원문 보기
Michael Berry, the man who predicted the 2008 housing crash and slump, has just issued another warning about the stock market in 2026. In particular, one of the largest American companies, according to Michael Berry, may be on the verge of a sharp decline. In this video, I'll tell you about Berry's new warning, and I'll also show you some very disturbing data about the state of the housing market and the stock market right now.
And this data reflects the fact that we are actually, for the first time in U.S. history, in what I call a dueling asset bubble. Both stocks, highlighted in blue, and the housing sector, highlighted in orange, have near-record valuations in 2026. We have never seen a period like this before in US history .
Specifically, we look at the housing- to-GDP ratio here, as well as the equity- to-GDP ratio. Both of these indicators suggest that the valuations of these assets are not supported by the growth of the underlying economy. This has been the case for the past three or four years, and the stock market and economy continue to move .
However, it is worth asking the question: how long can this boom last if the fundamentals do not support it? Given that the Shiller Cape P/E ratio shows the stock is currently trading at 41 times its 10-year earnings, this is the second most expensive stock price outside of the 1999 bubble. The only other time it came close to that was 1929, just before the Great Depression.
And one might suspect that it is this stock market bubble that is fundamentally supporting the economy now. Many people were confused about how people continue to spend and companies continue to make profits. Well, it's because of this record stock market bubble. One has to wonder how long this can last, especially given Michael Berry's latest warning.
Big Short investor Michael Berry is currently sounding the alarm about Nvidia, which is the largest company in the world, and also about something called AI token maximization. Michael Berry warned that Nvidia's stock losses look vulnerable, and that the trend of AI token maximization will not continue in the future, according to a Business Insider article.
He says the conditions for an aggressive decline are as strong as they have been in stock history. He added that Nvidia's next drop could be more dramatic than the last three major crashes: 56% in 2018, 67% in 2021, and 43% in 2025. Maybe some of you know about Nvidia or at least have heard of them.
They are now the largest company in the world. Their market cap is over $5 trillion, and they essentially sell GPUs to build data centers. The chips used in these data centers are then used to train artificial intelligence models, and there is currently a huge demand for these chips from companies like Meta, Microsoft, Oracle, and Google.
They're buying up these chips all over the place because they're all building data centers. And this led to an incredible increase in Nvidia's stock price , as well as their revenues and profits. But on Michael Berry's Substack, he revealed something very interesting that I want to show you. This graph is taken from a book by Michael Berry called Cassandra Unleashed .
Actually, this is a great article to read. I would recommend you subscribe to it. I'm subscribed to it and saw this graph that shows that Nvidia's top three customers account for 64% of Nvidia's accounts receivable. That is, in essence, almost 2/3 of Nvidia's revenue comes from three customers. Berry postulates that the largest customer is probably Microsoft, and we can assume that some other large customers could be Google, Meta, OpenAI, Oracle.
And one can only wonder how sustainable Nvidia, the world's largest company with a market capitalization estimated at over $5 trillion, is, how sustainable it is for the world's largest company with only three customers that take up 2/3 of its revenue. I mean , after all, how diversified is this business?
How sustainable is this business? How well will this business survive changes in the environment in the coming years, when perhaps one of these customers disappears? It's a kind of incestuous relationship between many of these biggest companies here in the US. Many of them buy and sell services and goods to each other, increasing their own income, often using debt to finance this.
And in fact, if you want more evidence that we may be in a stock market bubble right now, take a look. SpaceX just had an IPO. They're on this chart from the Wall Street Journal. You can see their market cap, their total value is now higher than Amazon's and almost as high as Microsoft's, which is pretty crazy.
This is another kind of anecdotal indicator of where we are in the market, which is SpaceX, a company that does several things. They launch satellites into orbit, they launch astronauts into space, and really their revenue driver is Starlink, which provides internet to many people around the world. So it's a somewhat diversified business, but it's fragmented, and their total revenue is now about 14 billion a year.
$14 billion in annual revenue is valued at nearly $2.5 trillion. This is madness, everyone. The price- to-earnings ratio for SpaceX is 100 times. So, their market capitalization is currently more than 100 times their revenue, which is actually 19 billion. So, I am corrected. 19 billion in revenue, according to Google Gemini on SpaceX.
You can see that Microsoft has 320 billion in revenue and Amazon has 640 billion. Microsoft and Amazon also have profits of between $32 billion and $90 billion, while SpaceX is losing money. Yet somehow SpaceX is now as valuable on the stock market as Microsoft and Amazon. And that's exactly what happens in the late stages of a bubble, everyone.
