SoftBank Group Corp

TSE:9984

Q3-2025 Earnings Call

[Unknown Executive]

Thank you very much for waiting, everyone. Now we would like to start the SoftBank Group Corp. earnings results announcement for the 9-month period ended December 31, 2024. First of all, I would like to introduce today's participants.

From left, we have Yoshimitsu Goto, Board Director and CFO; Kazuko Kimiwada, Corporate Officer, Chief Sustainability Officer, Senior Vice President and Head of Accounting Unit; Navneet Govil, CFO, SB Investment Advisers and SB Global Advisors; Jason Child, Executive Vice President and CFO. Today's announcement is live broadcast over the Internet. Now I would like to invite Yoshimitsu Goto to present you the earnings results and business overview. Mr. Goto, please.

[Yoshimitsu Goto, executive]

Thank you for the kind introduction. This is Goto speaking. Thank you very much for gathering -- sharing the time with us today. So this -- actually 2.5 years to present the earnings. Before me, actually, that our Chairman and CEO, Masayoshi Son, has been presenting the earnings. But since then, he start saying and motivated to only focus on AI revolution. So he's been working very hard for the past 2.5 years, and people wonders what Masa is doing. And time to time, I was trying to kind of brief you what he's doing. But this time, I believe this quarter, now that we are able to clarify and share the message that Masa wanted to convey. And for that, we had 2 big events. Of course, our road to ASI is the -- of course, our objective.

And sometime soon that we will be seeing AI is going to be exceeding human beings capability. So we've been thinking how should we approach to this road and how can we minimize the time to be spent to come to this era. And these are the 4 focus if we ask to limit to the 4 elements. First, AI chip. Of course, this is the biggest advantage that we have in the group because design capability that has been building is going to be still very much essential piece for the future development of AI. And based on that, we have various study and hypothesis. And based on that, we are to prepare the computing power. And for that, we need a huge size of the data center as early as possible so that we'll be able to take an advantage and utilize for the AI evolution. And for that, power is necessary to support such data center.

And in addition, robotics, which we've been making an effort for the past 10 years or so, that is another important area together with logistics that the AI can be a big leverage for. So robot is still underway, but the first things to first, which is supplying the chip and also supporting the chip for AI. And that comes to the Stargate project announcement in White House, together with President Trump, we had an opportunity to share our project. And here, you see old friend of Masa, Mr. Larry Ellison, was also together with him. And also the strong partner, Mr. Sam Altman from OpenAI was also there for that announcement. In the past to ASI, what kind of commercial service can we provide? And that came to Cristal intelligence.

You may have an opportunity to be able to see the earnings announcements by SoftBank Corp the other day, and there was some explanation about Cristal intelligence. And also, there was an announcement together with Mr. Sam Altman from OpenAI. We had more than 500 companies with about 1,000 business person joined together to hear Masa and also Sam's, what they need to say. In addition to that, we had many people from the financial institutions as well. We cannot evaluate right now for this service yet, but we would like to continuously study what will be the most necessary part or the task that we need to be solving with this service. So we would like to spend some time for that.

And here on that, I would like to share the highlights for this quarter. Important KPI, net asset value, a little bit shy of JPY 30 trillion and also loan-to-value, 12%, both are very much steadily progressed. In addition, cash position, this is another important KPI in this quarter, 3 months, we have increased. So that gave us to the JPY 5 trillion of the cash position. Net income, year-to-date, JPY 636.2 billion, year-on-year basis, JPY 1.1 trillion increase. Number four, as I mentioned, Stargate, Cristal intelligence. Those are the announcement we made. These 4 points are the highlights that we have done for the quarter and a little bit after this quarter. And let me start going into a bit details with our key KPIs.

Net asset value, loan-to-value and cash position, both -- all of those that are having a very steady progress. As for net asset value, at the end of December end, it was a bit of -- a bit short of JPY 30 trillion. But if you look at the very latest number, JPY 33.5 trillion. And you can see here the breakdown up until the end of December, breakdown is a little bit of a difficulty in share price. But after January, we've been seeing quite a good pickup on share price from Arm and T-Mobile, one of the biggest contributor. And right now that the pro forma basis, net asset value is about JPY 33.5 trillion. And I believe this number is very much close to the record high level for net asset value considering from the start of our business.

Of course, it is still just the beginning of AI era. And the important thing is to keep the momentum. Every quarter announcement, whenever it's good or whenever it's bad, we don't want to be too happy or too sad about the share price volatilities. But I would like to focus on 1 year, 5 years, 10 years and make sure that we would like to keep the momentum increasing. And that will be the most important trend for the investment company like us. Cash position. As you see, in the past 3 years, we've been keeping a very ample financial resources. At the end of December, we have some increase in cash position because we had a margin loan capacity utilizing Arm share has been increased by JPY 0.8 trillion. And also some issuance of bonds, we have increased of JPY 0.4 trillion.

