1. Podcasts
  2. Planet Money
  3. Two ways big bets on data centers might affect you (Two Indicators)
Planet Money

18 Nov 2026 · 18m · 8 chapters

Episode artwork

Two ways big bets on data centers might affect you (Two Indicators)

In short

How big data-center investments by hyperscalers are being financed with large volumes of investment-grade corporate bonds, and what risks this creates for investors and the broader AI supply chain.

Guests

Zachary Griffiths (CreditSights; U.S. investment-grade debt specialist) and Jonathan Mondillo (Aberdeen Investments; manages about $170B fixed income). Later, Matt Sheridan (Ohio State University finance lecturer; analyzed credit).

Key claims

Demand exists for AI data-center bonds but pricing/yields are tightening; debt repayment depends on long, complex lease structures and project completion, creating construction, obsolescence, and lease-renewal risks. Oracle is highlighted as a potential weak link due to heavy debt (about $169B), reliance on OpenAI, and “circular deals” involving NVIDIA and OpenAI.

Notable examples

Alphabet’s $25B bond sale with 4x orders; hyperscalers issuing $90B+ in bonds (Vanguard) and $130B+ in 2026; Oracle’s force majeure notice for an OpenAI data center in New Mexico; Oracle CDS rates rising and S&P downgrade to just above junk.

Written by AI. May contain mistakes.

Chapters

0:37

The Boom and Risks of Data Centers

Exploring the impact of AI data centers on finance and productivity.

“Tech giants are borrowing hundreds of billions of dollars to finance AI data centers.”
3:38

Understanding the Bond Market

Analyzing the role of bonds in funding data centers.

“They are flashpoints in local debates about what should be built in communities.”
6:43

Investor Demand and Risk Factors

Examining risks tied to data center investments and bonds.

“Is it like Mean Girls, like The Limit Does Not Exist?”
12:26

Oracle's Unique Position in the Market

Discussing Oracle's strategy and challenges in the data center space.

“Oracle is not like the other children in the big tech company classroom.”
14:01

Oracle's Business Transformation

Explore how Oracle shifted from software to data centers amidst financial pressures.

“And about a decade ago, credit analysts viewed Oracle bonds as a safe investment.”
15:12

The Risks of Oracle's Debt

Learn about the implications of Oracle's debt and reliance on OpenAI.

“It's turned Oracle from a business that generates cash into one that burns money.”
16:19

Credit Risks and Market Reactions

Understand how credit investors view Oracle's financial stability and risks.

“This bit has more implications for the wider economy.”
18:44

The Systemic Importance of OpenAI

Investigate the interconnectedness of Oracle, OpenAI, and the broader economy.

“Lindsay Tyler is a credit research analyst at Morgan Stanley.”
A man walking through an airport, asking his phone a question

Ask about this episode

Get the gist, explore an idea or ask about something said.

Claude

On Planet Money, what are the key points of this episode?

Searching your listening

Transcript

Automatic transcript. May contain errors.

0:00This message comes from Apollo. Apollo provides access to long-duration capital and a broad range of financing capabilities, helping companies drive growth at an efficient cost of capital. Learn more at Apollo.com slash capital. This is Planet Money from NPR. If you own stocks in the U.S., it's very likely that your money is tied up with mega tech companies. That's because the 10 biggest stocks in the S &P 500 are names like Microsoft, NVIDIA, and Amazon. Regardless of whether you've invested in stocks or bonds, the fate of these companies could impact you. Tech giants are borrowing hundreds of billions of dollars to finance AI data centers.

0:42If your 401k has money invested in a U.S. bond fund, your savings could be fueling the build-out. This boom in data centers could lead to both productivity and financial gains, but it also exposes new risks. And this is where it gets interesting. It's one of those stats where my jaw dropped while looking at a computer screen. Back in September, a research group, Epoch AI, reported the cost of achieving a basic level of AI performance is down 1 ,300%. Imagine if we could get that much better at podcasting over the same period of time. Then you would be paid pennies per year. Because I didn't really think through the economics of that.

1:21Oops. And the economics might not be much better for companies investing in data centers. The cost of tokens, or the fuel that runs AI engines, is plummeting for some models. So while demand is currently out the wazoo for data centers, the future is more of an open question. The economics of AI are changing, the cost of debt is rising, and AI is getting more efficient.

