저자: Kelvin · 외국인
요약 보기
S&P500 밸류에이션(Shiller PE 42.1, P/S 3배 이상, 버핏지표 230%)이 역사적 고점이고, 시장폭 협소(200일선 상회 종목 54%)와 소수 대형주 집중, 저소득층 소비·신용연체 악화를 하락 경고 근거로 든다. 반면 실적 성장(1분기 EPS 27.1%), 높은 EPS 상향 등 반론도 제시하며, 결론은 전체 지수는 버블이 아니지만 집중도 리스크가 크고 AI 관련 고평가 종목은 주의가 필요하다는 것. 단기 폭락 시점은 예측 불가하나 향후 10년 수익률은 낮을 것으로 봄.
다듬은 글
The stock market just hit another all-time high. But at the same time, Michael Burry, the guy who predicted the 2008 financial crisis, just doubled down on his USD 1.1 billion bet that it's about to crash. So, is the S&P 500 sitting on top of a historic bubble that's about to burst? Or are the bears wrong again, just like they've been for years?
In this video, I'm going to show you exactly what the data says, so by the end you'll know what to do with your money. Alright, let's start with the valuations first. Right now, the Shiller PE ratio is at 42.1. This is even higher than during the dot-com peak in 1999, which soon after saw the S&P 500 crash by 49%.
Likewise, the S&P 500's Price-to-Sales is trading at over 3 times revenue. That's not only way higher than during the 2000 dot-com bubble, but it also shows that in terms of revenue, we are now way more expensive than anytime in the past. Then the Buffett Indicator, which compares total market cap to US GDP, is currently sitting at 230%.
In the past, Warren Buffett himself had warned that anything above 200% is "playing with fire." So that's Warning Sign number 1, where every single major valuation metric is in the top decile of its entire history. Second, market breadth. When a market is genuinely healthy, you would see 70% or more of stocks trading above their 200-day moving average.
But right now, it's only at 54%. That means that while the index is hitting record highs, most stocks are not. Goldman Sachs calls this "one of the narrowest breadth readings in recent decades outside of the dot-com bubble, with the median S&P 500 stock still 13% below its 52-week peak." You can see it in the performance numbers too.
From 2023 till now, the market cap weighted S&P 500 has returned close to 100%, whereas the equal-weight S&P 500 only returned half of that. What this tells us is that a small number of very large companies are now carrying the index, while the rest of the 490+ stocks are on average, lagging badly. And whenever the concentration becomes this extreme, the whole index would become hostage to just a handful of names.
Now if you want to start investing, you want to keep your costs low. That's why I use Interactive Brokers. With Interactive Brokers, you're paying just US$0.35 per trade on the US stock market. That's one of the lowest in the industry. They also give you access to Ireland Domiciled ETFs, which can save you a chunk on dividend withholding taxes.
And if you're not sure what to invest in, they have free company research tools built right into the platform, so you can do your homework before putting your hard earned money in. Check out my link in the description to get started. Alright, now back to the video. Third, the index is being carried by just a handful of companies.
In Q1 2026, the S&P 500 reported 27.1% earnings growth year-over-year. However, most of that earnings growth came from just a handful of companies. In particular, Alphabet, Amazon, and Meta alone drove 71% of the net dollar earnings increase for the entire index. This means that the bull market we are seeing isn't really a true bull market in the American economy.
But rather, this is a bull market in just 3 very specific names. In fact, this has happened before in history, and it didn't end well. Back in the 1970s, the Nifty Fifty were the top 50 blue-chip companies that were considered too good to ever sell. But when the 1973-74 bear market came, it still hit them hard.
Just like Mag 7 today, the blue chip companies back then weren't bad businesses. But when the market is this concentrated and multiples this high, that's what makes them dangerous. Fourth, while Wall Street is celebrating, the consumer data is telling a very different story. Just recently, the University of Michigan Consumer Sentiment Index came in at 48.2.
This is the lowest reading since records began all the way back to 1952. And according to McDonald's CEO, consumer sentiment is continuing to deteriorate. While higher-income consumers remain resilient, lower-income consumers are cutting back on their spending. The Bank of America card data tells the same story.
While the total household spending is growing at 3.2%, lower-income households are only growing at 1.1%. This means that the bottom half is stalling. Not just that, credit card delinquencies have hit 3.3%. That's 50% above pre-pandemic lows, and within 1 percentage point of the 2009 financial crisis peak.
At JPMorgan, delinquencies are 2.3%. And at Synchrony, which skews toward lower-income borrowers, it's 4.8%. That tells you lower-income borrowers are already under severe financial stress. So while the market is pricing a soft landing, Main Street is not living one. Last but not least, more and more smart money are now sounding the alarm.
The most famous one is of course Berkshire Hathaway, which is sitting on USD 397 billion in cash, and has been a net seller for 3 consecutive years. Then there's Stanley Druckenmiller who's now warning that we are now "toward the top of the valuation range historically." And Paul Singer at Elliott Management has also set aside up to USD 7 billion as dry powder, while waiting for a dislocation.
Meanwhile, Bank of America's research shows that 60% of their historical bear market signposts have now triggered. And that reading has historically been associated with proximity to a market peak. So, those are the warning signs. But before you be like "Panic Ahh!", let me also show you the other side of the argument, as to why the market may not crash just yet.