To cover for next 2 years of the bond redemption is about JPY 2.1 trillion. So if you look at this number, JPY 2.1 trillion is the cash position financial policy that we've been communicating to the market. And actually, you can say -- you can tell that this is a very safe level of cash position. Loan-to-value, this is also an important KPI that asking everybody to follow. And right now 12.9%, we are too safe. Again for this quarter as an investment company, we would like to seek for the even higher performance so that it is going to be -- I believe that this is a message that maybe we can make a little bit more challenge for the equity investor as well as the credit investor.

Next, consolidated results. Gain on investment, income before income tax, net income and net sales. We are making steady growth in all areas. As a holding company, we are not really paying too much attention on sales, but the gain on investment was up by JPY 2.7 trillion. Income before income tax, up by JPY 1.1 trillion year-on-year. Net income, we posted JPY 636 billion, up by over JPY 1 trillion. So again, compared to the last same period, we saw a huge improvement. Looking at the numbers by the quarter, however, you see some ups and downs. In the third quarter, we saw a loss in investment and net loss as opposed to gain. But this is a very small aspect of the total picture.

If you look at gain or loss on investment more in detail, for example, 2 numbers, investment business of holding companies and SVF, which comprise of this number mainly. Vision Fund posted a loss of JPY 339 billion. Investment business of holding companies posted JPY 108 billion loss, but this JPY 108 billion loss by investment business hold companies, if you look at in detail, we posted a loss of JPY 377 billion of Alibaba shares because the share price went down. But remember, we monetized Alibaba share, but we still have some shares for contract. So the value of share price and value of derivative, we have to look at both. Down of share price is presented straight here. And T-Mobile share went up by JPY 185 billion. So net-net, we lost JPY 108 billion. But when share price went down, derivative value goes up.

So even though we saw a loss in terms of investment business of holding companies, especially looking at Alibaba, even though we saw a loss of JPY 339 billion, but we saw also derivative gain of JPY 686 billion, then you can say the numbers are netted out. So when the term will be expired, we will see all net out and 0. You could exclude Alibaba at all, then you could see the result of our real -- in real-term performance. And impact of foreign gain or loss on net income, which could be very significant. Foreign exchange rate is a different animal compared to a corporate performance and corporate management.

As yen gets weaker, we posted JPY 540 billion loss. pushing down net income. That was the result of third quarter. Yen was JPY 142 per dollar as of end of September, but it got even weaker down to JPY 158 as of December 30. That had a positive impact on NAV, up by JPY 3 trillion, thanks to the share price and also foreign exchange rate. From accounting perspective, weaker yen has positive impact on equity, but negative impact on consolidated net income.

Now SoftBank Vision Fund performance. As I mentioned a little bit earlier, from October through December last year, we posted $2.3 billion. For SV 1, $1 billion; and SV2, almost $1.2 billion. That was the loss from investment. If you look at Vision Fund 1, if you take a look at the left-hand side, public portfolio companies, some companies saw value up like Grab whereas Coupang saw value loss partly due to political turmoil in Korea, end of December. As for Vision Fund 2, most of the portfolio companies are private companies. For this particular quarter, those companies saw lower fair value. And also weaker yen and Korean currency against dollar had some impact on their performance.

By the quarter since FY '22, as you can see, in the second quarter, we saw a positive number. And in the third quarter, we saw a negative number, which was disappointing, but we are confident with the trend itself. As for cumulative investment return, Vision Fund 1, they don't invest in new opportunities anymore. Against investment cost, about $70 billion already exited. And for Vision Fund 2, they have many portfolio companies that have a great potential in the future, and they are actually growing. So we are hopeful that those portfolio companies will improve their performance. Talking about the portfolio, year-to-date, we saw 4 listings this year.

And on the right-hand side, you see total fair value of late-stage portfolio, $33 billion. For example, SoftBank Vision Fund 1 has ByteDance and Fanatics and DiDi, which can be said as a private company. As for Vision Fund 2, there are a lot of IPO candidates and other companies are almost ready for IPO like Klarna and PayPay has a great potential, and OpenAI to which we made a follow-on investment. So again, we have great expectations from those portfolio companies.

Of the portfolio companies who got listed this quarter, Swiggy. It was the largest tech IPO globally in 2024. Total offering size was INR 113 billion. And we believe that this can be even growing further for mid- and long term. And SoftBank Vision Fund made new investments, as you can see here, including Databricks, which delivers integrated cloud-based platform to enterprises and DayOne pioneers cutting-edge data centers in Asia. And Helion, it develops next-generation fusion technology to produce low-cost clean energy. And QuEra, it develops scalable quantum computers. So in the last quarter, those were the major investment made by Vision Fund.