When you see this hype around IPOs, SpaceX just did an IPO, Anthropic is probably next, OpenAI is probably after that, and they're all going to be rushing to raise as much capital as they can by the end of 2026. And many retail investors are going all out for it. They are buying up shares of a company that, I don't know, maybe in 10 years, maybe it will be thriving, maybe it won't.
We do n't really know. I think you could say it's probably crazy that SpaceX is now valued as highly as Microsoft and Amazon, but that's exactly what's happening in this stock market right now. So for those who don't think this is a bubble, I would just advise looking at these individual metrics alongside the data I showed you at the beginning.
Since the US stock market is valued at approximately $76 trillion, and the Wilshire 5000 index trades at 250% of GDP. So the stock is literally worth 2.5 times the value of the entire US economy. At the same time, housing, the orange line, accounts for 146% of GDP. And what you can see that's really interesting about this period in 2026 is that it's unique.
We haven't really seen a period in US history where both stocks and housing were, let's call it, in a major bubble relative to the fundamentals of the underlying economy. You can see that in 2006 it was just housing. Housing accounted for 163% of GDP in 2006. You could see a big bubble, and then it burst.
But stocks weren't in such a big bubble back then. Going back to 2000, 2001, it was stocks that were in a "bubble" at 160% of GDP, while housing was only about 100% of GDP. So this is a very rare scenario, fast forward to today, where both of these assets , assets in the U.S. economy, are trading at valuation levels that are not supported by the underlying size of the economy and economic growth.
Many people over the past few years have been saying that this is simply the new reality. Many people say that the old rules no longer apply to valuing the stock market and the housing market. Many of these people argue that things like GDP, income, inflation, rents no longer matter, and that valuations are traded on some other basis, some other forward multiplier that is perhaps difficult to understand.
No one can explain it. And what worries me is that the underlying state of the economy is now becoming inextricably linked to these asset bubbles. That consumer spending in America is increasingly being driven by the fact that people are looking at their 401(k)s and looking at their Zestimates, and feeling comfortable with their paper wealth, so they're spending money that they don't necessarily have.
And you can see this very clearly on the graph of personal savings rates in America starting around 1955. You can see that this is the percentage of Americans' paychecks that are saved each month. And in April 2026, this savings rate fell to almost a minimum of 2.6 %. This is close to the serial minimum, only 2.6% of salaries are invested in savings.
The long-term rate is now well above 5%. In fact, in the 70s, 80s, and 90s, this figure was closer to 10%. Back then, people saved 10% of their salary. They are currently saving only 2.6%. The only other times we've seen anything like this were during the housing booms of 2005, 2006, and 2007. We also saw something similar during the inflation spike in 2022.
And just before the dot-com crash, the last time such a drop in personal savings rates was observed . So this is a worrying trend. Americans are saving less and less of their paychecks. In fact, in the short term, some people might see this as a positive phenomenon, especially large publicly traded companies , because the less Americans save, the more they spend.
So, the inverse relationship to this personal savings schedule will be interest expense and consumer spending. And, of course , consumer spending is still holding steady. Retail sales in the United States, according to the U.S. Census Bureau, continued to grow. And if you look at the earnings growth rate of the S &P 500, data from Robert Shiller, you can see that the earnings growth rate is still high.
38% year- over-year revenue growth through May 2026. So this blue line going up means that the reported earnings growth of public companies is strong, which means that there is some momentum in spending, and that these companies may also be able to achieve operational efficiencies and reduce costs through artificial intelligence.
Interestingly, this year-over-year earnings growth graph predicted recessions very well. So when earnings growth starts to fall, as it did in early 2001, it was a signal of the dot-com crash, and eventually earnings fell by 50%. When profits began to fall in late 2007, it signaled the global financial crisis, and profits fell by 90%.
Profits began to fall during the lockdowns, which was a signal of this recession. And you could see that almost every period, more or less, when profits fall corresponds to a recession. You can also see the Great Depression of 1930, the two parts of the Great Depression of 1930-1938 also had a significant drop in income.
So, we are not on the verge of an inevitable stock market crash and recession right now, because as long as earnings are growing at 30% year- over-year, you won't see a stock market crash tomorrow. You won't see a recession tomorrow. However, what you should watch out for is if there is eventually a slowdown in earnings growth.
If companies start reporting earnings growth, maybe 10% year-over-year , 15% year-over-year, maybe 6 months later earnings growth slows to stagnation, that would be a signal that we are in a more imminent recessionary environment. So for those of you worried about an imminent stock market crash, it does n't look like it.
But what concerns me is that, in my opinion, earnings are very much tied to stock valuation. And as earnings eventually slow, stock prices will also slow, leading to an even greater slowdown in consumer spending, because remember, people aren't saving enough to support spending . So, if their equity in the stock market declines, as well as their equity in the housing market, it could reduce their consumer spending, which would lead to lower earnings growth, which in turn would lead to lower stock valuations and could cause a vicious cycle.
