Google’s Chip Push, SK Hynix’s $29 Billion Buyback

0:00The most effective people at work aren't working harder than everyone else. They're working smarter, inside better systems. Superhuman Go, from the makers of Grammarly, is the AI assistant that works inside every tool you already use, always ready, already aware of what you're working on. It's a teammate whose only job is to help you be better at yours. With Go working with you, you can show off what you do best. See what Superhuman Go can do at superhuman.com. Today's episode is brought to you by ChatGPT for Business. As a listener of this podcast, you're looking for ways to help teams move faster, make sharper decisions, and turn scattered context into work they can use.

0:41ChatGPT for Business can help. ChatGPT for Business gives teams a shared workspace with admin controls, permissions, and access to work and codecs in ChatGPT. This means your business can move from question to answer and code to rollout quicker. Join over 10 million business and enterprise users worldwide already using ChatGPT for work. Download the ChatGPT desktop app or contact sales to learn more. When you're running a business, the best days are the ones where priorities stay on track. For midsize and large companies, risk can affect multiple parts of the organization at once, from property and liability to cyber and regulatory challenges.

1:20At that level, managing risk becomes an ongoing discipline. At the Hartford, the focus is on helping businesses manage risk before it turns into something more disruptive. And when losses do happen, that work is paired with insurance coverage shaped by years of underwriting, risk engineering and claims experience. Learn more at the Hartford dot com slash risk mitigation. Policies provided by Hartford Fire Insurance Company and its property and casualty affiliates, Hartford, Connecticut. it. Bloomberg Audio Studios, podcasts, radio, news.

2:00Bloomberg Tech is live from the heart of Silicon Valley with Ed Ludlow in San Francisco. This is Bloomberg Tech coming up. Marvell and Google expand a partnership for chip development with Marvell giving the search giant the right to buy as much as$12.2 billion in stock. Plus, SK Hynix plans to buy back$29 billion of shares and return more profits to investors to stabilize its stock after falling more than 50 % in two months. And startup Fuse raises$100 million to bring commercial nuclear fusion energy to the US and its allies. We're going to speak with the CEO. So 90 minutes into the US trading session and technology stocks are under some modest pressure.

2:45The big catalyst in markets this morning was the Treasury stepping in with a plan to boost buybacks of longer dated bonds. You look at yields coming down across the curve, but on the 10 year in this case, chip stocks actually lost momentum and now underperforming in that case. One outlier in that story of the chip sector is Marvell giving Google the right to buy more than$12 billion of stock. It's about a deeper agreement for cooperation on chips, chip design across a number of facets. At one point in the session, Marvell on track for its best day since the first week of June. Let's get more with Bloomberg's Ian King, who leads our semiconductor coverage out of San Francisco.

3:23This is custom silicon in a way. Marvell is another name that Google is looking to diversify the chips it has access to. Yeah, I mean, the devil is in the details. if you read the list of products that Marvell might possibly be making for Google, they're kind of the backup singers in the band, but there is one important thing in there, which is the AI accelerator. This is the type of chip that NVIDIA makes. It's the cornerstone of NVIDIA's product lineup. And therefore, this shows or perhaps shows that Google is becoming more ambitious about its own chip efforts and is looking to use Marvell to help it.

4:03In the most basic terms, we know about the Google TPU, which Google has worked on with Broadcom. So Broadcom helps to design and then take that chip to production. Does Marvell do the same thing? It basically is a partner, but Google would still sort of own the IP of the chip. Yeah, no, I mean, these two companies, Marvell and Broadcom, are the absolute leaders in this market where if you are whatever kind of company you are, you have a bright idea. We need our own chip to do this thing. But you don't really know how to do it, right? So you go to them with your bright idea with some of your IP you've already developed, and they package it all together, and then they take it off to TSMC and get it manufactured for you.

