We want to make sure that this facility gets up and running successfully, then we will use it to drive further growth in cloud and AI. Next, the Consumer business. Revenue was JPY 1,602.5 billion, up 9% year-on-year. Mobile revenue increased by JPY 25.5 billion. Operating income in the Consumer business was JPY 322.9 billion, down 2% year-on-year. However, it is steady, and the progress rate is 58%, and we remain on track to achieve full year operating income growth. Now I would like to touch upon mobile ARPU. So in the second quarter, mobile ARPU increased by JPY 230 year-on-year, a solid result. We will continue enhancing our network and improving our services. Our smartphone churn rate improved year-on-year in the second quarter. Since the second half of the last fiscal year, we have changed our customer acquisition policy. Rather than chasing net additions, we focus on customers who will stay with us over the long term. This change in policy has contributed to the improvement in churn. I believe it was an important management decision. We will continue to prioritize profitability when acquiring customers, and we will pursue sustainable growth. Next, the Media & EC business. Revenue was JPY 894.7 billion, up 9% year-on-year. Revenue in the Media business also increased by JPY 8.9 billion. Operating income in Media & EC was JPY 126.5 billion. Excluding last year's onetime factors, this represents growth of 16%. Next, the Finance business. Revenue was JPY 241.9 billion, up 27% year-on-year. We are seeing strong growth led by PayPay. Operating income in the Financial business was JPY 67 billion, up 74% year-on-year. Now I have some good news about HAPS. Our trial services were successful in August. On August 9, the aircraft departed from New Mexico in the United States, it crossed at the Pacific Ocean. And on August 23 and 24, we conducted a trial service off the coast of Kochi Prefecture. The flight lasted 30 days, covering approximately 30,000 kilometers. We have video footage of the moment that LTA or Lighter Than Air, aircraft lifted off. [Presentation]

Here's an overview of the HAPS trial test. First, we tested communications performance. We confirmed quality equivalent to territorial networks in Japan. We also succeeded in indoor communications and local government emergency alert tests. These can be difficult for satellites. As for emergency call, as the tests were conducted at sea, we tested only that Japan Coast Guard's emergency number 118. We verified that the calls worked properly. Second, HAPS-to-drone communications and remote flight control from HAPS and high-resolution video transmission were tested and done successfully. We tested AI-RAN functions on board HAPS. Using an onboard edge processing server, we confirm that processing latency was reduced by more than 40%. This was a world first. In optical wireless communications, we successfully demonstrated laser ranging and tracking. This was also a world's first. We confirm that optical wireless communications technology used in space can be applied to HAPS. We see these successful trial services as a major step toward commercialization next year. Last but not the least, let me discuss our Beyond Japan initiative. We have decided to exhibit at Mobile World Congress, Barcelona this coming March for the first time. We want to take this opportunity to introduce new business opportunities to telecommunication carriers around the world. We will have the largest booth among telecom carriers. We have been invited many times before, but we kept saying no because we were not ready yet. But finally, we have products and organizations are in place and fully ready. So we have decided to exhibit. Please come and visit our booth and see where we try to go. Let me summarize today's presentation. First, revenue and profit increased in the first half with revenue growth in all segments. Second, cloud and AI drove growth and enterprise operating income increased by 20%. Third, reflecting steady business performance, we revised our full year forecast upwards. And fourth, I announced that SoftBank will exhibit at MWC in March for the first time. Under the theme Activate AI, we hope to show how our businesses are steadily expanding. Once again, if you have a time, please come and see us. We welcome you warmly. That concludes my presentation. Thank you very much. Now we will move on to the Q&A session.

We'll first take questions from participants here at the venue, followed by members of the media joining via Zoom and analysts and institutional investors joining via Zoom. [Operator Instructions] Thank you for waiting. [Operator Instructions]

I'm Yagi from Nippon Keizai Shimbun. Two questions. First is about cyberattack. So for the past month, it's been increasing. Can you tell us more specifically how much is increasing? And within the SoftBank world, in 2020 -- fiscal year '25, the number of attack was about 60,000 average per month and a total like 300 million. So what is the increased rate in October when you were attacked? And the second question is about you're conducting emergency inspections -- and at the SoftBank World, and so you have 1,800 systems under the system running and 700, the source code developed by in-house and also open source. And you also mentioned about open source as well.

