Is The AI Bubble Starting To Pop?

November 19, 2026

Joshua Scheer

The AI boom was sold as a technological revolution. It has also become a financing experiment. Chipmakers, cloud providers and AI startups are pouring money into data centers on the expectation that demand for artificial intelligence will eventually generate enough revenue to justify the cost. The chips, construction and electricity have to be paid for along the way. As Ben Norton argues in a new episode of Geopolitical Economy Report, that gap between what companies are spending today and what they expect to earn tomorrow is where the risk is piling up.

Norton follows the money through a network of companies that can be one another’s investors, suppliers and customers. A chipmaker can invest in a cloud provider that buys its chips, while an AI startup can raise capital from a company that also sells it computing power. Those relationships help finance the buildout, but they also tie companies’ fortunes together. Norton looks at the debt pressures facing Oracle and CoreWeave, the spending commitments of giants like Alphabet and Meta, and the challenge for OpenAI and Anthropic of turning demand for AI into a business that can sustain the cost of running it.

The politics of the buildout are playing out in communities asked to host these facilities. Norton points to opposition to new data centers across party lines, as residents weigh the promised economic benefits against demands on local water supplies and power grids. A company may see a new facility as essential to competing in AI. For the people who live nearby, the questions are more immediate. Who gets the electricity, who pays for grid upgrades, and how much say does the community have before a deal is struck?

The financial stakes reach people who have never bought an AI stock directly. As Norton explains, workers saving for retirement through broad U.S. index funds can still have substantial exposure to the handful of tech giants that carry much of the market’s weight. He also traces the exposure overseas through investors and semiconductor suppliers tied to U.S. spending. Drawing on parallels with the dot-com bust and the debt buildup before 2008, Norton asks how far a downturn could spread if a few of the companies financing the AI boom can no longer afford to keep it going.

Ben Norton’s Geopolitical Economy Report offers independent reporting and analysis on the U.S.-China chip fight, the rise of BRICS and how U.S. foreign policy shapes the global economy. The outlet is funded entirely by small contributions from readers and viewers, and you can support its work through Patreon, PayPal or a paid Substack subscription.

You can also make a donation to our PayPal or subscribe to our Patreon.

Please share this story and help us grow our network!

Is the AI Bubble Starting to Pop?

A transcript of Ben Norton’s commentary for Geopolitical Economy Report

Editor’s note: This transcript was prepared from the supplied captions and lightly edited for punctuation, spelling and readability. Section headings were added. The transcript preserves Norton’s analysis and the claims he makes in the segment. The claims and figures are presented as stated in the commentary. Brackets mark a timing reference that was unclear in the captions.

Ben Norton: Is the AI bubble in the United States starting to pop? Over the last few months, we have warned that regardless of what metric you use, the U.S. stock market is in one of the largest bubbles in history. And this affects not only many Americans who have their retirement funds invested in the stock market, but also people around the world, given that the U.S. stock market is very heavily concentrated in a small handful of big tech companies and these big tech companies operate all around the world.

And starting this October, we began to see reports in major financial media outlets warning that the AI bubble may be bursting. Fortune magazine noted that the AI bubble is cracking. Wall Street is nervous. CNN also published an interview with Gary Gensler, who was previously the chairman of the U.S. SEC, the Securities and Exchange Commission. He was also an investment banker at Goldman Sachs. And he warned that the AI bubble will likely pop in the upcoming months and cause an economic recession.

And this is not just empty rhetoric. We’ve seen that major investors have been reducing their holdings of U.S. big tech stocks and the Nasdaq index which is dominated by tech companies has fallen markedly in the past month. In fact, a survey conducted by Bank of America found that investment fund managers rank the AI bubble as the biggest threat to global markets, not just the U.S. market. And investors are reducing their exposure to U.S. tech stocks by the most in 20 years since 2006.

There are a few key reasons being cited to explain why this is happening.