4:41And I'm not going to ask you this question unfairly, but circular financing? I mean, this arrangement is basically a warrant to give Google the right to buy stock, along with, in parallel, a partnership to design multiple different chips. Two ways to look at this from the perspective of the chip makers. look, we're giving, in this case, investor money to a valid customer because, you know, this expands our business. It's a good use of our resources. The other way to look at it is, well, hold on a minute. If this is such a good deal, why are we giving something back to our customers? Using the technology.

5:20And, of course, this is something we've seen with the number of names in the chip space when it comes to AI accelerators. Bloomberg, Ian King, thank you very much. Let's stick with this Marvell and Alphabet story. Allspring Global Investments Portfolio Manager Matt Whitmer is with us. And in Matt's ETF that he oversees, Alphabet's a number three holding, I think. You know, this is a diversity of silicon story. It's a finding chips anywhere you can story. But as somebody that is focused on tracking Alphabet, how do you interpret it? Yeah, I think, I mean, I think it's a, first of all, thanks for having me.

5:52But I think it's an opportunity across the board to make sure that you're just, have enough supply. I think that's really what gets down to the nuts of it is just making sure you've got tons of demand out there and you've got to be able to make sure that you can meet supply. And having a broad-based supply chain, I think, is going to be helpful in the long run. Matt, you subscribe to the five-layer cake thesis of the world of AI. Alphabet sits across a few of those layers now, if you think about it from the chip perspective. How do you feel about the circular financing question in this case? Yeah, no, I think it's important.

6:29I think it's a key debate that's out there. And, you know, whether or not this AI financing ecosystem is truly building some genuine capacity or if we're just recycling capital to some degree, I think only time is going to tell. For us, I mean, I think cash flows and ROIC are going to be important KPIs to watch kind of going forward. I'm looking at the composition of the ETF. So there's the five-layer thesis, but semiconductors are still a kind of important part proportionately. What's your thinking there? Yeah, look, I think that the AI opportunity is broadening across all five layers. And I think value creation is going to expand across those mega cap infrastructure and implementation into real-world applications.

7:15I think concentration could limit some of the exposure to a full kind of AI opportunity set. So we've been advocating for a barbell approach just to make sure that you're maintaining kind of a core exposure to incumbents while also kind of adding some of the edge beneficiaries. I think it's really going to help capture multiple phases of growth and really elongate the cycle for investors out there. And so even in that one spot on the infrastructure, Chep Semi, we want to make sure we're broadening our exposure there. Looking at NVIDIA, AMD, LAM, Broadcom, ADI. So we've got a lot of exposure with just inside that one layer of the AI cake, if you will, from that perspective.

7:57So we want to make sure we're spreading our bets across everything right there. We're talking about the Allspring LT large core ETF. and you guys talk about the holistic portfolio approach, there is more than just those chip names. Where is the next phase for you that you need to be focused? Yeah, I think as the AI infrastructure builds out, I think obviously it's been rapid capital intensive. That's really benefited the hyperscalers and the infrastructure layer. As that build progresses, spending is going to increase and that's going to shift towards the implementation phase of the AI value chain.

8:30I think industrials is a really interesting area where you'll start to see some of that. There's a lot of investors that are seeking exposure that combine both kind of secular and cyclical characteristics. And I think industrials can take care of a lot of that. There's a lot of AI use cases that are starting to spread across capital goods and diversified industrials. It's significant. It's expanding. They're doing that with enhanced product design and engineering, factory automation. You think about labor productivity. They're even using predictive maintenance services. And there's even dynamic pricing models in there that I think are going to help a lot.

9:05Matt, in a few minutes' time, we're going to talk with the team about the momentum in Anthropic for the third day in a row. And there was a really interesting point in your notes about Salesforce. The idea that sometimes a software or technology is entrenched and getting people to change their habits is hard. You see Salesforce being in a pretty good spot and the idea that bespoke solutions from the frontier labs that will come for them, you're not so sure. Yeah, look, they are fully embedded with their customers. They've got great access to customer data. They've got that operational scale. And I think you're going to see AI only enhance their product set.