So regarding the number of attacks. So last year was about 300 million on average last year and monthly, it happens about that many times per month. So it will roughly double. And from August to September, October per month, about 100 million each month, roughly at this pace. So it's not decreasing, but rather increasing. So I think this will further increase, I think. And whether the patching has been applied or not, regarding the vulnerability assessment, as you said that we are expanding to the open source code as well -- open source as well. And so 1,800 or even more. Urgently, we are conducting vulnerability assessment. It's a vast volume. I could not really say that we have completed all, but we are working on a daily basis.

Additionally, so you mentioned that the number of attacks is increasing. So AI involvement is also one factor, it's said. So can you tell me your opinion why it's increasing?

Well, the attack using AI is increasing and also the speed of attacks is also increasing. So therefore, it is getting more difficult to handle by human hands. So announced regarding IDCF. So it was only 7 minutes for attack. So the crime time is only 7 minutes up until now, and alert sounds and our checking team assessed and then to take countermeasures that will determine the appropriate course of actions. So there are some decisions that were made in someone else's meeting. So we no longer have time to do such things. So we have also switched our operation where we respond accordingly. And we first blocked the connection. So that's the operation now. And in particular, the number suddenly started to increase. It seems like Japan is being targeted a lot. And also the U.S., Germany, Canada, those countries are being attacked. And it's been said that security is very weak in Japan. So Japan Security has been talked about for a long time about its weakness.

Any other questions on the floor?

Kimura from NHK. I have 2 questions. First, about IDC Frontier cyberattack. Looking at disclosed information, it looks like there were some customers whose data was not restored. So I wonder if you have any idea about compensation for those customers whose data was not restored. And next question about patching open source software and immediate action you took we understand. But going forward, from a long, midterm perspective, do you have -- do you think that your security policy will be changed in the long run?

Again, about IDC cyberattack, we understand we caused a lot of trouble to customers, and we take it very seriously. Softbank employees are helping customers to address these issues together with those customers. In accordance with the contracted terms, we want to sincerely help customers from a compensation perspective. And to your second question, things like 5.5-Cyber or Mythos are specifically designed for cyberattacks. And those models became available in May or June. And we thought that in 3 months or so, open source similar model will become available. And it happened on 15th of July. That was started by Kimi followed by Qwen and a lot of different open weight software became available. And every day, I don't know whether attackers are using those AI tools or not. But when open web model becomes available, generally speaking, I feel that cyberattacks became more and more prevalent. The cycle of new model release is shorter and shorter. And I believe on a weekly basis. That kind of model becomes available. Then we test with the latest release model, but almost every time we test, we discover vulnerabilities, new vulnerabilities. It's kind of embarrassing to say on 15th of June, we did vulnerability assessment of our systems, and we discovered over 10,000 vulnerabilities. And we organized the press release together with Masa because we thought that it will be very, very significant issue pretty soon. And for the next 3 months, we applied patches to all over 50,000 -- sorry, 10,500 vulnerabilities. And Astra model became available and then we tested against the model. and we discovered 35,000 vulnerabilities, new vulnerabilities. That how powerful the new model is. So we are to be very honest with you, very exhausted. But yet, we are applying patching almost every day. And whenever a new model comes, we will test once again. So to answer to your question, about policy in the long run. Again, system is designed and built by humans to use automating driving as an example, there is a transition period from human-driven car to fully automated cars. There is a transition period when risk is very high. Frontier AI has pros and cons. Frontier AI should help enriching customers' lives, but cyberattacks will be considered as cons of Frontier AI. And if every system is modernized and to be able to address all cyberattack threats. We have not reached that point yet. So again, this is a transition period, and we need to make sure that we address immediate threats every time we discover new threats and going forward.

Next question from Asahi Newspapers.

I have a question regarding cybersecurity. So related to GMO access and they requested neutralization measures to the cyber threat. If that was been applied and actually it has been done or the blocking was actually conducted, would you please give us any comment? And also last week, on the 6th, so there was an announcement that including the blocking, so there will be some measurement will be taken.

So I cannot give any comments related to the response with the government. So whether blocking is good or bad. So one of the ways to protect individual right, blocking is one way to do. And to the second question, so we are doing our business handling and communications, confidentiality of communication in our business. So since it involves the secrecy of communications, it is important for the standards to be clearly defined by law. So what about the question regarding whether the blocking has been done or not? I would like to refrain from answering to that question. Thank you.