The midterms and opposition to data centers

The first is the midterm election that was held in early November in the U.S.. The Democratic party delivered a massive blow to Donald Trump and the Republicans and following the election, U.S. stocks fell by $3.22 trillion in just one day. And this was in part because some major Democrats are call ing to rein in the buildout of AI infrastructure and call ing for more governmental regulation of the tech sector. Exit polls conducted on election day found that 76% of American voters say they oppose a new data center being built in their community.

And this includes 87% of Democratic voters who say they oppose a data center being built in their community and 62% of Republican voters oppose it along with 78% of self-declared independents. And it’s quite easy to explain why so many Americans are opposed to data centers. First of all, they pollute the environment around them and they’re very noisy. But also they use a lot of water and electricity and the costs of water and electricity in the U.S. have skyrocketed in the past few years since AI LLMs, large language models, started rolling out in 2022.

This has made the cost of living crisis even worse in the U.S.. So, even though Donald Trump has pledged that he’s going to veto any legislation introduced by the Democrats, the fact is that AI data centers are extremely unpopular among the majority of Americans. And the Democrats are going to try to push for regulation. So even if it doesn’t actually get passed by the Trump administration, the fact that this is such a chaotic and controversial issue is leading many investors to pull back from the big tech sector and AI infrastructure buildout.

So that’s one reason being cited to explain why we may be seeing the early stages of the popping of the AI bubble.

Nvidia and the technology stock sell-off

Another key factor was the ouster of Jensen Huang, the CEO of Nvidia, who was pushed out by China hawks in Washington. And this is extremely important because Nvidia is at the very center of the AI bubble. All of the other companies in the AI industry need Nvidia chips in order to operate their data centers. So the companies like OpenAI and Anthropic that are training AI models and the hyperscalers that are building out massive data centers like Amazon and Microsoft and Meta and Oracle, they all need lots and lots of Nvidia chips in those data centers that are then used to train AI models.

So if there’s instability at Nvidia now that Jensen Huang has been pushed out, this could cause ripple effects that could destabilize the entire industry. So those are two major factors that are now being cited to explain why the AI bubble may be popping or at least in the early stages of popping. However, I should point out that even before the midterm election and before Jensen Huang was pushed out of Nvidia this October, there were already reports being published warning that the stock market was already fall ing in October and there were fears of the AI bubble bursting.

In fact, if you look at the Nasdaq composite index, which largely consists of big tech companies, you can see that the peak of the Nasdaq was on the 6th of October. And since then, there has been a significant correction. And this is because beneath the surface, before the midterm election and before the ouster of Jensen Huang, there were already significant problems that were brewing. And we’ve seen some major financial institutions warning about these issues. For instance, the European Central Bank, the ECB, has warned that the massive binge of AI debt, these big tech companies have been taking on so much debt to build data centers and other AI-related infrastructure.

This binge of debt, this bubble has historic parallels with the subprime bubble that popped in 2008 according to the European Central Bank. So this is a major Western financial institution drawing direct parallels to what we’re seeing today and what happened in the lead-up to the great financial crisis of 2008 and 2009. So today we’re going to be looking in depth at the evidence explaining why the AI bubble may be popping. Of course, we can’t be certain, but a lot of evidence suggests that this could be the early stages of the popping of the AI bubble.

And as is always the case with bubbles throughout history, one of the key reasons for this is simply that so many of these companies have taken on so much debt in order to build out AI or whatever the new technology is in previous bubbles. We saw similar trends happen and we’re now getting at the point where earnings cannot justify that debt and there are fears that the entire bubble can burst. Yahoo Finance estimated that AI-related companies lost $5 trillion collectively in their market capitalization this October.

Now the U.S. stock market as a whole did not fall very significantly largely because many investors rotated out of tech stocks and into other stocks. But if you look at some major tech companies, including the biggest AI companies and chip companies, you can see that they lost trillions of dollars in market capitalization in the past several weeks.