1:47Hello and welcome to Planet Money. I'm Ricky Mulvey. And I'm Waylon Wong. Today on the show, two stories we've been reporting on The Indicator, Planet Money's short daily podcast. First, we'll look at this glut of AI-related bonds, the loans building these data centers, and we ask if all of this borrowing is sustainable, both for the companies taking out the loans and for your 401k. Then we'll zoom in on one company that's going full speed ahead financing data centers, and some warning signs that it could be a weak link in the AI deal world.

2:25Stay tuned for the following announcements and messages from our sponsors. This message comes from Midi Health. Most women never get real answers about their hormonal health. Midi is changing that with virtual care covered by most major insurances. Understand your hormones. Visit joinmidi.com to book your first visit today. This message comes from Bilt. Rent and mortgages keep getting more expensive. That's why BILT created three credit cards designed to offer rich rewards on everything you spend on. You can earn up to 1.25x points on every housing payment with no transaction fee, which can be redeemed for flights, hotels, a future housing payment, and more.

3:06Apply at joinbilt.com slash money. Terms apply, subject to approval and eligibility. Built cards are issued by Column N.A., member FDIC, pursuant to license from MasterCard International Incorporated. This message comes from Dell. Shop at Dell.com for Black Friday and find great PCs like the Dell XPS 13 laptop with a long battery life and powerful performance with Intel Core processors. Upgrade your whole setup for less. Dell.com slash deals. Data centers are all over the news these days. They are flashpoints in local debates about what should be built in communities. Now, if you're one of those people who don't want a data center in their backyard, and that's the majority of Americans, you might want this bond-fueled build-out stopped in its tracks.

3:55Then, of course, if you're rooting for the stock market to keep going up, you need those data centers to power the hopes and dreams and profits of the tech giants. But enough about stocks. I want to talk about bonds. Again? Why? I always want to talk about bonds because the bond market is like this thing that people never think about. And yet bonds have financed so much of human history, wars, bridges, everything in between. That's it? Just wars and bridges? Yeah, just two things, wars and bridges. But so many bridges, so many beautiful bridges, Ricky. Airports? Airports are good, too. And right now bonds are funding data centers.

4:31Bonds are interesting. I mean, that's what we're always saying to various kinds of receptivity. This is fellow bond geek Zachary Griffiths. He works at an independent research firm called Credit Sites. His specialty is U.S. investment-grade debt. That refers to safe and highly rated bonds. U.S. Treasuries are investment-grade, and so are bonds from megatech companies like Alphabet, Amazon, Meta, and Microsoft. These companies, along with Oracle, are considered the main hyperscalers. This is a term we didn't really use that much until a few years ago, and now it's getting thrown around all the time.

5:08Yeah, it's come about because it refers to corporations that run huge data centers. These hulking structures take a lot of money to build, and the hyperscalers have a lot of cash. Still, it's not enough for the kind of massive infrastructure projects they're planning. So they're getting ahead of their future spending needs by borrowing money now. And they're not just wading gently into the bond market. According to Vanguard, the five hyperscalers sold more than$90 billion in bonds last year. And they've already sold more than$130 billion so far in 2026. Zachary says these hyperscalers have made a splashy entrance.

5:48For the corporate bond market for a build-out, these numbers are truly staggering. He says the hyperscalers are doing deals where they sell$25 billion of new bonds in one go. That's a big shift from a few years ago when these companies weren't very active in the bond markets. If you go back just a few years, some of these names that are now at the very top of the issuance or debt outstanding list wouldn't have showed up in the top 10 or 20. And remember, most of the hyperscalers are considered very safe and creditworthy. The bonds they're selling are now competing with U.S. treasuries for investor dollars.

6:26That's helping push up yields for treasuries. We talked about that on a recent episode. But investors like hedge funds, pension funds and insurance companies, they don't have unlimited dollars. So we wanted to know, is there enough demand out there for all of these new AI related bonds? That's our first question. Is it just this like limitless appetite for this stuff? Is it like Mean Girls, like The Limit Does Not Exist? It's funny, The Limit Does Not Exist has definitely been thrown around quite a bit. Ricky, you remember this from the movie Mean Girls, right? When Lindsay Lohan's character clinches a victory for her team at a high school math competition?