9:42If you think about what's going on with Agent Force, I think they're really well positioned to be able to not only keep those clients that they have right now, but I think the product set is going to get better, like I said, fully embedded inside their customers. So I think it's just that that access to data, if they can keep that, I think they're going to be in a very good position to produce some really good results here going forward. Matt, what I don't see in the ETF is any of the memory names. Why? Yeah, no, it's, you know, Micron's an interesting example of that. It's a name that we've kicked the tires on.

10:18Look, we're trying to build portfolios of high quality companies that for some reason are inefficiently priced relative to their long term growth prospects. We look a lot at cash flows. Cash flows are really to us what can drive the true value of a company is that future stream of cash flows. Micron's had over the last five years, they've had several quarters where they've had negative free cash flow, and that's kind of kept us away. I think I'm right in saying that Allspring LT Large Core EDF, slight outperformance relative to the S &P 500 year to date. But what are the main drivers for you right now?

10:50Yeah, once again, it gets back to the company's ability to generate cash and what are they doing with that cash at the end of the day. So we really want to make sure we understand what companies are doing. They're driving those cash flows back into the core fundamentals of the company. They're then investing in emerging growth opportunities. And then finally, they're shareholder friendly. So kind of in that order. So we're long term investors. We have a consistent, disciplined approach and we're focused on the fundamentals. Matt Whitmer, Allspring Global Investments Portfolio Manager, ending us on memory, shareholder-friendly and cash.

11:23Let's talk about shares of SK Hynix, particularly the ADRs. This is the US-listed shares. The company is making a major move to win back investors after a sharp slide in shares, unveiling a$29 billion share buyback plan while pledging to return more of its free cash flow to shareholders. For more, let's bring in Bloomberg executive editor for tech, Peter Elstrom. This is more about the Korean listed shares, but it is an effort to stabilize the stock and it's multifaceted. What is SK doing? Yeah, that's right. As you mentioned, SK Honex is talking about buying back$29 billion in stock, 40 trillion won.

12:04And there are a couple areas to lean into here. First of all, the timing of this. Often when companies announce these kinds of buybacks, they're talking about over a year, even over two years. SK Hynix is talking about buying back these shares by mid-November at this point. So it's$40 billion being or$29 billion being allocated very, very quickly. This also may just be the beginning of their returning cash to shareholders. There may be more buybacks in the future. What they said is that they're going to increase the amount of cash that they return to shareholders in one form or another to more than 50 percent of their free cash flow.

12:41We did the math on that. SK Hynix is making a phenomenal amount of money right now. That could total something like one hundred and seventy billion dollars over the next two years. They're making a lot of cash because of these increases in the prices of memory chips. They'd also say post buyback they'd cancel those shares. And I remember when they did the ADRs, they issued more stock, which really upset Korean retail investors. They felt like they were getting diluted. What's amazing is that SK is also on the hook for a lot of cash, right? They're expanding in Korea domestically. And as SK Group Chair Che told me, they have ambitions to the United States as well.

13:18Right, right. Yeah, you raise a good question. Why would they raise this money in the United States through the ADRs and then also dish it out in Korea? There are a couple of things going on here. First of all, there was a bit more bullishness in the AI market a few weeks ago when you talked to Che about their listing there and they were able to raise that money. They also want to globalize their investor base. They want to get more U.S. investors. They want to make it easier for U.S. investors to buy into their shares in particular. Also, there's a mismatch in the valuations right now. The ADRs in the U.S.

13:49trade at a higher multiple than the Korean shares. So effectively, what SK Hynix is doing here is they've sold shares at a higher multiple in the United States, and they're buying back their lower cost shares in Korea. So they're taking advantage of this arbitrage trade, which I think we've talked on this show before. Individual investors are not able to conduct that arbitrage because of some technical reasons. There's a cap on what they can convert, but SK Hynix is able to proceed with that. Bloomberg's Peter Elstrom is going to be back later in the hour. We're going to talk more about the big stories from Asia Tech later on.

14:24Coming up, ahead of its potential mega IPO, Anthropix Credit Facility is set to exceed$10 billion. We have the details next. This is Bloomberg Tech.