AI spending, cash flow and borrowing costs

Today, we’re going to explain why investors have been selling tech stocks and why there are fears that one, AI may not prove to be as profitable as many investors assumed. Two, reports show that AI has not led to significant productivity increases. And three, many of these companies have invested so much in AI buildout that they simply don’t have any cash and are now taking on more and more debt. And investors fear they could not pay that debt off. And by the way, we can tell this is all AI-related because there is one major big tech company that has actually been outperforming the market at a time when all of the other big tech companies have been underperforming the market.

And that is Apple. And what is unique about Apple is that it has not been spending very much on AI capital expenditure capex. It has largely been sitting out of the AI arms race while all of the other Silicon Valley giants have been spending hundreds of billions of dollars per year to build out data centers and other AI-related infrastructure. And by the way, this is why Apple has now overtaken Nvidia to become the most valuable company on Earth by market capitalization with a market cap of more than $5 trillion.

Investors have rewarded the company for not going all in on AI unlike all of its other competitors. So one of the ways you can measure this is looking at the free cash flow of a company. As the name suggests, it’s a very simple concept. It is the cash that a company has left over after paying all of its expenses and capital expenditure. So if you look at Alphabet, which is the parent company of Google, you can see that for the past 20 years, this corporate monopoly has been printing cash, which makes sense.

It has basically a monopoly over search on the internet. It had so much cash that it was sitting on, but in the past several months, suddenly Alphabet’s free cash flow has gone extremely negative. The same thing has been happening to Meta, which is the parent company of Facebook, Instagram, and WhatsApp. For the past two decades, Meta was printing cash. But for the first time in history, Meta’s free cash flow is now negative because it’s spending so much on AI capex. And you can see the same trend if you look at the combined cash flow of all of the hyperscalers.

These are the major big tech companies that have created enormous cloud computing platforms. If you look at the free cash flow of Amazon Web Services, Google, Meta, Microsoft Azure and Oracle, together their free cash flow has gone negative and this is why these hyperscalers have been taking on enormous sums of debt which is quite new for the sector. If you look at a chart, you see that in 2026, the amount of debt exploded. And this comes at a time when paying that debt is getting more and more expensive because interest rates have been rising.

And in particular, yields on U.S. Treasury securities, U.S. government debt, have also been rising, which means there’s a kind of crowding out effect because essentially these big tech corporations are competing with the U.S. government in trying to attract investors to buy their debt, which continues to push up borrowing costs more and more.

Valuations and circular financing

So what all of this shows is that slowly the mania in the U.S. stock market has worn off and people are starting to see the reality which is that we are in the largest bubble in history. The Wall Street Journal referred to the AI buildout as the biggest economic bet in U.S. history. It noted that spending on AI-related infrastructure is larger than the spending on the infrastructure related to all other previous technologies that caused massive bubbles including canals, railroads, electrification, highways, and the telecommunications and fiber optic infrastructure during the dot-com bubble, the internet boom in the late ’90s which popped in 2000, which is quite similar to the AI bubble, much like what we’re seeing today.

Now, I’m not disputing that AI does have some practical uses like the internet and railroads and electricity. These were revolutionary technologies. But all of these previous revolutionary technologies also caused bubbles just like the bubble we see today. And regardless of what metric you use to try to measure how overvalued U.S. stocks are, you can see that as Bloomberg put it, U.S. stock valuations have never been higher. If you take an average of many different metrics, like for instance, several different ways to measure the P/E ratio, the price-to-earnings ratio, or market capitalization of the stock market as a percentage of GDP.

If you average that all together, the AI bubble is larger than the dot-com bubble and it’s larger than the bubble that peaked in 1929 that later fueled the Great Depression in the 1930s. And what’s especially dangerous about this bubble in the U.S. stock market is the unprecedented degree of concentration. The Wall Street Journal reported this October that two companies, AI chip makers, Nvidia and Micron, made up 30% of the earnings boom in the U.S. stock market. And if you take just 10 AI companies, they made up 2/3 of corporate profit growth.

