Michael Burry Just Bet on a $3 Trillion Financial Time Bomb

So, this report just came out from the Wall Street  Journal last month, showing AI spending on chips   and data centers is far higher than investors  have been told. It points to $3 trillion in hidden   spending being fueled by unsustainable debt, which  calls Nvidia's $5 trillion rise in recent years,   or Palunteer's $2500% rise in the last 3 years  into question. If these valuations have been   built on faulty data that doesn't tell the  whole story, what happens when these bills   come due? But investor Michael Bur has gone one  step further, claiming AI is a bubble that could   bring down the rest of the economy when it  implodes. And Bur isn't just talking here.   He is actively positioning his portfolio for  a major market downturn in AI. Explaining his   trades in a recent post, saying, "If this looks  like preparation for a larger fall in the market,   that is because it is." So, Michael Bur became  famous for predicting the housing bubble before   the whole market collapsed in 2008, profiting  over $700 million for his investors and over $100   million personally off that one trade. He's also  made big profits shorting other large tech runs,   including a decent profit shorting Tesla recently.  But he's also made several negative predictions   that never actually came true. This chart  lists a bunch of them. But after stepping   away from the public markets, deleting his Twitter  completely, Bur has now come back with a private   Substack where he is showing everything that he  is buying and selling and sharing some charts that   show us just how deep the AI debt worries may run  because even if you don't trust Bur's conclusions,   the data that he is showing is worth understanding  for any investor. So, let's just start by what is   he actually betting against. So in August, Bur  shared that he had added to his short positions   against the big AI players, Micron, Nbius, Nvidia,  Oracle, and Palanteer. But his biggest bet right   now is against SOXX, a semiconductor ETF. So  instead of betting against a single chip company,   he is betting against most of the chip industry  all at once, an industry that is up over 100%   in the past year. On top of this, he holds a  large negative bet against QQQ, an ETF of the   100 biggest companies in the NASDAQ, which today  is dominated by big tech and AI companies. Now,   we don't know the exact size of those trades.  Bur no longer reports to the SEC after he closed   his fund at the end of last year. So, these  estimates aren't perfect, but they do show us   that Bur is betting against several steps in the  AI supply chain. Now, in addition to these shorts,   Bur does also hold several long positions,  which I covered in a separate video where I   looked at what some of the biggest investors were  putting money into right now. But for this video,   I want to laser focus on why Bur is betting so  heavily against AI and specifically semiconductor   stocks. In recent months, many people have called  Bur a broken clock who is right twice a day.   He claims he's usually just early, but I want to  look at the data and the three biggest reasons   behind his bets so we don't just have to trust Bur  we can decide for ourselves. And we'll also look   at what investors can do to not just survive, but  actually profit from what is happening right now   in the AI market. So, this all kicked off last  month when the Wall Street Journal published   this deep dive into the massive amount of spending  that AI companies have been pouring into hardware   and data centers. Now, we already knew that around  $600 billion will be spent in 2026 alone. But this   investigation found that there may be a lot more  money committed to these data centers than meets   the eye. Money that isn't factored into these  stocks earnings and thus their stock prices.   And the key here is what's known as offbalance  sheet liabilities, where companies have promised   to pay money in the future and they may even have  a contract, but they haven't actually spent the   money or listed it as debt on their balance  sheet. Now, to be clear, this isn't some huge   secret. I mean, it's listed right here. This is  a footnote from Google's latest 10Q report. So,   it's there, but they don't make it obvious. Or  take Meta for example. They're now building their   Hyperion data center in Louisiana. But neither  Hyperion nor the $27 billion in debt shows up on   Meta's balance sheet. Instead, that debt is owned  by Blue Owl Capital, which Meta partners with,   slash is also a tenant. But even that is not a  true secret. It's not like they're fraudulently   hiding these loans, but it also does not show  up on third party websites like Google Finance   or in most investing apps. So, if you look up  their leases or their debt ratios, these numbers   are not factored in, which is a big deal because  some of these are like 25-year lease commitments,   which even if there's no revenue coming in for  AI in the future, they are guaranteed to have   to make those payments going forward. And Bur has  a very strong opinion on all this, saying, "The   fish have gorged themselves on off-balance sheet  liabilities, backstocks, unccommenced leases,   purchase agreements, all ways to spend future  money without having to list it as a liability.   And it's $3 trillion in future promises. So more  than is actually listed on the balance sheets. And   all that spending does not show up in the balance  sheet or income statements for these companies.   Now, to be fair, this isn't money that's  immediately leaving the company. Alphabet had $811   billion in future promises, but that includes  things like 25-year leases on data centers,   which obviously isn't paid all at once. But there  are other cases that look more concerning. Nvidia   just committed $15 billion to back financing  for Open AI. Kind of like a parent co-signing a   kid's first car loan. Only this backing is worth  three times more than Nvidia's entire publicly   disclosed debt levels today. So, you better hope  it's apparent that OpenAI doesn't crash their car   because Nvidia's investors are going to be on the  hook for that loan either way. And this is where   we get this famous chart where it feels like  AI is a circular economy that's just propping   itself up with loans and borrowing and deals. But  that takes us to Bur's second big worry about the   AI market. And this is more useful to investors  because it reveals the areas of the market that   are the most vulnerable to what is happening.  