Any other questions?

Yamamoto from Nikkei. About cyberattack, and also, I have a question about consumer business. First, about cyberattack. You mentioned the latest model of OpenAI, Astra. So I wonder do you have a plan to deploy Astra in Japan. And PaaS, you mentioned that you automate up until the point of assessment. So when do you think it's going to be completely automated? And again, the time line to reach that level.

About the GPT-6-Cyber, OpenAI's latest model, which was released pretty recently, and we applied for usage, and we are preparing for testing. So that model, we have not had an access to the model yet. I don't know if I answered to your question, but let me try. Again, to make sure I understand correctly, what's your question?

So again, when can you fully automate the process of Patching as a Service?

Again, we automated patching, but whether the system can be restored and work well, after being patched. There are a lot of systems integrated each other. And whether all systems work after patching being applied, we have humans to check. So that's kind of bottleneck that we are facing at the moment. If system is working stand-alone, it can be up and running immediately after patching, but a lot of big enterprises have a lot of systems integrated each other. So we are trying to automate the verification process as soon as possible, but hopefully, by the end of the fiscal year, we want to complete the automation. But we are not ready to share with the exact date.

And the second, consumer business, the price change, you mentioned that you are seeing increasing ARPU, while churn declining. So pricing revision, which you did, how do you feel? Are you happy with the price revision?

Yes, it has been going well as expected. By that what I meant is churn is decreasing as we expected. And we started off as a net loss, and we were concerned. But since August, we have been seeing upward trade -- trend, excuse me. So we are glad that we did revise the pricing. So you are mentioning that churn is -- or net loss is declining. I don't know whether price change directly impacted or not or whether customer took it, received well. So we believe that customers received our price change positively.

Next question.

I'm Sekiguchi from Keitai Watch. A question about consumer business. I have 2 questions from here. The first one is about smartphone sales. So the first half year, how the sales on iPhones and the new model of iPhones started selling in October and so which is a popular smartphone every year. And so since October, this iPhone with dual features has been released. So please tell me first how these factors affecting customer behavior?

Yes. So this year, the iPhone sales is a bit different from the previous years. And we started with the high-end model sales. And especially this year, well, customers seeking high-end products haven't really changed anyway the number sold is actually higher than last year.

So is there any story that SoftBank's tactic went well?

Well, nothing special. I think it's just a trend of the market as well.

And another question is also related to the consumer business. At the end of September, Rakuten Mobile's roaming contract ended with KDDI, which means that the Rakuten's area became narrower and network quality has a certain level of impact. So in October onwards, is there any impact on the enterprise business or consumer?

I'm not sure if that is the impact of this roaming contract effects, but MNP has been on a declining trend since the first half of the year. So we have been implementing various policies for quite some time. So from Rakuten users, there have been some who switched carriers. We are not running any special campaign. So I don't think there will be a significant impact.

MJ from Bloomberg. I have a question about our business in the U.S. SB Neo, that is for a long term, I believe that you want to launch a new cloud business with 10 gig in the end. But sentiment about AI infrastructure is being worsened recently. So is there any change about your strategy or business plan?

Well, SB Neo is just a preparatory company. SB Neo is to build data center with a GPU chip to deliver cloud services. So unless there are enough number of data centers available, that kind of business won't work. So in the states, the recent move in the states may be a negative in the short term, in general. So if -- it will be very, very difficult to run that kind of business in the States, maybe we look for different countries.

Do you mean France?

Well, not specific country, any other countries that have opportunities.

And next question.

Tajima from Toyokasei. My question first is about the business environment change due to AI. So cybersecurity is one and also in the U.S., AI development, speed deterioration as well as the result of the midterm election. So also the AI data center rules. So what is your opinion as of the situation right now? And also regarding the business side, the domestic AI cloud business is growing steadily. And SB Neo, you mentioned, so this AI business, is there any blocker you may think of? And another question is about TEPCO Holdings and alliance with them. So in order to proceed with the AI business in Japan, so securing the power is very critical. And I think that is also the concern you had? And Mr. Son at the SBG's General Shareholders meeting in June, so the alliance with the TEPCO Holdings, he is one of the candidate he mentioned. So what is the situation right now?