7:02Of course I do. And the underrated musical that came out a few years ago on big screens. The Limit Does Not Exist. The Limit Does Not Exist. Our new state champions, the North Shore math leagues. So, is there a limit to how many AI-related bonds investors can buy? The limit may not exist, but what does need to shift is how much spread you're getting or how much yield you're getting. Zachary is talking about the returns that investors are asking for. The higher the risk, the more compensation bondholders want. Here's a recent example. In August, Alphabet sold$25 billion in bonds. That's Google's parent company.

7:43Zachary says Alphabet got orders worth more than four times that amount. That indicates healthy demand. Still, investors wanted to be paid a little bit more in yield. They would lend Alphabet the money, but at a higher price. The Alphabet deal is a great case study in what we think is likely to be the environment going forward. There's demand there, but the pricing is starting to get called into question. So, is there investor demand? Yes, but with an asterisk. The second question we want to ask is how are bondholders going to be paid back? Some of these bond deals are tied to specific data center projects that are funded through a complex web of private credit firms and legal entities with cryptic names.

8:26We did an episode about this kind of financing last year. Jonathan Mondillo helps manage about$170 billion worth of fixed income at a firm called Aberdeen Investments. Each contract seems to be written a little bit different. There is a lack of standardization there. And that's a risk, ultimately. One risk Jonathan identifies is construction risk. Will the project get done on time? And looking further out, there's a risk of technological obsolescence. What if the hyperscalers get it wrong on their forecasts and they're stuck with data centers they don't actually need? These are long-term lease structures, 15, 20, in some instances, 30-year lease structures or lease terms.

9:07And, you know, who's to say that there's need for that amount of space in the data center in 20 years' time or in 30 years' time. That's why you set up a shell company, then it's not your problem. Jonathan talks about lease terms because that is a common arrangement for these data centers. Companies like Meta are planning to rent these buildings rather than own them outright. And this adds another complication. The bonds for some of these projects won't fully pay out for more than 20 years. Jonathan says he's seen deals where the leases come up for renewal before the bonds mature. That's another risk that's been a concern to us.

9:43So lease renewal risk ultimately ties into that technological obsolescence risk. I mean, a lot could happen in 20 or 30 years in the AI world. Maybe the expected productivity boom from AI doesn't materialize. Maybe some tech companies will quit the data center arms race. Maybe the current fears around AI safety prompt the hyperscalers to pull back their plans, either willingly or because of government regulation. Maybe AI does bring about some kind of existential threat to humanity. All right, that's enough for Anxiety Corner. I'm just getting started, Ricky. We'll talk to you when we're all paperclips.

10:20Interest rates have been climbing fast. That pushes up the overall cost of borrowing across the economy. You've probably felt this if you recently looked at a mortgage rate or paid a credit card bill. These are scenarios that investors have to consider when they're deciding whether to lend money to the hyperscalers, money that they might not make back for decades. And again, if you have savings invested in a bond fund, that's your money at work. But for now, people are still showing up to buy these bonds. Tech companies are looking to tap markets outside the U.S. as well. Alphabet recently sold bonds priced in Australian dollars for the first time.

10:55Each new mega deal, however, brings a fresh test of investor appetite. After the break, Adrian Ma and I look at the case of Oracle, a major player in software that leaned into data centers, and now the bond market would like to have a word.

11:19Stay tuned for the following announcements and messages from our sponsors. This message comes from Midi Health. Most women never get real answers about their hormonal health. MIDI is changing that with virtual care covered by most major insurances. Understand your hormones. Visit JoinMIDI.com to book your first visit today. This message comes from Mint Mobile. Unwrap huge savings this season. All plans, even unlimited, are$15 for new customers. Go to MintMobile.com slash switch. Upfront payment of$45 for 3-month,$90 for 6-month, or$180 for 12-month plan required,$15 per month equivalent. Taxes and fees extra.