And what’s so dangerous about this concentration risk is that so much depends on these big tech companies that are all paying each other or promising to pay each other and in many cases not actually delivering through a web of circular financing. And at the middle of this complicated web is Nvidia, which had previously been the most valuable corporation on Earth by market capitalization, although it was recently overtaken by Apple in this correction this October. And so much depends on Nvidia because Nvidia is designing the AI chips that are used by all of the other AI companies.

But if the other AI companies like OpenAI and Anthropic and Microsoft and Intel, if they cannot pay, they signed all of these deals but have not actually delivered. If they can’t pay, then there may be a cascading series of defaults that could cause a larger financial crisis.

Neoclouds and the risk of default

And some prominent financial analysts fear that this is exactly what could happen. And at the heart of this problem is a new kind of company which is called a neocloud. The neocloud companies have been providing the cloud services at data centers for many of the AI companies. The AI companies contract with the neoclouds. Now, I’m no fan whatsoever of Elon Musk, to put it very mildly. This is the world’s richest oligarch who’s been spending hundreds of millions of dollars to manipulate U.S. elections.

But regardless of what I think about him, he has been warning, and of course, given he’s the world’s richest oligarch, he’s very influential. He warned that more than 200 neocloud companies may face default and bankruptcy and consolidation. We may see some larger companies buy them out potentially the massive hyperscalers. And the perfect example of this is the company CoreWeave which provides cloud services for AI companies. We’ve seen that in the past six months CoreWeave stock has tanked by more than 50%.

And the financial press has been warning that CoreWeave faces very significant credit risks and investors have priced in more than 60% odds that it will default. And if CoreWeave does default on its debt, this could cause a cascading series of crises in the stock market and potentially the larger U.S. economy and the global economy as I’ll talk about later. Bloomberg published a very important article titled CoreWeave rout sends warning sign to rest of the stock market.

Bloomberg quoted a financial analyst who said, “If the AI bubble is going to burst, it’ll likely start as a credit event as neocloud and risky AI debt defaults creating a cascade of devaluations, reduced spending and contraction in multiples and growth.” And the reason this is so dangerous goes back to this chart I looked at earlier showing the circular financing scheme of the companies making up the AI bubble. And you can see that at the heart of this is Nvidia because Nvidia wants to sell its AI chips to all of these companies.

CoreWeave, for instance, provides these cloud computing services for AI companies and it buys tons of Nvidia chips for its data centers. But if CoreWeave goes bankrupt, it’s no longer going to be able to buy more Nvidia chips and then other companies may stop buying Nvidia chips. And then Nvidia’s earnings fall. And you can see a cascading series of crises happening here. And by the way, it’s clear that Nvidia is worried about this because the CEO of Nvidia, Jensen Huang, is going out publicly and saying, “There’s nothing to see here.

Don’t worry. ” He insisted that neocloud demand continues to be “off the charts. ” And what’s even more ridiculous is that Jensen Huang has been publicly defending CoreWeave and saying don’t worry we know that there are all these signs that this company probably cannot service its debt. But he said don’t worry I’m optimistic. Okay. But in order to be optimistic you have to ignore the data. You have to ignore the evidence. And one of the concerning pieces of data that we’ve seen emerge in recent months is the rising spreads on credit default swaps, CDSs.

What does that mean in finance jargon? This is a kind of insurance policy. If you buy a bunch of corporate bonds from a company like CoreWeave or Amazon or Oracle and you’re afraid that the company could default and not pay you back as the bond holder, what you can do is you can pay for a CDS, a credit default swap, which is an insurance policy with the financial institution. And the way it works is the bank or insurance company, the seller of the CDS promises to pay you the face value of the corporate bond if the company defaults.