Well, regarding the AI business expansion and deployment, as I also mentioned about the bright side and the dark side. So from where -- what perspective you take. So it depends on the perspective you take and thinking ahead 10 years, AI will eventually hit some limits. So I don't think it's something like something for specialists. I don't think so. So under the certain level of regulations, how it should be used and this kind of discussion will be actively made onwards. But once AI development stops, and -- but now the foundation of AI is done, and this going to continue and proceed or evolve even further, data centers or semiconductors and all these industries, I don't think that these businesses will not -- I don't think it will not decline. So there is that intention, I understand. We'll go through cycles of setting down and then becoming active again, repeatedly. That's how I think about it. And our Hokkaido Tomakomai data center is going to use green energy and starting with 300-megawatt data center and up to 1 gigawatt. And so this is how I think. So we have already secured power for this, and we are confident that we can fully accomplish this. And also in Kyushu, there are various locations currently under discussions. And Tomakomai, and Sakai, both in the industrial area. So data center is a factory. So factories and residential areas need to coexist in the community. So how we could coexist, and we need to -- based on perspectives obtained in each region, focusing on locations, we believe are suitable. So since we do have NDA about the second question, so we are not in a position to release any information.

Next question is from Sano, freelance journalist.

Because of more and more expensive memory prices that could have an impact on smartphone. In other words, users may find it difficult to buy smartphone because it's too expensive. So under the circumstances, do you have any specific plan or initiative to address that initiative -- sorry, issue or concern?

To be very honest with you, the latest models of smartphones in general cost JPY 300,000, JPY 500,000, it's too expensive. It's more like a computer, not just a phone. In order to promote sales we have -- a lot of carriers have taken initiatives. To be specific, high-end devices, there are some early entrants looking for high-end devices. So the number-wise, high-end device is selling well, but is it sustainable or not? I suspect there will be a limit to some point, not only expensive memory prices, but also there are a lot of capabilities now embedded in devices. So that could lead to high price range. So we need to make sure that we have a full lineup, high, mid-, low-end devices. I think that's an approach we should take. Not too much focus on high-end device, is our answer to your question.

Do you follow up on that?

What about mid-, low-end devices that kind of more impact from more expensive chip prices, for example?

I cannot say we are not going to raise prices for mid-, low-end devices. But we try to make lower-end devices as affordable as possible, but we don't expect memory prices will go down soon. So next year, we might consider raising prices for mid-, low-end devices.

And next question from the venue.

I'm Koyama, freelancer. I have a question about cybersecurity. IDC Frontier, so have you conducted the vulnerability assessment in advance? Even so, it was attacked or subsidiary did not have that assessment yet. So how do you think about countermeasures against the information leakage through the subsidiaries or subcontractors, companies? So -- and it seems that companies using your Patching as a Service have been able to prevent them quite effectively. Please share some cases.

So to answer your question, so whether the patching had been applied advance or not? Well, 700 companies had own source codes. For those, it was after applying patches. So including the subsidiaries and they had conducted. However, the actually the ones attacked was the open source. So there is a vulnerability on the open source was not well aware on the site. So we had a plan to conduct assessment on the open source, but it was not our priority back then. So we had to regret and then we need to admit that our response was somewhat delayed. So regarding the subsidiaries, whether PaaS were applied or not. The major ones, LINE Yahoo. So they have a lot of Internet-related things. So they have been applying patches. So as I mentioned earlier, on a weekly base, new models are emerging. So even we apply the patches with the previous model and still, there is a hole to -- for a new model to attack. So every week when new models emerge, then we have to also keep taking countermeasures.

The second question is about security again. So the other day about DOCOMO, so it was not a vulnerability issues, but the system linkage or authentication-related matter. So -- but other than vulnerability and patches, was there any failure in the system linkage? Or were you able to verify in that depth? Or it was just the vulnerability assessment?

So there are 2 ways to assess vulnerabilities. One is the -- with the source code itself is checked by AI. And -- and that's a source code is needed. Another one is that conduct a hacking test using AI. And so we call that black box test. And so from the entering point to how far it can go. And so that is also a system-related checks.

Ishino, freelance journalist. About churn, I have a question. You mentioned that you want to focus on long-term users or users will stay with you for a long time. How do you -- how can you identify those customers that are likely to stay with you for a long time?

So the attributes of those customers, well, when customer sign up for our devices, we offer different products like optical wireless. And those who signed up those kind of products, they are likely to stay with it for a long time. So what we would do is to ask whether customers are interested in other products like electricity or optical fiber wireless. If they are willing to sign up, we can say those are long-term users. You never know until they stay for long. But that's how we try to identify potential long-term users.