14:46Alphabet paid just under 7 % to borrow longer-dated funds in its debut Australian dollar bond offering, its highest ever, underscoring how hyperscalers are facing rising costs to finance their AI ambitions. The multi-part deal saw Google's parent company selling$3.9 billion of debt with the yield on the 20-year tranche. The rising costs also underscore a global trend of surging borrowing costs as long-end government bond yields are at multi-decade highs. Another top story is talks of going public continue. Anthropik's revolving credit facility is anticipated to rise above its approximately$10 billion target, according to sources.

15:28That That would be a substantial increase over a 2.5 billion five-year facility Anthropic secured last year, though the sources say discussions are still ongoing. Bloomberg Shreda Nadarajan joins us now. There's something really interesting here, which is basically based on our reporting. Anthropic is proposing to extend the facility. It's the response of the banks who are trying to get in place for a potential IPO that I find most interesting. Explain that. Pay for play, did someone say? Look, all kidding aside, it's become the reality of modern-day IPOs. The balance sheet play has become such a critical component.

16:05Any time that a company is looking to go public and banks that are angling to find a role on a public offering will try and show every bit of support that they can for the company to show that they're committed to it and want to make sure that they snag an important role in the IPO. And as have been reports that started a few weeks back that Anthropik is in the market trying to expand its revolving credit facility. The feedback we're hearing from the market, from the banks, that there is a lot of interest, a lot of people want to sign up. What is the final amount that Anthropic will end up with isn't clear.

16:36What they would want to cap it at isn't clear. But what is clear, that there is certainly enough demand out there, enough willingness for Anthropic to be able to announce a new revolving credit line that could be well in excess of$10 billion if they wanted it. And I know every time you hear questions about or headlines about a revolving credit line or a credit line, there is some level of concern. Why is the company taking on new borrowings? But it is very important to understand that this is not a new borrowing. This is liquidity, not leverage. This is cash reserve to impress public market investors.

17:07It is to show that this company can withstand any sort of financial surprises. And that's a show of confidence. And that show of confidence especially helps right in the lead up of pitching these public market investors on what could potentially be one of the biggest, if not the biggest IPO of all times. I appreciate that because$10 billion is a lot and people do see the word credit facility or revolving credit line and wonder about it. Do we know where the ranking on banks is right now with Anthropic, like the players that are justling for position? I think even unlike sort of SpaceX, where even weeks before the IPO, we got a much bigger list of the expanded roster of banks on the IPO.

17:47This list has remained tight. It's, I believe, Goldman Sachs, Morgan Stanley, and maybe one or two others. But closer to the IPO, you can be rest assured that this list will expand. But they're keeping the information tight for now. But as we get closer and closer to that public offering, to that public listing, you will, in the final lineup, definitely see a much bigger list of banks that will perhaps go well over 10 banks, if not much more. This is one of the must reads across Bloomberg this morning. Bloomberg, Sridhar Nadarajan, Thank you very much indeed. When it comes to setting the rules for AI, many CEOs are being accused of overstepping.

18:22Fei Fei Li, CEO and co-founder of World Labs and the so-called godmother of AI, weighed in on the issue in the latest episode of The Circuit with Bloomberg's Emily Chang. Some of the most powerful CEOs in AI right now have been accused of having god complexes because they seem to think they know what's best for humanity. What do you think of that term? I think it's dangerous for any individual to think they know better than everybody else. As a CEO, as an entrepreneur, our founding premise is we are here to contribute to the society and to empower people. It's not our position to make every single decision for people.

19:04But isn't that the path we're on? I mean, these are really powerful people. I think it's too early to tell, Emily. I'm not being naive. I know what you're talking about. I also have my concern. I don't think having God complex is constructive and I wouldn't want to do that myself. But we have a democratic society. This is another reason I think our civil society should not give up on AI. We need the voices. We need the collective governance. But let's not throw the baby with the bathwater. The governance shouldn't be let's stop AI because this is a technology that can discover cure for diseases, can empower students and teachers, can help our elderly.