And in return, you have to pay what’s known as a spread or just simply a fee, a premium to the seller of the CDS. This is your insurance policy. So as the spread increases, what that means is that you have to pay more and more in order to maintain the insurance policy. Your insurance gets more expensive because the risk of default has increased. And if you look at some major companies, especially Oracle, you can see that the CDS spreads have risen by several hundred basis points in the past few months.

This is a major warning sign that investors fear that companies like Oracle may default on their debt. In fact, Bloomberg warned that Oracle’s credit risk has hit a near [time period unclear] high because it’s been taking on so much debt to build these data centers for AI. And by the way, if Oracle faces some serious problems and can’t pay its debt, it’s quite possible the Trump administration may try to bail out this company because the founder of Oracle is the billionaire oligarch Larry Ellison who is a major supporter of Trump.

In fact, a Trump adviser referred to Larry Ellison as the shadow president of the United States. And the U.S. government forced the Chinese parent company of TikTok, which is called ByteDance, to sell control over the TikTok algorithm in the U.S. to Larry Ellison’s company, Oracle. So politically, there is a lot invested here.

Investors pull back as AI companies look to go public

But anyway, getting back to the AI bubble, the point that I need to stress here is that there is more and more evidence showing that we could see the beginning of the popping of the AI bubble. Everywhere you look, you see more and more signs. For instance, Goldman Sachs, the major U.S. investment bank, published a report noting that hedge funds have been dumping tech stocks at the fastest pace in a decade. Moreover, some major pension funds are now following the hedge funds.

CalPERS is California’s retirement fund. It’s the largest pension fund in the U.S. and it has cut its exposure to the Magnificent 7, the seven biggest tech companies by 26%. It cut its exposure. And that’s not all. Another major warning sign that’s flashing red is that even OpenAI is now having trouble raising money. And OpenAI was supposedly, we were told, going to be the most profitable company in human history that’s going to bring us to the technological singularity and AGI and all of that propagandistic nonsense.

Well, late October, the Wall Street Journal reported that OpenAI was not even able to raise half of its $30 billion target in its latest round of funding. And this is because many investors are becoming quite skeptical and recognizing that this AI sector is in a bubble. OpenAI is losing billions of dollars every year. They’re probably not going to pay back their investors. Investors will lose money. So people are no longer buying in. And this comes at a time when the two biggest U.S. AI companies, OpenAI and Anthropic, are preparing for massive IPOs, initial public offerings, in which they will sell stock in their company to the public.

At the same time, Anthropic recently published its S-1, which is the document it has to file with the Securities and Exchange Commission, the SEC, in order to prepare for its IPO. And Anthropic acknowledged that it had a net loss of more than $80 billion in the first half of 2026. Now, that very important data point was often not mentioned by many of the AI bulls talking about how great Anthropic is and this amazing S-1 report that warns that AGI could lead to the end of civilization and destroy the world and all of this nonsense.

Meanwhile, if you actually look at the boring financial data, you can see that yes, Anthropic’s revenue increased. And all the bulls said Anthropic’s revenue grew by so much, it’s amazing. Okay. Anthropic’s revenue in the first half of 2026 rose to $16.29 billion. Okay. But its net loss was $80.75 billion. So its revenue did increase but its loss increased by so much more. Moreover, if you closely read Anthropic’s S1, you can see that the company acknowledged that its recent growth “may not be sustainable over time.

” And its growth is already slowing by the way. But what’s so ridiculous is that Anthropic is still estimated to be worth $1.5 trillion or even two trillion dollars when it launches this IPO. And some financial analysts have pointed out that three AI-related IPOs, SpaceX, Elon Musk’s company that already had its IPO earlier this year, Anthropic, which is planning very soon, and OpenAI, which plans on having its IPO soon, although it delayed it. Together, these three companies are estimated at more than $5 trillion, which is more than all of the other tech-related IPOs of big tech companies in the U.S. in the past 45 years since 1980, which is outrageous.