OpenAI, which everybody is interested in and looking at the moment, we made primary investment in September 2024, and we did a secondary investment in January. Total investment so far is $2 billion. As you can see, monthly website visit compared to other players, OpenAI is growing exponentially and, by far, the biggest AI entity. And SoftBank Group has a partnership with OpenAI and significance of the partnership can be presented in a chart like this. And DeepSeek, by the way, is shown here in green line at the far right bottom. And as a new move in AI sector, we should welcome someone like a DeepSeek.

Now let me go into Arm. Very smooth quarter. Sales, royalty revenue, highest ever and non-GAAP operating income actually exceeded analyst consensus. Speaking of royalty revenue, there are 2 major growth drivers. One is V9 penetration. V9 compared to previous version V8, price per unit royalty is around double. So in mobile or cloud market being adopted gives us a big contribution for the royalty revenue, and we are expecting that this will be even more penetrated over the market. In Compute Subsystems or CSS, this is several licenses combined to one package products as in a service, so that the people who use does not need to design from each chip from IP, so that they can reduce the development time and also cost and at the same time, can expect the higher capability of chip itself.

And also more and more penetration that we've been seeing of Arm technology adoption. Variety of the services are now provided utilizing Arm technology. And as an example here, Graviton4 from AWS, Azure Cobalt 100 from Microsoft and Axion from Google. those are all using ARM-based chips. And license revenue actually have some fluctuation ups and down because based on the time of the agreements or the time of the renewal agreement or the tenure of the agreements, that may fluctuate the license revenue. So here, we have numbers for annualized contract value and the bar graph shown below is a trend. And if you look at the ACV, annualized contract value, actually it is making a very steady increase here. So even with the year-on-year basis, ACV went up by 9%.

In recent earnings announcement, it was announced the guidance 2024 as a current guidance, revenue guidance between $3.94 billion to $4.40 billion (sic) [ $4.04 ] billion and the non-GAAP operating expense up to $2.07 billion and the non-GAAP fully diluted earnings per share $1.56 to $1.64. And we have a little bit of decrease in top numbers for the revenue. But the important thing is if you look at midpoint, you see actually the midpoint has increased from $3.95 billion to $3.99.

And Stargate project that I would like to make a little bit of a comment on. We would like to accelerate the AI revolution with the support -- with the Stargate and to make that time happens that we would like to support the infrastructure side. And here is the project overview. SoftBank Group, OpenAI as a lead partners, and we have Oracle, MGX. MGX is the investment vehicle from Middle East. Chairman will be served by Masayoshi Son and operation lead will be OpenAI. SoftBank Group will be leading finance. So over next 4 years, we will be deploying $500 billion. And first, we would like to start deploying $100 billion. So $500 billion, that's JPY 75 trillion in Japanese yen. That's quite an eye-opening number.

But it is not that we, SoftBank Group, will be raising money with the assets or cash for the tens of trillions of yen. We would like to set up the project together with lenders, equity investors that we would like to proceed the project together with them. There will be a variety of data center that we will be addressing and the size may differ one by one. But of course, as a mega data center, as a Stargate project, we would like to pick up both data center and work on address that. And after accumulating such efforts and projects, we are assuming it will be achieving or reaching close to or maybe on $500 billion. And in other way that we can show as a project illustration, the earliest or the first one will be, for example, we call it Project 1 and the next one or maybe in parallel, we may be addressing Project 2, Project 3, Project 4 and so forth.

So each one of the project may be this city, in this state, this real estate available here that we would like to create and build data center and deliver chip systems and also need a supply of power. So that whole project will be addressed by this Stargate project. So amount is large; however, how are we going to raise such money? Maybe some of you are very familiar on this, but the total amount of the project for those partners of this project, we'll be playing an equity funder and also having a tranche approach by risk so that we'll be able to secure the necessary amount in total. So equity and maybe preferred tranche, a very top, most safe product like senior debt.

So below there, we may have some subordinated mezzanine and so forth. So based on each tranche, we will have each respective investors around the world. So some investors would like to go for a better return, even they need to cover some risk or some investor prefers the safe, but steady return. So project finance, when we call it, how much -- market availability there may be another question you may have. Roughly speaking, in the last 2 years, we don't have a last year number, but 2022 and 2023 wise, in 1 year, between JPY 350 billion to JPY 380 billion was actual numbers for a year. So project finance market is such a huge market as well. So this time, we will be addressing a project, which will build AI infrastructure.