The next question is about the revised purchase program or revising the residual value of the device price that is being discussed at MIC or Ministry level.

From my perspective, from SoftBank's perspective, I think it's supposed to be like that. By that what I mean is if you buy a car branded Crown. The dealer A and dealer B may have different prices for the same car, but with residual price of devices can be a factor for lowering price competition, if you will. So I think a healthy discussion is underway on that front at Ministry level.

Next question from the venue.

Hori, writer. Back in October, October 19, My SoftBank stopped temporarily on the web -- for the web version, but the app was available and most process was made on app. However, some -- there are several menus that can only be accessed on the web version actually. So in that case, users -- they had to go to the SoftBank shop. So when users had to just go to the shop -- so is there any compensation or are you considering any kind of compensation for support for the customers?

So I was not aware of that the fee was different. So if I will assess and if those had to pay some fees, then we would like to refund. Thank you for pointing out.

And another question. So it's been 20 years since MNP started this October. So looking back 20 years ago or so, when Mr. Son held more than 50% and the share ratio also changed. So do you have any opinions that if you have any words to reflect or to mark this milestone?

Well, I haven't done enough yet. So I still like to do -- pursue my wish with SoftBank.

Nagata From Nikkei CrossTech.

I have a question about the capital alliance with Seven & i Holdings, which you announced in July. It's been 3 months. I wonder if you have any update that you can share with us?

Yes, we have progressed well. Since the beginning of October, LINE Mini app became available, mobile order of delivery service at a convenience store 7-Eleven was launched and 7-Eleven app and PayPay became linked -- sorry, the users who have their PayPay app with Seven & i's app linked, we started offering deals for them. So we want to work together and want to strengthen our partnership.

Any other questions from the venue. We would like to take questions from the Zoom. [Operator Instructions] Ishikawa-san, please unmute and ask your question, please.

Ishikawa, freelancer. So next year, you are presenting at MWC for the first time. So please tell us about what is the status of the AI-RAN, AI-RAN alliance.

So regarding the AI-RAN alliance, we are expanding right now. And AI-RAN, both hardware and the software have been established and alliance members -- with alliance members, we have started conducting assessments.

Next, we want to take questions from analysts and institutional investors. First, Masuno-san, Nomura Securities.

I have 2 questions. First, ARPU for the 3 months, July, September, 6% compared to last year. And if ARPU goes up that much, you should expect income more, but I see our income was down a little bit. So I wonder if it was due to some kind of expense matters? But for the second half of the fiscal year, do you expect both ARPU and income growth?

The first question, amortization of acquisition cost was heavy burden in the first half. But in the second half of the fiscal year, that impact should be smaller. So I think that should contribute to the bottom line.

Next question. For the -- CapEx in 6 months was JPY 26.7 billion, which was small to me. So I wonder if you are slowing down, investing in the AI.

Well, next year and the year after, we're going to launch a lot of data centers and then CapEx will grow accordingly. Akiyama-san, anything?

For the first half of the fiscal year, we didn't have a lot of acceptance of AI-related facilities. But next half, Sakai data center, Tomakomai data center will be accepted. And then the result will be represented. So the investment in the second half of the year will be a substantial amount. That's how I understand your answer.

Next, Kikuchi-san from SMBC Nikko Securities.

About the upward revision of the forecast and also the next fiscal year's forecast. And so this time, revision is due to the steady growth in the consumer and enterprise and also LINE Food also made some revisions and also SB Energy, the profit from the SB Energy. So in the second quarter -- so I would also like to ask about any revisions in the spring of next year? And so regarding the SB Energy, the sales -- the profit from the sale. So what would be the positive outcome of that?

So the SoftBank itself is focused on the cloud AI sector is now in the monetization phase that we went to. That's why all of these are showing upside. So cloud AI is 37% increase. So I believe that is the driving force. But I believe that it could have been awarded a little more in both operating income and net income. Looking at the SB Energy's sales profit. And you may be considering the security-related measurements, but I believe that you are showing a bit conservative figures though. Well, yes, if we consider the first half of the year, then it could be more than that. But the figure we show present has to be the one that we can -- we are confident to achieve. So that's why -- and also considering the interest rate increase was also reflected in the figure we planned. So if that could reflect on the second half or not, we needed to see how it will impact. But -- and if do we have any major investment or some surprise, it's not that we do have anything like that to become conservative.