the market is not all equal. And after that,   we'll look at why Bur is worried about this  spreading into a broader market downturn and what   we can do about it as investors. One thing you'll  notice about Michael Bur is he focuses on using   his money effectively. For me, that usually means  finding smart ways to make my income go further,   like using tax advantaged accounts like an FSA, a  flexible spending account for tax-free healthcare   dollars. And this video sponsor, FSA Store®, helps  you get the most out of your funds. FSA Store® is a   curated store to buy products that qualify for FSA  spending, over 2500 of them. So, if your income   tax is 30%, that's like saving 30% on eligible  expenses. And we're getting into the end of the   year, so if your FSA dollars might expire, you  might as well use them instead of letting those   pre-tax dollars go to waste. But what's great is  FSA Store® isn't limited to medications. They've   got sunscreen, allergy meds, even pain relieving  foot massagers. FSA Store® sent me this Caring Mill™   by Aura Migraine Relief Mask. It's pretty great  after staring at a screen for hours or during the   afternoon slump. It's like a 15-minute powernet  plus massage, heat, and vibration. It even plays   music. But the best part, it's FSA eligible. So,  you can buy this and everything else on the site   using your FSA card. No submitting receipts. Just  go to fsastore.com/fintech or scan the QR code.   So, if you have FSA dollars, they're taxfree and  they could expire at the end of the year. So, it's   worth checking out their collection of products.  Visit fsastore.com/fintech or scan the QR code   to start shopping. So, thanks to FSA Store®, where  everything is eligible for sponsoring this video.   Smarter saving is smarter investing. And now,  let's look at the second reason for Michael   Bur's bet against the AI market. So, you have all  of these offbalance sheet liabilities, but those   by themselves do not cause an AI meltdown, but  they do introduce fewer points of failure in the   market. So Bur's other big fear is how much  actual debt is being fed into a system that   is already less stable than it should be. Because  it's one thing if you have $3 trillion promised on   future profits, but today much of big tech's cash  flow has already disappeared into hardware costs   they can't afford to spend anymore. Meaning that  $3 trillion time bomb has to be funded by debt.   In August 2026, Nvidia announced a partnership  with six investment and private equity firms   with the goal of raising more than $500 billion  for AI buildings and equipment. That is a target,   not money sitting in a bank account today. But the  goal is to get outside investors to start paying   and writing loans directly for this buildout.  So this now adds an extra layer to that chart.   Investors and lenders can now give money to a data  center project. that project can then buy Nvidia   systems. Then Open AAI or another AI company can  promise to rent the computer power. So now Nvidia   can record a chip sale today even though the final  money to pay for the whole system must come from   years of future rent. And we can see how this  whole situation looks a little dangerous just   in the latest Nvidia OpenAI deal as an example.  So let's break this deal down. We have SB Energy   which is building a data center campus in Ohio.  Open AI has agreed to rent space and power there.   Pretty normal business things. So, lenders are  willing to give money to SB Energy to fund the   construction of the buildings because they expect  those rent payments to come in. But Nvidia has   now had to step in and promise to cover some  of OpenAI's rent and power payments if OpenAI   doesn't pay up to that $ 105 billion number across  the nine different sites they're building. And in   return, OpenAI promises to only use Nvidia's chips  in those data centers. Then on top of all of that,   every company in this deal seems to own a piece of  every other company with the investment firm Soft   Bank owning both SB Energy and a big chunk of Open  AI. That is the circle that Bur is worried about.   Money flowing through a system is good. That's  just how an economy works. But in this case,   it's a series of contracts all built onto each  other, like stacking blocks higher and higher.   And ultimately, the only money flowing into this  system is going to have to come from OpenAI or the   end AI users. But right now, OpenAI doesn't make  a profit, and they won't anytime soon. So, they   need investors and lenders to pay for all this.  And that's basically the bet to keep this whole   thing building higher and higher long enough that  they can become profitable because if they don't,   there's nothing holding it up. And we've seen even  some traditional banks make some interesting moves   here. The banks that loaned money for an Oracle  data center, for example, are actually selling   off those loans to private credit, an area of  the market that already looks vulnerable to   any outside shocks. And Michael Bur summarizes  all this by saying that the increasing capital   that all these companies are spending is coming  from debt. And this puts the AI boom on a clock   with a countdown because every loan has interest  and every loan must be paid back eventually. It's   not that these companies are putting profits into  this hardware, they're borrowing for it. And Bur   thinks this countdown timer will hit sometime  in 2027 and certainly before 2028, but recent   moves from the Federal Reserve may move that  clock even earlier. Barkclay's Bank now expects   the Fed to raise interest rates for the next two  quarters, meaning all these companies are going   to be paying even more money in interest, and by  