First, while it's true that the market is trading at historically high valuations, many of those high prices are backed by real earnings, and that is fundamentally different from the 2000 dot-com bubble. For example, back in 2000, Pets.com raised USD 82 million in its IPO, burned through it in 9 months, and went bankrupt.
Webvan raised USD 375 million to deliver groceries and collapsed in under 2 years. And across the index, companies were trading at absurd revenue multiples with nothing to show. Even Cisco, the crown jewel of that era, was at 200 times earnings. At that level, the stock would require impossible growth forever to justify the price.
And eventually, it fell 86% when the bubble popped. On the other hand, the big tech names driving the index today are a different story. For example, Nvidia is trading at roughly 26 times forward earnings on USD 215 billion in actual revenue, while Apple alone generates USD 100 billion per year. These are not speculative bets on future earnings, these are businesses that are already printing money at scale, and their high prices are backed by real earnings.
Difference #2: Earnings Growth Is Real and Still Accelerating Second, not only is the S&P 500 earnings growth real, that growth is still accelerating. In Q1 2026, 84% of S&P 500 companies have beaten their Earnings Per Share estimates. That is the highest beat rate since Q2 2021. Not only that, while EPS expectations are typically revised lower at the start of a quarter, for this time around, analysts have also raised their Q2 2026 EPS estimate by 2.1%.
That is the largest intra-quarter upward revision in 5 years. So if those numbers come through, in terms of valuation, the market will actually be even cheaper in 2027 than it appears today. Third, the bears have been consistently wrong for years. So one of the most common arguments that the market will crash is that it has high valuations.
However, if you had followed this theory, you would have stayed on the sidelines when the Shiller PE first crossed 30, and missed out on over 200% worth of returns. In fact, throughout history, there have been a lot of pundits who tried to call the market top, only to fail miserably. And when it does eventually drop, they'll be like, see I'm right.
Invesco found in 2025 that the relationship between CAPE and one-year forward returns is approximately zero. This means that , over the short term, you can't reliably use today's high valuations to predict when a crash will happen. And history backs this up. In 1929, the bubble stayed elevated for about 5 years before it crashed.
Japan's 1989 bubble ran for 4 to 5 years before falling 80% over the next 18 years. And the dot-com bubble stayed above CAPE 30 for 3 years before it finally popped. This tells us that the markets can stay expensive for a very long time before they finally correct. However, that doesn't mean CAPE is useless.
At longer time horizons, it does matter. JPMorgan found that P/E ratios have a much tighter correlation with 5-year forward returns. And when you go even further to 10 years, the correlation increases even further. So while high valuations today won't tell you when the crash will come, they do tell us that returns over the next decade are likely to be much lower than what you've seen in the last decade.
However, with that being said, it doesn't mean the risk of a crash isn't real. Right now, the 4 biggest tech companies, Amazon, Microsoft, Alphabet, and Meta, are on track to spend over USD 700 billion on AI infrastructure in 2026. And they aren't stopping anytime soon. However, JPMorgan ran the numbers and found that USD 650 billion in annual AI revenue would need to be generated "into perpetuity" just to deliver a 10% return on that investment.
And on the ground, MIT found that 95% of organisations see no measurable returns from AI adoption. And more recently, PwC's 2026 CEO survey showed that 56% said they've gotten "nothing out of" their AI investments. Now, bulls will point to the J-Curve, where major technologies often show flat productivity for years before the gains materialise.
It happened with computers, it happened with the internet, and it could happen with AI too. However, the risk here is that if AI ROI disappoints over the next 12 to 24 months, that's when hundreds of billions in datacenter investment gets written down across companies currently priced for the AI supercycle.
In that scenario, an earnings problem at these multiples could lead to a bubble popping. Ok, here is my take. Right now, I don't believe that the entire S&P 500 is in a bubble, at least not by any definition that matters. A bubble is when prices disconnect entirely from fundamentals. And parts of this market are showing exactly that sign.
However, the companies that are actually driving the index, whether it's Nvidia, Apple, Microsoft, Alphabet, these are businesses generating hundreds of billions in real revenue and real free cash flow. So while they may look expensive, I don't believe that they are in bubble territory. That said, what concerns me most is not the valuation level, but rather, it is the concentration.
40% of the entire US stock market is in just the top 10 companies. And historically, when concentration gets this extreme, the correction that comes is rarely painless. But here is the other side. The market can stay irrational longer than you can stay solvent. And history is very clear on this: the buy-and-hold investor will almost always do better than the "smart investor" sitting on the sidelines waiting for the perfect entry.
The bears have been warning about these valuations since almost a decade ago, and the S&P 500 has returned more than 200% since then. So don't panic-sell. Don't go to cash. But do look at your portfolio, especially if you are concentrated in AI names that have run 200 to 500% on nothing but momentum.
That is where the real bubble risk lives. As for the broader index, they are built on real earnings, and because of that, the bull run might just continue for a while longer, at least until they pops. Anyway, that's all for this video. Hope you found this useful. Like and subscribe and I'll see you in the next one.
Comments 0