It’s so ridiculous when you think about this because all those other big tech IPOs included the IPOs of all the other giants, Amazon, Microsoft, Google, we’re supposed to believe that OpenAI, Anthropic, and SpaceX will be more valuable, more important than those other corporate monopolies. I mean, it’s so ridiculous. But what’s sad is that a lot of people believe the hype, especially a lot of unsophisticated investors, retail investors, average people, they believe the propaganda nonsense and they buy into the IPO and they end up losing tons of money.

For instance, if you look at the SpaceX IPO in June of this year, there was a massive rally for the first few days and then there was a massive decline. So, if you had bought into the SpaceX IPO, you almost certainly lost money. It has been a horrible investment despite the fact that Elon Musk claimed in his S1, the filing with the SEC for the SpaceX IPO, that the size of the potential market for SpaceX is larger than the entire global economy because he wants to like recreate the Spacing Guild from Dune or something and have a monopoly on space travel and trade.

I mean, it’s so ridiculous. By the way, SpaceX is also a company that’s losing a lot of money. But the ridiculous valuations of these companies in these IPOs is yet another sign of the frothiness of this enormous bubble in the AI sector. And some financial analysts have pointed out a scary little factoid. The last time the volume of IPOs in the U.S. was this high were 1929 and 2000. So in the massive bubble that popped and fueled the Great Depression in the ’30s and during the dot-com bubble.

Lessons from the dot-com bubble

Now to be clear I don’t think that when the AI bubble pops we will see a Great Depression like the ’30s that was caused by many other factors not by just the crisis in the stock market. But I do think that there are so many parallels between what’s happening today and what happened in the late ’90s in the dot-com bubble. And it’s important to study the parallels between the dot-com bubble and what’s happening today. For instance, if you look at the IPOs of major tech companies that took place in the late ’90s during the dot-com bubble, you can see that the stock performance of many of them was horrible in the subsequent decades.

A perfect example of this, by the way, is Intel. Someone on Twitter pointed out that if you bought $10,000 of stock in Intel at the peak of the dot-com bubble in 2000, today it would be worth $9,000. And Intel recently, in the past few months, saw another massive increase in its stock price. And this is largely because Donald Trump bought a U.S. government stake in Intel. But in the past few weeks, we’ve seen a major correction and its stock price has fallen quite a bit.

Retirement savings and global exposure

However, with all of that said, there may be some people listening to me here and saying, “Ben, why are you talking so much about the U.S. stock market? I’m not rich. I’m not invested in the stock market. ” Well, the reality is that this impacts many more people than you might think, including many average people and not only in the United States, but around the world. Let’s start with the U.S., for instance. Most Americans don’t realize that basically all of their retirement savings, if they have them, are tied up in the stock market.

Especially if you have a 401(k), you are invested in the stock market, which means that you are heavily invested in the AI bubble. Even if you just put your retirement savings every month or whenever in the S&P 500, the main stock market index of the 500 largest publicly traded companies in the U.S.. That is so concentrated in a small handful of big tech stocks. So when the bubble pops, this is going to lead to a massive correction that could hurt the retirement savings of many Americans.

But it’s not just Americans who will be affected. I mentioned that this will also have a massive impact on many other economies around the world. And Bloomberg published a very interesting report titled AI’s giant money loop is drawing foreign cash like 2008 and you do not want to hear parallels to 2008 which of course was the great financial crisis. In this report, Bloomberg pointed out that the AI bubble in the U.S. is actually a global AI bubble and there are many other economies involved in this AI spending spree.

And there has been a massive influx of foreign capital into the United States chasing returns. And when the bubble pops, this will also hurt investors around the world. This is what Bloomberg wrote: “It’s a circular flow of capital that’s contributing to higher yields on U.S. Treasuries and sovereign bonds, raising the cost of servicing debts globally. It spreads the risks of an AI bust beyond Silicon Valley to investors in Asia and beyond. ” And I talked about some of the circular financing happening among U.S. companies, but there are also foreign companies involved, especially South Korean companies.