Another thing is that JPY 600 billion was the target that you planned and the dividend is JPY 89, not JPY 90. What is the reason that you said at -- sorry, it's JPY 8.9, right?

Yes. Okay. So if we achieve JPY 600 billion, we said that we would provide JPY 9. So you are talking about that, right?

Yes.

We upwarded net income towards the end of the year, but we would like to discuss in our BOD meetings. Therefore, we have refrained from indicating any specific dividend amount at this stage. So JPY 9 for JPY 600 billion is that our dividend policy should be based on what we are confident to be able to achieve. So we want to clearly express our thinking on the dividend policy, and this something I'm currently particularly focused on. In the first 5 years, the dividends remained flat, although the payout ratio was very strict, the amount stayed consistent and steady throughout. So our focus is to continuously provide the dividend. So also taking the good balance between our sustainable growth. So that's how we would like to make our judgment, but we would like to pursue what we have committed to do.

And another point is that next fiscal year's plan and in the midterm plan, though, it was just disclosed 6 months ago. So net income, excluding the onetime factors, so are you able to achieve your plan?

By next earnings result, we would like to come up with more specific figure. But -- so the next fiscal year should be higher than what we have set for this fiscal year. And so rather than focusing on the figure itself, we would like to exceed what we have targeted.

Next, Tokunaga-san from Daiwa Securities.

I have 2 questions. First, about the NTT DOCOMO's price revision. And for SoftBank, you may have some discussions internally about price revision, maybe next year or when are you going to decide, depending on how the market dynamics goes and how users accept new price plans?

Of course, we all consider price plans and discuss price plans. But again, memory for servers are increasingly expensive and payroll is also getting expensive. So for that kind of rising cost, needs to be taken into account when we discuss price plan. So we will keep watching how the market goes before making a final decision.

Then my next question is about data center like Tomakomai in Hokkaido. So you mentioned that in the U.S., you are cautious, but maybe in Japan, there is a supply demand issue. And may be different -- other carriers may have different discussions, but what's your plan for data center building going forward and customer pipeline?

Well, we have received inquiries from many customers like in a double-digit number of customers. But we want to be rather cautious because Japan has a very expensive electricity price and building costs and the long lead time for buildings in Japan. So if we accept all requirements, requests from overseas customers. Do we have enough capacity left for Japanese customers? That's one question we have to answer to us. And also in Sakai, we are building compute for physical AI. And once that kind of compute platform is ready, that kind of platform should be designed specifically for different industries. And the inference requirement will be getting bigger and bigger. So from SoftBank's perspective, we want to look at domestic customers because we have limited resources in data centers, and we want to treat such precious resources allocated for domestic customers. And if there's an opportunity overseas, once our model is ready to be sustainable in the long run, we want to deploy to overseas customers. So you said that you have a lot of request or demand from customers for Hokkaido. Those customers include overseas customers? I'm talking about customers in Japan. And in fact, demand in Japan is getting higher and higher. And the moment we build our capacity, it's going to be committed soon.

We'd like to take questions from one more person before closing. Tanaka-san from ITmedia.

I have 2 questions about telecommunications services. DOCOMO will revise the pricing of ahamo from December. And also the data capacity will be 40 gig. So as for SoftBank, Y!mobile line would be the rival of ahamo. So do you have any plans to increase the data capacity like 40 gig?

We have no plan to do so. The DOCOMO's campaign from August to increase the capacity, and that did not actually impact us at all. So I think we are competitive enough.

So the next question is about Opensignal's assessment and KDDI took 5 consecutive years. So SoftBank ranks the second, very close though. So how do you see this result? And how we're going to improve your quality?

Yes, we have been in the second place for consecutive years expertise. However, regarding the downloading quality, we received the first place. So we barely allowed the field to pass. So that is the nature of our relationship, but I want to keep working hard so that we do not lose. Thank you very much.

This concludes SoftBank Corp's financial results briefing for the second quarter of fiscal year ending March 31, 2027. A recording of today's briefing will be available on demand on our corporate website at a later time. We hope you will find it useful. Thank you very much for taking the time to attend SoftBank Corp's financial results briefing for the second quarter of fiscal year ending March 31, 2027. . [Statements in English on this transcript were spoken by an interpreter present on the live call.]