19:49So that's the baby here. But right now, we are in a very messy period. This is also why I feel personal responsibility to speak the truth I believe in as a scientist and an educator. that was world lab ceo fei fei lee speaking with bloomberg's emily chang you can find the full episode on youtube and on the terminal watch it tonight on bloomberg tv at 6 p.m eastern okay coming up big tech breaks out the charm offensive will it be enough to win over local resistance to millions of dollars of data center projects we have the story next this is bloomberg tech

20:32The people who seem to get more done than everyone else. They're not working longer hours or running on more caffeine. They've just stopped wasting time on the stuff that doesn't move work forward. Switching apps, re-explaining context, hunting for files. Those aren't small inefficiencies. They're hours wasted every week. Superhuman Go gives you those hours back. From the makers of Grammarly, Go is an AI assistant that sits inside every tab and tool you already use, always available and already aware of what you're working on. Ask it to draft something, summarize a long thread, pull up a file, or prep you for a meeting.

21:13Go handles it without you ever leaving the page you're on. This is what it looks like when AI actually fits into your work, instead of adding to it. It's like having a teammate whose only job is to help you be better at yours. Go keeps up so you can move forward. With Go working with you, you can show off what you do best. Superhuman Go. Find out more at superhuman.com. That's superhuman.com. Quick one before you jump back in. You're listening for ways to help teams move faster, make sharper decisions, and turn scattered context into work they can use. ChatGPT for Business can help. Chat GPT for Business gives teams a shared workspace with admin controls, permissions, and access to work and codecs in Chat GPT.

21:58This means your business can move from question to answer and code to rollout quicker. Join over 10 million business and enterprise users worldwide already using Chat GPT for Work. Download the Chat GPT desktop app or contact sales to learn more. Everyone's talking about how AI is transforming work, especially in sales. While the landscape shifts, one thing remains the same, the thrill of closing a deal. Whether it's a gong or a confetti machine, every team has its celebration rituals. Adio is designed for that moment. It's the agentic CRM that turns customer signals into actionable insights, helping you close deals faster.

22:34With revenue agents and automations working around the clock, you'll have everything you need to scale your go-to market efforts. Elevate your wins with Adio. Start your free trial at adio.com slash iHeart. Let's go over to New York where Bloomberg's Yohara and is standing by. Yohara, take it away. Hi, Ed. It's time now for Talking Tech. First up, TikTok is exploring a feature that would allow users to send each other money in direct messages, a move signaling the app's expansion into financial services. The evidence of the feature, which is already available in Southeast Asia as TikTok Pay, was found inside of code hidden in the current version of the platform's American iPhone app.

23:17Plus, Amazon is planning to expand its drone delivery operation to at least five more states in the coming months. Prime Air drones will begin delivering packages in several new markets, including Chicago and Atlanta, with the expansion eventually reaching nearly 500 U.S. cities and towns. That's a six-fold increase from its current footprint. And the AI cloud company Nebius is said to be pricing the sale of$4.5 billion in convertible bonds Wednesday today. After the closing bell, sources say the firm is marketing the bonds, maturing in 2030 with a 0 to 0.5 % coupon and 40 to 45 % conversion premium.

23:59Ed? Thank you, Yohaira. Big Tech is turning on the charm, spending on upbeat ads, donating to local charities, and making new concessions, all to win over local communities who are concerned about the data center build out. Bloomberg's Brody Ford, part of the team that dug into this in another must read on Bloomberg today, there's a data point that jumps out at me. In the first three months of this year, $130 billion worth of data center projects were basically put on hold or disrupted by local resistance more than all of last year. Take that and tell me the story. I have been shocked about every couple of months.

24:38It feels like the data center opposition just keeps growing. It's become this really bipartisan issue that cuts a lot across a bunch of different ideologies. I mean, folks effectively are worried that it's going to mess up their power or their water, or they're just upset about AI. And this is kind of the most tangible thing they can fight and fight. They have been right. I mean, 75 data center projects delayed in just three months. And we got to remember, there's so much money tied up in each of these centers in many cases. that even, say, a six-month delay can mean something pretty material for these companies' earnings.