Usually, when you call it project finance, you may think of a power generating plant or communication facilities or some production lines. But speaking of infrastructure, compared to communication infrastructure, power infrastructure, AI infrastructure is still new to the market. So how much return can we expect? That's something that we need to speak with lenders and investors, and they will have their own calculations for that. So we would like to design a good instrument, which may suit their demand and their preference. Of course, we haven't set the numbers yet. But in an annual basis, $350 billion to $380 billion project finance market is there, assuming there that we may be able to raise $100 billion out of that or if there is any other attractive instruments, investor, of course, will be going for that.

So I believe that the product or instrument is going to be very important so that we would like to seek for the win-win design for this instrument so that we'll be able to attract lenders or investors who would appreciate each tranche so that we will be able to work project by project so that we can make a robust result out of that. So for us, the money that we need to bear is going to be very bottom, which is equity portion only. Therefore, we believe we will be able to address this. For example, JPY 1 trillion project is there, JPY 1 trillion project very bottom equity portion, depending on the project, however, it can be 10% to 20% of the portion can be equity as an example here. For example, JPY 1 trillion project, equity, say, 10%, then that's JPY 100 billion. JPY 100 billion divided by 4 companies, 4 partners here.

So we don't bear JPY 100 billion total, it's going to be tens of billions. So that's the kind of amount we are looking at. So we do like to calculate and we may be able to expect to some extent how much project. Once we have, for example, JPY 100 billion of the project, how much equity do we need to put? And that amount can be well managed within a financial policy, keeping our financial policy and address that. So that's how we are looking at for our portion. So like we have been communicating with investors and market that the financial policy is always our focus, and we'd like to pay a very close attention to that to address projects. And as technology partners, we have, of course, NVIDIA, Microsoft, Oracle, together with Open AI, those are the industry leaders is going to be our key technology partners.

Now let me talk about Cristal intelligence, which was announced by Mr. Son the other day, he appeared himself for the first time in a long time. Earlier this month, together with Sam, he made announcement, strategic partnership. SoftBank Group, Arm and SoftBank Corp, which is our telecom unit. Together with OpenAI, we will work together to drive Cristal intelligence. What is Cristal intelligence? We want to accelerate adoption of AI agent for enterprises, so what we want to do is to securely integrate all systems and data of companies so that we can provide most innovative AI customized for them. And we want to help accelerating their operation processes and improvement of their businesses.

More in detail, if you can see the slide -- talking about data or information to run the business, a company could have these kinds of information or data, as you can see on the left-hand side, including e-mail, business discussion recalls, meeting materials, et cetera, and system source scores. So those data will be input into Cristal and Cristal intelligence learns to improve the quality of inference. And output will be a capability for a company to transform their business operations and turn them into AI agent, and they can operate and automate their operations and business management. But we want to try by utilizing a lot of data that we have before releasing to a customer.

Looking back 17, 18 years ago, when we launched iPhone in Japan, what we did was to have group companies to use iPhone, completely change the business phones to iPhones and let them use iPhone for their own businesses before marketing them to the customers. Also, we saved a lot of paper. Again, my point is, if you want to be committed to a particular product, you need to try by yourself before sharing it to customers. So learning from those experiences, we are going to do the same, which is to have SoftBank Group companies to use this AI capability and market to Japanese enterprises in Japan, the sales effort will be mainly driven by a SoftBank Corp.

And group will pay $3 billion as an annual purchase. And once commercial products are ready, we have group companies to use the products, and we will get a fee from those group companies on pay-for-use basis. We don't know how much and which of the group company will use the Cristal, and we need to figure out what will be the most attractive product that we can sell to the market. So that's something that we want to do before releasing to a commercial market. As for the structure, as you can see here, SoftBank OpenAI Japan will be held equally by OpenAI and intermediate holding company, which is held by SoftBank Corp, 51% and by SoftBank Group, 49%. And SoftBank will consolidate SB OpenAI Japan. And we hope that we can get started as soon as possible and start developing a product and want to help transforming all industries in Japan, and SoftBank will have exclusive right to market the product in Japan.

This is Sam's message that he delivered when we made an announcement the other day. This partnership with SoftBank will accelerate our vision for bringing transformative AI to some of the world's most influential companies, starting with Japan. Again, we want to launch the service as soon as possible. Last but not the least, our finance strategy, which remains the same. We adhere to financial policy, maintain LTV below 25% in normal times. We want to maintain at least 2-year worth of bond redemption in cash and we secure recurring distributions and dividend income from SPF and other subsidiaries. This policy remains unchanged. For this year, as the finance unit, our mission is to simplify business operations. We want to keep at the same pace as how business goes.