12:00Initial plan term only. Greater than 50 gigabytes may slow when network is busy. Capable device required. Availability, speed, and coverage varies. See mintmobile.com. This message comes from Dell. Shop at dell.com for Black Friday and find great PCs like the Dell XPS 13 laptop with a long battery life and powerful performance with Intel Core processors. Upgrade your whole setup for less. dell.com slash deals. Oracle is not like the other children in the big tech company classroom. Sure, the software giant is investing in artificial intelligence, and the company's revenue is booming. But Oracle has irked the credit market by loading up on loans to make its data center dreams come true.

12:42See, Oracle may be the weak link in a crisscrossing daisy chain of businesses that's holding up the AI boom. My biggest concern in all of this is the circular deals. Oracle's a part of circular deals with OpenAI, but they have one with Meta. If you look, NVIDIA's all over the place in these circular deals, Microsoft's in a few. Matt Sheridan is a senior lecturer in the finance department at... The Ohio State University. I took Matt's investing class as a student. He was one of my favorite professors, and this was an excuse to call him. Matt analyzed credit before working as a professor, too. The biggest issue with Oracle is their debt load has nearly doubled in the last two years.

13:22They're sitting at$169 billion in total debt. They're the fifth largest investment grade issuer now. I think they're the second largest that's non-financial, only Amazon has more. Oracle is one of those companies that you've probably heard of, but may not deal with directly. It's a multinational tech giant that sells enterprise software and computing power. Yeah, Oracle's software involves the less glamorous programs that run businesses, the databases that keep track of inventory, health records, and supply chains. The vast majority of big companies use its databases and software. And investors really like Oracle's software business.

13:58It's steady. It makes a nice profit. And about a decade ago, credit analysts viewed Oracle bonds as a safe investment. They were a high-margin software company. Now they're capital-intensive data center build-out. This was a huge shift in business strategy. Oracle went into the AI boom in a weaker financial spot than other tech giants. The company burned a lot of cash when it bought a health records company a few years ago. Larry Ellison is the co-founder, executive chairman at Oracle. He thinks that training AI models is a huge business opportunity. Here's what he said at a company keynote last year.

14:36It's the largest, fastest growing business in human history. Bigger than the railroads, bigger than the Industrial Revolution. I mean, it is a whole new world that is dawning. So why aren't Oracle investors happy? Well, even aside from Larry Ellison's dreams, there's his son David's dreams, too. The senior Ellison guaranteed$40 billion to support his son's Paramount Skydance merger deal with Warner Brothers Discovery. And that made investors nervous. But back to the data centers. One reason is the sheer size of the debt. It's turned Oracle from a business that generates cash into one that burns money.

15:18And another is that the path back to profitability may take some time. And the third reason is that long path may not be certain. Oracle is relying on OpenAI a lot. Sam Altman and the folks at OpenAI promised Oracle they would pay about$300 billion over the next five or so years. Oracle's data centers help power ChatGPT. Right. So you can ask Jachibiti how to make like a weekend salmon recipe. And OpenAI's financials come with a lot of open questions, mostly because you probably can't see them. The company delayed going public and its books are not open. And Matt says if OpenAI's financials were out in the open, its value would also be based on its actual numbers, not just the story it tells investors.

16:03So if you can't really trust that OpenAI's revenues are going to be growing at parabolic speed, that trickles into Oracle that there is some concentration risk there. Oracle and OpenAI are not just in a customer-seller relationship. This bit has more implications for the wider economy. OpenAI pays Oracle to use its cloud infrastructure and says it'll pay a lot more in the coming years. So, Oracle buys chips from NVIDIA to build this cloud even more. And then NVIDIA invests money into OpenAI so that OpenAI can buy computing space from Oracle, which, remember, is buying those NVIDIA chips. I'm just picturing a circle.

16:43And according to credit investors, Oracle is a weak link in this circular economy. Oracle has already signaled that it may run into trouble paying back its loans on time. The company is currently building this massive data center in New Mexico for OpenAI to use. Oracle told its creditor that it may need to delay its loan repayments on this data center using something called a force majeure notice. And this is when a borrower tells its lender, hey, something totally unexpected happened. We can't make our delivery or payment. It's actually common in kind of the commodity space where you have forward contracts to deliver, say, oil.

