The Information Finance

It’s too Soon to Call an End to the AI Boom

Art by Clark Miller
By
Guest

I’m the senior finance editor here at The Information. I’ll be writing this new column on money, dealmaking and bubbles of all sorts every Wednesday.  If you have any feedback or suggestions on topics I should explore, feel free to reply directly with your feedback.

We’re launching this newsletter to cover the increasing role finance is playing in tech. The development of AI will demand $5 trillion or more in funding, according to JPMorgan Chase. The quest for cash has set off a global scavenger hunt among the tech giants and AI wannabes. That’s pushing the tech industry into the complicated world of bonds (more than $200 billion raised this year for AI) and structured finance. The idea of raising just millions feels quaint.

The Information has expanded its finance coverage to meet this moment. We’ve added new finance reporters covering lending—bonds, private credit and structured finance—and we’re tracking cash coming from the world’s biggest pools of money: sovereign wealth funds, private equity and venture capital firms, and global pension giants. We’re watching the balance sheets of the tech giants, which so far have been the biggest source of cash for AI. We’ve also boosted our coverage of mergers and acquisitions, with a focus on the impact of AI on industries.

This column will highlight some of our best work, and I will give you my take on the most important financial stories driving tech. I haven’t seen it all, but I’ve come pretty close. As a reporter, I rode the tech bubble up and down, writing detailed dissections of some of the biggest blowups, including Nortel, Enron and Lucent. Later, I helped lead The Wall Street Journal’s coverage of the global financial crisis, digging into the ridiculously complicated financial arrangements that nearly brought down the financial system. 

I admit, I’m feeling echoes of previous bubbles. I did a double take when I saw the name of Magnetar listed as a big funder for CoreWeave. The hedge fund with the astronomical name won big in the housing bust by betting on dodgy mortgage loans. I’ll be writing about the current boom, bubble or not, and highlighting the risks and rewards at play.

I’ve spent the past few weeks talking to the bankers and investors leading the financing of AI. I’m convinced a crack in the market isn’t coming anytime soon. Investor demand is very strong, and it’s too soon for any real problems in the financing machine to show up.

That doesn’t mean the path will be smooth, as the current market tumult shows. I think the easy money has been raised. There are not many SoftBanks or Middle East sovereign wealth funds that can write multibillion-dollar checks at a moment’s notice. The tech giants are already levering up and growing more protective of their cash piles. 

There’s a lot more money to be raised for AI. Bankers and their clients are coming up with new strategies to fill that need. A prime example is the deal by Meta Platforms to fund its Beignet data center in Louisiana. As our Miles Kruppa wrote, the arrangement was tailored to meet the needs of Meta, which wanted to keep its balance sheet clean, as well as of investors who wanted slightly higher yields without taking additional credit risk.

Demand was so strong for the bonds that Morgan Stanley, which ginned up the structure, was able to sell $27.3 billion in bonds to a select few big investors. A higher yield helped. The only grumbling I heard about the deal came from people who didn’t get a piece of it.

What’s become clear is that a lot of capital can be brought to bear in deals backed by trillion-dollar companies. Smaller, more speculative companies are already facing pushback from investors. That is playing out in the stock market, which might feel ugly but is only down about 5%, and the bond market, where yields are up a bit but only after a big rally. 

New financing deals are still coming every week, and investors are gobbling them up. Amazon this week sold $15 billion in bonds in an offering it upsized from $12 billion due to high demand. When the ultimate backstop is anyone but a tech giant—such as Oracle, OpenAI and xAI, bankers are finding ways to make them appeal to investors. Bankers can be surprisingly creative when there are fees on the line. 

Look at it from the perspective of a fund manager with cash on hand and clients who expect top-notch returns. If your competitors are diving into AI, you can't risk falling behind them. Underperforming for a few months is OK, but doing so for a few years can shorten a career. 

It’s hard to argue that there will be a bust anytime soon. Even if AI growth rates slow, most companies have the financial wherewithal to make payments for the next few years. And if one of the riskier plays stumbles, one of the healthy companies would likely snap up the data centers it was trying to build. 

There are a few risks. The most basic is higher long-term interest rates, which seem increasingly possible given rising inflation. No amount of screaming by President Donald Trump will bring down the yield on the 10-year Treasury bond. Another is execution risk. Investors are closely watching the giant Stargate data center project backed by Oracle and OpenAI for delays and snafus.

Even seemingly real threats to the AI buildout may turn out to spur more investment. Earlier this year, Chinese AI company DeepSeek scared the market when it announced a top AI model that it trained using a fraction of the computing power demanded by U.S. developers. 

That could have made investors wary of funding the AI buildout. Instead, it raised a more important question: whether China would take the lead in AI. Whatever you think about a government backstop for AI lending, fear that the U.S. is falling behind would create a strong incentive for the government to step in. The Trump administration hasn’t hesitated to spend its money on what it deems to be in the national interest. 

None of this means there won’t be a spectacular bust later this decade. Part of our job at The Information is to figure how and when that would happen. I’ve gone on about debt because that’s where big crises usually start. My experience tells me there are concentrations of risk lurking in the shadows of finance. That could be where the next crisis starts. We’re happy to take nominations.