2027, we're going to see increasing pressure on   the hundreds of billions of dollars in debt that  is funding this buildout. Now, this pressure won't   hit everywhere all at once. The AI market is  not one uniform thing. So, let's look at the   parts of the market Bur is most worried about, as  well as the specific warning signs that investors   can watch out for. So, Bur's biggest bets are  focused on the hardware players in the AI market,   especially the NeoClouds like Coreweave and  Nebius, each of which are already paying hundreds   of millions of dollars in interest for the debt  that they've taken on to grow, which has led Bur   to say about Nbius that this is what the top of a  boom looks like. Now, to be fair, these are also   just aggressively growing companies, and taking  on more debt can help you grow faster. They also   have real revenue and real paying customers. But  if there is an AI slowdown or an AI credit crunch,   who's going to have more leverage to negotiate  contracts and loans, Google and Nvidia or these   smaller Neocloud providers? The issue here is that  these smaller players need two things to be true.   One, they need their customers to continue signing  more and more contracts. and two, they need their   lenders to continue offering more and more debt  to fund it. So far, that hasn't been an issue.   Both of these pillars look pretty strong, but they  do point to three warning signs that we can look   for. So, even if we don't trust Bur's total market  crash prediction, we have smaller pieces of data   that we can look at and act on. So, number one,  if customers don't use these companies compute,   if they have excess capacity, then the growth  story here might be over. Second warning sign,   if they can't find new lenders for their loans.  And third, if we see a rapid rise in interest   rates that makes their loans too expensive for  these companies to cover. Any of these could   be a reason to sell out of these stocks because  they have to balance supply and demand with the   rates from lenders and keeping those in balance  is a tightroppe walk. But there's a third reason   Bur is worried about the market and it goes beyond  semiconductor stocks. It's the risk that any major   fall in AI could spread to the rest of the market  and cause chaos to use Bur's word. And after that,   we'll look at three specific things that  investors can do to protect ourselves if   Bur's prediction comes true. Right now, AI stocks  make up nearly onethird of the S&P 500, the 500   biggest companies in the US. So, a major AI crash  is already going to hit retirement counts of most   long-term buy and hold investors who are holding  index funds with Bur even warning back in 2022   that most of the market's money is just passively  invested. Meaning, as all the money flows into the   biggest stocks in the market, that props up any  bubbles that appear. But there's another concern   which is how much of the market is propped up on  investors borrowing margin debt where investors   borrow money specifically to invest in a market  that seemingly keeps going up. This is a chart of   all the customer margin debt in the United States.  It shows debt reaching a record $1.5 trillion in   June 2026 and currently sitting around 38% higher  than last year. And for retail investors like you   and me, that rate of debt growth is way higher.  Robin Hood reported its margin loans crossed   $21 billion in their second quarter, up 127%  year-over-year. Now, we don't know what stocks all   that debt flowed into, but it's a reasonable guess  that some of it has flowed into the AI market. So,   if AI were to dip substantially, that is more  investors who could be forced to sell into   the dip to cover their margin loans, potentially  driving the market even lower in a feedback loop.   And as I covered in my video on what causes stock  market crashes, those feedback loops downward are   what defines a true crash. But this channel is  not about panic, it's about preparation. So,   what can we do to prepare ourselves if Bur is  right about what's happening? So, first off,   I personally would never short a stock because  even if you correctly guess that a stock will   decline if your timing isn't perfect, you can  still lose money. Bur himself almost went bankrupt   when it took 3 years from when he predicted  the housing crash in 2005 to when it actually   collapsed in 2008 with his investors revoling and  trying to pull their money out early. So, instead   of shorting the market, I would do three things.  First off, if you have a large amount of margin in   your investing account or even high debt levels  like a credit card while also trading stocks,   now could be a good time to pay that down. It  protects your downside if there's a major dip   before 2028 like Michael Bur predicts. Number two  is to check your allocations. If you own an index   fund and you invest in individual stocks, let's  say you own Google, you might be overexposed to   some of the big tech companies because these big  companies are so overweighted in these indexes.   buying an international index or even a small cap  value index could help even things out. And these   allocations really matter because we know there  are still deals out there. Even Michael Bur is   buying stocks. But certain areas like the NEO  clouds, we now know what warning signs to watch   for. And the most important is number three, which  is to avoid panic selling. Even if you're close   to retirement, there are calculators for how  much you should have in stocks versus bonds to   weather a market downturn. For everyone else,  study after study shows that staying invested   through the downturn tends to outperform trying  to time the market. And over a 20-year period, the   market has never had a negative return. But the  most important thing is to have a plan ahead of   time and stick with it. Stay invested. I'm Curran  Francis, and if you found anything valuable, just   consider subscribing to the channel. And here's a  video of what five of the biggest investors in the   market are buying right now, including a deeper  look into what Michael Bur is actually buying