South Korea is a powerhouse when it comes to the semiconductor industry with two major chip companies, Samsung and SK Hynix. And by the way, in the past year, we saw one of the largest bubbles in the history of the South Korean stock market, which massively inflated over late 2025 and early 2026 and then popped and we’ve seen a significant correction in the South Korean stock market. And this has really hurt a lot of people who have lost their life savings in this enormous bubble.

So so much in the global economy, not just in the U.S. economy, is depending on the continued capital expenditure of these AI companies building out more and more data centers and chips and other equipment.

The revenue gap and the payoff from AI

And Bloomberg warned that in the United States alone, AI faces a $4.9 trillion revenue gap. The companies building out data centers need $8.5 trillion in revenue through 2030 to cover their investment costs, but they’re only estimated to get revenue of $3.6 trillion. So the math simply does not add up. Maybe we’re seeing the beginning of the popping of the bubble. No one knows. But when it pops, it’s going to have very severe consequences. And it’s becoming clearer and clearer that we may be coming to the end of the AI bubble.

A great example of this is a report that was published in late October by Apollo Global Management. Apollo is one of the major asset management firms. It runs one of the world’s largest private equity funds. And it published a report titled, “No signs of AI payoff outside the tech sector,” noting that AI “has delivered zero financial payoff to corporate America outside of tech.” “There are still no signs of AI boosting productivity growth. With no impact in the bottom line, mainstream corporate America is pulling back and starting to view token spending not as a source of growth, but as a cost center that weighs on margins.” “Mainstream corporate America does not need the latest model and is unwilling to pay for it.”

Here are some of the very revealing charts in this report. It notes that revenue per worker is increasing for U.S. tech companies, but not for the rest of the economy. Moreover, the profit margins of these massive big tech companies in the Magnificent 7, their profit margins have gone up significantly, but outside the big tech sector, there has not been a significant increase in profit margins. Similarly, Apollo showed that there are no signs of AI use boosting productivity growth in non-tech companies.

And finally, AI spending is extremely concentrated in a small number of companies. This is the record level of concentration that I warned about. And if one of these major companies like Oracle defaults, can’t pay its debt, goes under, we could see a cascading series of crises.

Interest rates and the timing of a possible bust

And then finally, another major factor to consider in all of this is that the U.S. central bank, the Federal Reserve has been raising interest rates. And it’s expected that the Fed will raise rates at least once, maybe two more times. And the reason this is relevant here is because as the Fed raises interest rates, what that really does is it makes the cost of borrowing more expensive. And these big tech companies are now taking on more and more debt. Earlier I discussed how all of the hyperscalers have burnt through all of their cash and are taking on hundreds of billions of dollars of debt per year.

So they are now competing with the U.S. government more and more with U.S. Treasuries and the bond market for investor capital. And the reason that this is noteworthy is because if you look back at the dot-com bubble, you can see that before the dot-com bubble popped, there was a very similar trend. The Fed was raising interest rates in 1999 and 2000. And this is one of the factors that contributed to the popping of the dot-com bubble among other factors. By the way, we also saw something very similar in the lead-up to the great financial crisis of 2008 and 2009 when the real estate bubble popped.

The Fed was gradually raising interest rates in 2005 and 2006 before the real estate bubble peaked in 2007. So there are certainly parallels here. And you should keep in mind that when these bubbles in the stock market pop, it often happens over a period of months or even years. It doesn’t just happen overnight. The fact of the matter is that no one knows exactly when the bubble will pop. We can say with certainty that it will pop. All bubbles in the past have popped. We don’t know when.

And as the renowned economist John Maynard Keynes put it, that’s because the markets are moved by the animal spirits, not by reason. Or as Keynes also put it pithily, markets can remain irrational longer than you can remain solvent. I’m Ben Norton, the editor of Geopolitical Economy Report. Please like and subscribe. I’ll see you next time.

Subscribe
Notify of

0 Comments
Most Voted
Newest Oldest