But again, at the same time, we want to really stick to our financial policy. For FY 2024, our financial strategy is to investment for future to grow net asset value. One example is our investment in OpenAI and Stargate project and Cristal intelligence. This is just the beginning. And there are a lot of ways to explore, but the goal is the same. As for invested amount, looking at Q3, less than $1 billion or $600 million, but we made $1.5 billion in January, so somewhere around $2 billion is something that we want to see as investment. But after the announcement of Stargate, it depends on how the project goes, but we may accelerate our new investment going forward.

Share buyback, which we have been doing for some time, the progress to the max allowance is 42%. In summary, NAV, cash position, net income, big announcement. Again, we want to realize ASI, and we want to lead in the efforts. And we will explore different opportunities, and we are making steady progress towards the summit. Thank you very much, and we'd like to happily take your questions.

[Unknown Executive]

Now we would like to take questions. First, we would like to take questions from the floor.

[Operator Instructions] MJ from Bloomberg.

[Unknown Analyst]

You mentioned that the investment may be accelerated. And regarding Stargate, you mentioned that it's going to be project by project basis, but $100 billion execution is going to be the amount, and I believe that the timing or the schedule, when the project is decided and when -- how much will be the initial investment amount? Will you be able to clarify on those, please? That's my first question, please.

[Yoshimitsu Goto, executive]

Yes, thank you for your question. If I can talk about that, I would love to. However, the project has not been fully started yet. Of course, there are candidate sites and so on. And we have members that start discussing on that. So the first case, first project, we would like to execute within this fiscal year. Overall project I said that we have an existing speed of investment pace each through the -- our balance sheet or through Vision Fund, so there are existing investments going on. On top of that, we have Stargate that we would like to begin deploying with $100 billion, and that is going to be project by project, and we will bear a little by little.

Therefore, that will be kind of incremental on top of the existing pace of the investment. That's why I said we may have some acceleration of the investment amount. As for project, whenever we're ready, when it comes to size, schedule, financing structure, I will share with you. So please bear with us for a moment.

[Unknown Analyst]

And my second question, there are many media covering OpenAI new funding round. Are you, SoftBank Group, also participating? So 40 billion maximum may be possible from SoftBank Group to participate in funding round by OpenAI. Do you have any comment about that? And Vision -- the $5 billion, is that related to maybe a new investment in OpenAI? If I could ask Vision Fund [Foreign Language].

[Yoshimitsu Goto, executive]

Yes, let me go first. I know there are many articles, but those are all speculation, and I hesitate to make any questions regarding speculation. And as for investment increase, I believe the commitment increase for Vision Fund if there is any necessary investment, we can flexibly increase the commitment. So that's also case by case so that we may also increase. As a Vision Fund 2, we may increase, and it is possible for us to increase the commitment if we need to do so. Is that okay? Do you want Navneet to answer? Navneet, do you have a comment?

[Navneet Govil, executive]

So I would reiterate the point that Goto-san made. For Vision Fund 2, you're correct. In the December quarter, the commitment from SoftBank Group to Vision Fund 2 increased by $5 billion. It wasn't necessarily for any particular investments or the pipeline. As you know, in 2024, we made 22 new investments in a lot of AI companies. And this is to continue to allow Vision Fund 2 to continue investing in AI companies.

[Unknown Executive]

Next, from the floor. [ Morita from Toyo Keizai. ]

[Unknown Analyst]

First question about Stargate project finance. You mentioned that you designed financial instruments that are attractive, which is very exciting for me. But when it comes to designing debt finance instruments, is it going to be led by SoftBank Group? And you mentioned that building data center like in Texas, have you executed project finance for that particular data center in Texas?

[Yoshimitsu Goto, executive]

You are right. As for finance, SoftBank will take a lead in terms of finance. So my team will play a key role, leveraging the expertise and experience to work on financial instruments. Site-wise, we have not finalized. And whenever we can communicate with you, we will let you know.

[Unknown Analyst]

Second question about ASI. Looking at the presentation in second quarter, I remember you mentioned 3 things: data center, chip and robotics. And you added electricity or power in your presentation. What is the background? For example, nuclear fusion, Helion, which area you are most interested in, in terms of electricity or power? Or more like investment in power company, for example.

[Yoshimitsu Goto, executive]

I think, both could be possible. For example, someone like Helium that has technology, but that technology to use nuclear fusion for electricity power is not validated for commercialization. But demand is there in terms of power and electricity. And Stargate project, we are looking at about 4 years as a term, but it's hard for us to predict how the technology evolves. For example, nuclear fusion technology, is it going to be validated sooner rather than later? Then we need to make a business plan to address such technology as nuclear fusion.

Some people said that it takes 3 years. Some people said 10 years, 30 years. We don't know when such technology is validated, so we cannot wait until then. So we have to be flexible in line with technical evolution, not only Helion, but also we are exploring opportunities in the tech sector, and we want to make an investment through Vision Fund. The project size is huge, but we want to make sure that the project is very narrow focused.