And it seems like many different parts of this economy are interconnected. On one hand, we have the capital spending on AI, the AI boom , Nvidia supplying chips to companies like Oracle, on the other hand, the tech boom in stock prices and then the SpaceX IPO, and people going crazy and obsessed with it, and people feeling good about their 401(k)s and their stock portfolios because of it, and then they increase their spending because of it, and that gets into retail sales.
But what happens if even one of these elements slows down, right? For example, what happens if in 6 months we realize that maybe we don't need as many data centers, maybe we don't need as much compute, maybe some kind of token maximization that Michael Berry talked about slows down. And again, Michael Berry talked about this in his Substack Cassandra Unchained talk.
He talked about how he thinks many of these companies are probably at the peak of demand for AI tokens right now because they are training all these models for the first time. He also talks about how many of these companies have a culture where employees are encouraged to burn as many tokens as possible.
You know, an employee burns 30 million tokens to climb the leaderboard for the maximum number of tokens. This is what some of these companies have been doing, creating rankings of the number of tokens their developers have used. So , there was an incentive to just use tokens, tokens, and tokens. And this, of course , has led to huge revenue growth for companies like Anthropic and OpenAI.
But what happens when the models are more or less trained, or most of that training is complete, or maybe some of these companies start to cut back on how much they spend on tokens. The value of these tokens has increased, and we see that now some of these AI agents that operate autonomously can cost between one hundred and three hundred thousand dollars a year.
According to a recent Fortune article that argues that the cost of computing far exceeds the cost of a single employee, Nvidia's head of deep learning claims that AI is now more expensive than paying people to work. Which is a bit of a mystery. Over the last 6 months, we have heard a lot about how AI will lead to massive job losses.
I reported about this on this channel. Some companies have certainly carried out mass layoffs. Now, the narrative seems to be shifting to the point that perhaps AI will be more expensive than hiring humans. So how will this work? The truth is, I don't really know. However, I know that the math here doesn't make sense in fundamental ways.
And the longer we continue with the stock market valued far above earnings and the housing market valued far above GDP, the more pressure there will be on AI, on Nvidia, on these companies to continue to show crazy growth and crazy results. And if these companies fail at all, if revenue growth slows at all , if we see even a small downturn, that could potentially cause some problems and lead to a drop in stock market valuations, which could lead to a drop in consumer spending itself.
And if that happens, the question we have to ask ourselves is, what will happen to the housing market then? Well, I think some of the areas in the U.S. housing market that are going to be most susceptible to a potential stock market crash, and I'm not saying the stock market is guaranteed to crash, but the potential crash would be areas with a lot of shadow inventory, areas with a lot of second and third homes, in particular.
This is what we highlight on this map from the Reventure program. Areas where more than 10% of the housing stock is shadow stock. I think they will be the center of attention. We're talking about a lot of markets in Florida where people own second, third and fourth homes, people with large stock portfolios.
This includes a bunch of markets in the Northeast, second and third home markets in Maine and Vermont , and upstate New York. These are areas in Arizona and Colorado, including Aspen and some ski towns. We are also planning California and Lake Tahoe. I think the first areas to be hit by a potential stock market crash will be these shadow stock markets with high levels of returns.
That will be a lot of people who have large stock portfolios, as well as a lot of second and third homes. I think some of these second and third homes will be sold and listed if the stock market crashes. Additionally, I would also look at high- priced areas where the typical home costs over $ 700,000.
That's not a lot of areas, but they are tech-dominated markets: San Jose, Jackson Hole, Wyoming, San Francisco, parts of Utah, Steamboat Springs, Colorado, Hailey, Idaho, Breckenridge, Colorado. And what's really interesting is that many of these expensive markets also have a lot of shadow inventory.
So that's something I would keep a close eye on over the next 6-12 months. If we see a slowdown in the stock market, a slowdown in the economy, I think those are going to be some of the most vulnerable areas in terms of where prices could fall the most and where the biggest sell-off could occur . Until then, homebuyers and investors should keep a close eye on our 2027 forecast for your city and zip code.
Where do we think prices will go over the next 12 months, based on supply and demand in the market today? Knowing the answer to this question will help you a lot as a home buyer or investor right now. So, go to www.reventure.app and check out the 2027 forecast for your region to know where prices will be headed next year.
Prices vary greatly in different cities. The bluer the zip code, the greater the forecast for price declines. The redder the zip code, the more we believe prices will rise. Go to www.reventure.app now to see the 2027 forecast for your city and zip code.
Comments 0