원문 보기
The market looks healthy. The S&P 500 is sitting near its highs. The headlines are calm. But underneath, Michael Bur says the bare market has already started and most investors haven't noticed. On October 8th, Bur published a new warning built on a chart he calls the caldera. Here's the picture on October 6th.
Out of 503 stocks in the S&P 500, only 15 were sitting at a new high. just 97 were within 10% of their highs. Meanwhile, 80.7% of the index, 406 stocks were down more than 10% from their 10-year highs. 54.9% 276 stocks were down more than 20%. And that is Bur's bare market ratio, 55%. So 27% of stocks have broken down.
New highs from the last two years now down more than 20%. That is his fresh draw down ratio up from 22% in August and the index looks fine because a handful of giants hold it up. Underneath more than half the market is already in a bare market. Bur shows this with a shape he calls the caldera. Think of a volcano in cross-section.
The peak is the index itself near its high. The crater walls show where members actually sit. How far each stock has fallen from its own 10-year high. A healthy market has a shallow caldera. A dangerous one has a deep crater. The index at the rim most stocks far below. Right now the crater is deep. What happened the other times it looked like this.
Start with the worst one. March 2000 the dot peak. Then 15 stocks were at a high the same as today. 87 were within 10% fewer than today's 97. But the damage underneath was worse on every single reading. 71% of stocks were down more than 20% versus 55% today. 53% had freshly broken down versus 27% today.
The median stock was already down 31% at the top. What came next? Over 2 and 1/2 years, the index fell by half. Every reading on the 2000 caldera is worse than today's. But 2000 matters most for what happened to the leaders and the broken stocks after the peak. But Bur says the closest match to today isn't 2000.
It's July 1990. On the 10-year measure, 1990 is nearest to today. And on fresh draw downs, half of it, a fresh draw down ratio of 13% versus 27% today. The bare market ratio was 47%. The median stock was down 19% and the damage was old. 97 stocks in bare markets had still not regained their 1987 highs.
17 days after that snapshot, Iraq invaded Kuwait. By October, the index was down 20% and 82% of its members were in bare markets. 1990 warns that a market can look merely tired until a shock turns tired into a route. The rest of the history is a speed round. In August 1987, the Caldera looked healthy, only 20% in bare markets, and 8 weeks later, 97.9% were.
Bur's own honesty note, 64 stocks were unpriced in his 1987 data. So, the true ratio sits between 17% and 30%. In October 2007, the ratio was 43%. And Bur is blunt. The ratio did not call 2008. Housing and financials fell one by one. Breth didn't warn first. That 2007 reading was arguably inflated. 166 of the 214 beaten down stocks were still under pre205 highs, leftovers from 2000.
In 2015, 2018, 2020, 2022, and early 2025, the pattern repeats. Narrow tops, then falls of 19 to 24%. Once in 2018 and 2020, Bur says the grief cycle was interrupted. The Fed reversed in 2018, and in 2020, the rebound came faster than the crash. The cycle doesn't always finish. Now, the table that matters most.
For every past top, Bur split the market. Leaders within 10% of their highs broken already down more than 20%. And then he measured what each group did over the next 12 months. Usually, both groups moved together. In 2007, both fell. Leaders down 38% broken down 45%. In 2020, both rose. But March 2000 is the only top where leaders fell hard and broken stocks rose.
Leaders down 33% in 12 months, 50% in 18 while broken stocks gained 14% then 8%. Desbury puts it, the index's bare market was the broken stocks bull market. Today, the split is 19% leaders, 55% broken, and the outcome is still a question mark. So, why would the leaders fall this time? Bur's answer is the AI buildout.
Quote, "The tech giants, but for Apple, are spending as if their lives depended on it when their lives clearly do not. The hyperscalers are committing vast capital to AI capacity before the returns are proven. They expect government to quote tolerate their oligopoly on AI." May will tolerate and fear it.
And quote, "Even Larry Ellison was seduced to bet his company on it. If monetization disappoints," Bur argues, falling expectations, weaker cash flow, and financing pressure could undermine the leaders holding the index up. And quote, "Compression will do its deed, and much of what is being spent will be just so much sunk cost." A risk thesis, not proof, but the mechanism that could turn a breath warning into a leader's decline.
And the financing behind the buildout is the part Bur finds most fragile. He traces a chain he calls other people's money. It starts with an annuitant, a retireese savings flowing to an insurer, then Bermuda, then a sponsor, a special purpose vehicle, a bank, a landlord, a tenant, and finally a chip.
Each link takes a fee and adds leverage. Bur calls them quote pec/insure financial pyramids financing the AI buildout. He warns the unwind runs in numbered order from the permanent buyer back to the builders with quote caks and groans and some loud snapping and crackling in private equity and private credit marks.
If the marks are wrong, the pyramid wobbles from the top down. Then there's the government's room to respond or lack of it. In fiscal year 2026, the federal deficit was $1.99 trillion, 6.2% of GDP. Receipts were 5.40 trillion. Outlays were 7.40 trillion. Those are Treasury's final numbers released October 16th.
Bur's point, deficits of 5 to 6% of GDP with unemployment near 4% and no recession make 2023 to 2026 unusual. In past cycles, deficits that big came with a crisis already underway. This time, the borrowing came first. If the private debt bubble cracks, Bur argues the government is in a very weak position to support it.
It has already spent the ammunition. So, what does Bur actually expect? not necessarily a crash. His base case, in his words, the market will fail in a manner that lifts the median broken stock. This is my base case. It will perhaps not fail in a big crash, but will drift lower, sometimes violently, as fundamental news from the land of compression makes its way into the AI narrative.
And the plunge protection team is very good at treating market plunges. But a fundamentally and technically deteriorating market with falling breadth presents a problem that cannot be easily cheered back to health. He also says the bull's last months could be wild and puts the top within 0 to 9 months.
And he's backing the rotation. On September 22nd, he disclosed buying five stocks 29 to 65% below their 52- week highs. Two honesty notes before we close. First, this indicator is not a crystal ball. It did not call 2008. It twice the grief cycle was interrupted by the Fed in 2018 and by the fastest rebound ever in 2020.
Sekbury himself says 1990 not 2000 is the nearest analog. Breth says the market is fragile. Not that a date is circled. Second a footnote from his July piece. Variance ratios have sat well under 1.0 at every horizon since 2011. Markets mean reverting, the mechanical bid losing steam. Same story, different lens.
So, here is Bur's question. If the expensive leaders are the risk and the beaten down stocks are where value is quietly building, what are you actually holding? The index's bare market could be the broken stocks bull market or the cycle could be interrupted again. Either way, the caldera says the market is not as healthy as the headline.
The full bare market breakdown is linked at the end of this video. This is educational only, not financial advice. What would change your mind about the leaders? Let me know below.
Comments 0