[Unknown Executive]

Then next question. Second row on your left.

[Unknown Analyst]

My name is [ Brian from Diamond. ] My first question is about joint venture with OpenAI. So as a usage fee, $3 billion or the JPY 450 billion to be paid. And SoftBank Corp, other group companies is going to pay SoftBank Group based on pay-per-view or the pay per use. So this JPY 3 billion, this may be a bit different way of using money compared to the past investments by SoftBank Group. So this is not the investment. This is going to be expense, I believe. Is that correct?

And also for this expense, I believe the -- I want to ask you how are you going -- is it going to be effective in terms of expense? As a CFO, how you look at this way of using this money? Is this regarded like R&D type of the expense or that you are not expecting the immediate return out of it? Or this is going to be sales for the joint venture, so it is okay to expense this amount or so. So this $3 billion of expense, how are you looking at that money? That's the question I have because the way of using money is a bit different from I used to understand from SoftBank Group.

[Yoshimitsu Goto, executive]

As a compensation for AI service, of course, like you said, this is going to be expensed. But $3 billion per year, how are we going to pay, has not been finalized yet. And first of all, we don't have a service available yet. And in a few months or in when that we're going to be able to use this service and how are we going to pay for pay per use basis, how are we going to set the pricing scheme? That's also going to be finalized. SoftBank Corp may use a lot. So it doesn't mean that the large expense needs to record. And Miyakawa, CEO of SoftBank Corp, mentioned several times in his earnings that it's going to be paid for use, and we don't know or they don't know how much they're going to use. So that is something that we can discuss internally.

So we're going to be kind of the experimental and we kind of take it to SBG and SBG is a small company as a holding company. So we will not be using a large portion of it anyway. So we will be asking those operating company to utilize, but we do need a little bit more time to finalize all that. So let me go into much detail. So $3 billion, that I believe is going to be recorded as a sales of the joint venture, if that's the case. R&D, it's not going to be go on because I don't believe that it's going to be breakeven right away, so that those money is going to go on for the development costs and everything.

So in the meantime, that also is kind of a going -- ongoing discussion. How much money from the JV point of view, who receiving this money, how much cost they need to expect to provide such a service. That's -- based on that, they may be able to see some margin or they may just go break -- even though they may go a loss. That's another thing that we need to take a little bit more time to think about that.

[Unknown Analyst]

And for my second question is regarding Stargate. $100 billion to be the first deployment. Then equity portion is the SoftBank Group's portion I heard, and that to be decided project by project, you mentioned. But looking at the graph that you show us, compared to total amount, equity portion is going to be quite limited from my understanding, only looking at the slide. So just for my reference, for $100 billion, are you going to be bearing just a little bit? And how much that will be? Do you have any kind of a good guidance within the range? How much financing do you need to do? So I believe the SBG portion will be raised by SBG by project basis. That's my understanding. Is that correct? Can you give us kind of estimates?

[Yoshimitsu Goto, executive]

So the project for the target, and we are going to have Project 1, Project 2, Project 3 and on. And for each project, we will be structuring the financing by ourselves. And when you look at project finance world, each respective project may differ when it comes to how much equity portion necessary. And that varies, but still in total view, it's going to be minority, of course. We have to kind of once do; otherwise, we cannot really tell. I don't want to go numbers first and that goes on.

So for example, we have a company called SoftBank Energy, which is a power generation company in United States. So they have their own power generating project one by one. So equity portion for that, sometimes 1 digit, sometimes teens, sometimes 20s. So it always varies based on the project. But what I can say is that it's still the minority. Equity is going to be minority, and that will be shared among those 4 partners. So based on that basic understanding, it won't be a huge amount of money.

[Unknown Analyst]

Let me add one more. So equity ratio may differ by project, you said. And by project, equity ratio is going to be decided. And as SBG, you will decide whether you're going to raise money or you use your own cash on balance sheet. Is that kind of a way of decision-making? So you have JPY 5 trillion of cash position right now. So is that going to be within that cash position? Or are you going to raise outside of cash position?

[Yoshimitsu Goto, executive]

So the money out of our cash position is going to be on the equity portion. So we are not going to building a bridge for the total amount or anything like that. That's not our understanding. So from the beginning, we would like to set the structuring of the financing, setting the participants and address that. So that's the project finance schedules or the milestones that we understand. So by project finance, SBG portion is going to be decided and SBG will raise money outside from the -- of the company by project. So project, it's not us that we're leading this financing. It's going to be project actually leading the financing. It's going to be nonrecourse financing, so it's not us that leading the finance. We -- our portion is going to be limited to equity. That's kind of the main part for this structure.