17:24And if there's geopolitical risk like we're currently seeing with the Strait of Hormuz, I mean, that's outside of your control. It's less common, though, in corporate America. Oracle issued the notice because of potential delays, but publicly maintains that the project remains on schedule. Now, this notice created some alarming headlines, but Matt says it actually could benefit Oracle. It allows the company to delay payments, though it's still on the hook for the full amount. Still, Oracle's credit investors are pricing in more risk. And this is showing up in something called a credit default swap, or CDS.

17:59You might remember these from the Great Recession. Think of this like an insurance policy that Wall Street people trade back and forth. If the company goes bankrupt, the CDS contract will pay you out. But you can also use it to speculate. You can use it to profit on the demise of a company. The rates on these swaps are rising. This means that Oracle is becoming riskier. Matt says this does not mean the company is on the edge of bankruptcy or another immediate credit downgrade, though. It does mean the credit market believes that Oracle is growing increasingly risky. Right. S &P, the big credit rating agency, they downgraded Oracle's credit in July.

18:39S &P basically said, we'd like to see you make a profit again. Now, Oracle is barely above junk status, or a status that would force a lot of investors to run away from Oracle bonds. Lindsay Tyler is a credit research analyst at Morgan Stanley. She says to imagine data centers is a luxury apartment complex. Demand is strong now, but it may not be forever. The analogy might be you have 10 apartments, you've pre-signed leases for those 10 apartments, And you're thinking, should I build another floor? Right now, with S &P rating Oracle just one notch above junk bond status, falling beneath investment grade would have serious consequences.

19:18I think it's a medium term risk. But in that instance, Oracle would have a materially more challenging time not only funding the build out, but also refinancing. If Oracle goes to junk status, then many investors in the AI boom couldn't invest in one of its key players. For example, many pension funds can't invest in junk-rated bonds. Matt says these circular deals and heavy amounts of debt could lead to real problems for the economy as a whole. And that makes me concerned that this, if this is a bubble, it's going to be more like a Lehman Brothers type event. In retrospect, it's easy to say they were too interconnected to fail because it almost took down the global financial system.

20:01Is open AI systemically important? well, if they get in trouble and then that ripples through to Oracle and that ripples through to NVIDIA and then the whole ecosystem starts to collapse, that would be a serious problem. And they're all becoming very interconnected. For now, the wheel is still turning. And by the way, Amazon, Microsoft and Google are NPR sponsors. If you don't already give a follow to this show and to Planet Money's other podcast, The Indicator from Planet Money. It's one business or money topic from the news explained each day in 10 minutes or less. These episodes of The Indicator were produced by Corey Bridges and Julia Ritchie.

20:44They were engineered by Kwasi Lee and Sina Lafredo. They're fact-checked by Sierra Juarez. Kate Kincannon edits The Indicator. I'm Ricky Mulvey. This is NPR. Thanks for listening. Do people say junk bond anymore? I think wheeling turrets, high yield, leveraged credit, include loans in there. Stay tuned for the following announcements and messages from our sponsors.

21:30Visit thehartford.com today and see what's possible when belief is put into action.

22:00NPR for 20 % off your first purchase.

From the publisher
Your money might be fueling the AI boom … just as some AI data center bets are starting to look riskier. If you own stocks, that money is likely tied up in giant tech companies like Microsoft and Nvidia. And if your 401(k) has money invested in the US bond fund, you might be lending tech companies the money to build data centers. 

Today on the show, two stories from our short daily podcast The Indicator from Planet Money on how risky these data center bets are for investors and the companies making them. And, of course, for all of us. 

Read: 

Follow: 

The original Indicator episodes were produced by Corey Bridges and Julia Ritchey. They were engineered by Kwesi Lee and Cena Loffredo, fact checked by Sierra Juarez and edited by Kate Concannon. This episode of Planet Money was produced by Vito Emanuel and Emma Peaslee. Alex Goldmark is our executive producer.

Music: NPR Source Audio - "Saunter" and "Super-Fi" 

Support public media with NPR+ and enjoy perks for over 25 podcasts like this one. This show’s perks include bonus episodes and sponsor-free listening. Learn more at plus.npr.org.

See pcm.adswizz.com for information about our collection and use of personal data for sponsorship and to manage your podcast sponsorship preferences.

NPR Privacy Policy

More from Planet Money

All episodes