[Unknown Executive]

Any other question from the floor?

[Unknown Analyst]

[ Shikata ] from Nippon Keizai. I have 2 questions, Nippon Keizai Shimbun. And first, it's reported that Elon will propose to buy OpenAI at $97 billion and Sam Altman shows he's not going to accept that offer. So actions like that by -- do you think that it would have an impact on Stargate feasibility or other projects that you want to proceed with?

[Yoshimitsu Goto, executive]

Yes. We heard the news, and we were a little bit surprised. And Elon has his own AI vehicle, so that means he can be our competitor. In that sense, something like this could happen, but it's up to OpenAI to make a decision. We will see how it goes. And following investment in OpenAI, even more, it's kind of a speculation, so we cannot answer to the speculation at the moment.

[Unknown Analyst]

Next question is about, you mentioned that you refrain from talking about further following investment in someone like OpenAI and the Stargate. It seems like you are confident with financing the project. But when it comes to investment size as big as JPY 5 trillion, then I wonder you are going to enter the stage where you will accelerate talking or communicating with potential investors globally. You are talking about our cash position when you said JPY 5 trillion?

[Yoshimitsu Goto, executive]

Yes. In fact, too plenty position that some people may say and some people may say that it's inefficient if you have too much cash position. But we want to make sure that we are flexible whenever wherever something happens. Of course, going forward, as we spend, then the cash position will decrease. But still, we are in a very safe area. 2-year of bond redemption is about JPY 2.1 trillion. We have way more than that. On top of that, we have JPY 34 trillion -- or JPY 33 trillion of net asset, obviously, not including borrowings.

And Arm is going to continuously grow and OpenAI continues to grow and PayPay and other Vision Fund portfolio companies will continue to grow. And those could give us more opportunity for us to finance. We could have more options. And we need to continue investment. We need to see the plan to cultivate in 1, 2, 3 years ahead. So the investment style may be different from other investment companies. But again, net asset value is the key. And we want to keep net asset value even though we continue to invest.

[Unknown Executive]

Any other questions? On your right-hand side, first row.

[Unknown Attendee]

[ Toyoshima from TV Tokyo ]. So may be a bit redundant from your explanation, Stargate project, Elon Musk saying that you cannot raise such a huge amount of money. And do you have any response to that? Do you have any message to him? And if so there may be a debt financing, and you said that your participation is just a part of it. But if you may have difficulty raising from the other members and because Masa has made a commitment to Trump administration, how much risk would you be willing to take? And $100 billion, how are you going to make your commitment in maximum, amount-wise?

[Yoshimitsu Goto, executive]

First, I saw the -- Mr. Elon Musk's comment. And I don't believe that any emotional response make any good results, so I just wait and see how it goes. We have a long history of addressing project financing, and we have a lot of experiments. So maybe we can -- you can wait and see for that. And for your latter half of the question, having Stargate project start, we may face difficult moment for debt financing. But if we have facing -- we face the difficult time for debt financing, we should slow down the speed of the project, probably that will be the way. Because this type of the product may be very much acceptable for the debt financing market.

And if it's -- we face the difficulty, meaning that the financial market itself may be in difficult situation so that we shouldn't make in too much stretch.

[Unknown Attendee]

And Mr. Elon Musk mentioned about the unsolicited bid. And I believe that your position needs to just wait and see how it goes, but SoftBank Group, you are the Sam Altman side?

[Yoshimitsu Goto, executive]

More specifically, we are partnering with OpenAI for this business and for this joint venture. No change about that. And we are not having any conflict or anything with Mr. Elon Musk, so that's just plain relationship.

[Unknown Attendee]

And DeepSeek are successfully developing the very reasonable price of the AI. And you said that the competition is welcome. But for the long-term vision, is that going to be negative impact on your strategy?

[Yoshimitsu Goto, executive]

DeepSeek is recently received -- drawn the attention by everybody, so it's quite recent, and we need to see how they're going to be valued or how they're going to be evaluated. That's something that I would like to take some time to see. And competition is always there anyway. As Sam Altman was commenting that having a good service available in the market is good things for the industry. Important thing is we just pursue the better service for the market. I think that's very well said.

[Unknown Executive]

Any more questions from the floor?

[Unknown Attendee]

[ Samura ] from Nikkei Business. Question about target project, more particularly about power electricity. You mentioned that nuclear fusion is not validated technology. But I think a realistic solution for now could be utilized nuclear plant. What do you think?

[Yoshimitsu Goto, executive]

I think nuclear power is lower, very lower in our priority in terms of options or you are looking at renewable energy instead?

[Unknown Attendee]

Yes.

[Yoshimitsu Goto, executive]

Through the SoftBank Energy, we have been working on renewable energies mainly. And going forward, renewable energy should be the first to be leveraged going forward.

[Unknown Attendee]

The second question about Cristal intelligence. At the press conference, you mentioned that 1 company in 1 industry as a start. And I wonder, since then, how many inquiries or requests from other companies?

[Yoshimitsu Goto, executive]

We are checking, and we are not ready to disclose how many companies are approaching to us. And I believe that SoftBank Corp will make an announcement when time comes.

[Unknown Executive]

Any other questions from the floor? Then that's all from the questions from the floor. And now we would like to go on to the question from the Zoom.

[Operator Instructions] Now we would like to take question from the Zoom, Mr. Sato from Nikkei Asia.

[Fumika Sato]

Yes, this is speaking -- Sato speaking from Nikkei Asia. My first question is about Stargate. I believe that the main part is investment in data center. And as far as I understand, looking at the current growth of the data center, this is mainly for the learning for the graph. And the Stargate data center may be aiming for inferences. And are you going to put the Cristal intelligence onto it? Or is it ChatGPT? Which -- what product are you going to sell based on the Stargate?

[Yoshimitsu Goto, executive]

Infrastructure user for Stargate is going to be OpenAI, so they will be able to enhance the speed of development. They are needing in huge computing power, and this is a project for that. And as I mentioned earlier, there will be 3 portion. And the biggest one will be system with chip in data center, that will be probably the most expensive part. On top of that, we need a box for that, which is data center, and we need power to run this data center. So all combined, we call it as a Stargate project. So user is going to be OpenAI, and they will be utilizing this infrastructure to develop their services and products. So is it mainly used by OpenAI? As of now, yes.

[Fumika Sato]

Understood. My second question is about Arm. Arm is making a very steady progress, including share price. And to further enhance the additional value, you may want to contribute to more AI data center or GPU. So that world, you may be -- you may want to extend your business. So what is your strategy for that? And SoftBank Group is very much now close to OpenAI. And is there any synergy between OpenAI and Arm?

Can Arm make any contributions to that, Jason?

[Jason Child, executive]

So certainly, there is opportunity over time. I would say right now, initially with Stargate, the biggest opportunity for Arm is we are going to be the CPU of choice. And so when you look at the technology partners, that's going to be our primary contribution. And when you do the math on $500 billion of investment, that's going to be millions of GPUs and every GPU has to be pared to a CPU. So that's a pretty significant near-term opportunity. In terms of ability to get into GPUs, that's certainly a possibility, but certainly nothing that we've actually disclosed our intentions about at this time.

[Fumika Sato]

Any technical synergies with OpenAI? Right now, I believe you just mentioned about business opportunity, but any synergies, technological synergies with OpenAI?

[Jason Child, executive]

Yes. So well, on the Cristal project or Cristal Intelligence, we are providing [ Claude ] AI, which is a library that allows developers to be able to develop very, very quickly in terms of ultimately what some of the end applications could be. That is actually certainly something that could benefit ARM significantly. When you think about inference, the majority of inference actually happens on the edge, meaning it happens on smartphones, on PCs, in cars, robots, et cetera.

And so the opportunity, specifically with OpenAI is to try to push as much of the inference to the edge. And because ARM is the most ubiquitous CPU platform, and we believe most of the edge inference is typically a CPU workload, and that's because these are mostly devices that are power constrained and many of them are on battery. And so as a result, they're going to rely on a CPU. And so that's certainly something that will benefit us. So yes, there are many different synergies that we're working on. And I would say stay tuned for what some of those different applications will be.

[Unknown Executive]

Next question from the Zoom, [ Mr. Miyoshi from NHK ].

[Unknown Attendee]

Again, my question is why OpenAI? You made initial investment, following investment and a new business or partnership. So you mentioned that we should look at the monthly web visit. There are a lot of AI players in the world in order to deliver ASI and AGI, do you think that OpenAI is the closest to realization of AGI and ASI compared to other AI players? Goto-san, what do you think?

[Yoshimitsu Goto, executive]

Yes. I believe that you're looking at the chart once again here. So I think it's going to be very difficult for other competitors to catch up. It takes time. And OpenAI is that much welcome and well accepted. And who should we partner with? There's no doubt, OpenAI.

[Unknown Executive]

This concludes the SoftBank -- sorry, this concludes the question-and-answer session. Thank you very much. SoftBank Group Corp. earnings results announcement for 9-month period ended December 31, 2024, is now concluded. Thank you for your attention.

The video footage of this meeting will be distributed on demand from our corporate website. Thank you once again for joining SoftBank Group Corp. earnings results announcement for 9-month period ended December 31, 2024.