Latest / Deep Values: Value Investing Deep Dives / Michael Burry and Chamath Palihapitiya's Warning: The Case for a 77% Market Crash
Transcript
- 0:00Imagine waking up tomorrow. You roll over, check your phone
- 0:03and your stock portfolio. You know, the one you've been
- 0:05diligently building, sacrificing for just pouring a chunk of your
- 0:09paycheck into for years has just evaporated by 75%.
- 0:13Oh man, yeah, a total nightmare scenario.
- 0:15Right. And the craziest part of this
- 0:17scenario is that it's not because of some sudden global
- 0:21war. It's not a catastrophic banking
- 0:23collapse like we saw in 2008. It is literally just because of
- 0:27math, shifting demographics and artificial intelligence.
- 0:31I mean, it sounds like financial science fiction honestly, but
- 0:34when you strip away the emotion and look strictly at the
- 0:36underlying mechanics, it's actually a mathematical
- 0:39possibility that is flashing red across multiple data points
- 0:42right now. Yeah.
- 0:43And that is exactly our mission for this deep dive.
- 0:46We are unpacking to seemingly independent but really eerily
- 0:49parallel warnings from absolute heavyweights in the financial
- 0:53world about an impending 75 to 77% market.
- 0:57Crash. Exactly.
- 0:59So we're pulling from a detailed analysis by George Gammon, who
- 1:01breaks down this rather dire mathematical warning from the
- 1:04legendary investor Michael Bury. Right, the big short guy.
- 1:07And then we're pairing that with an analysis from Terry Chen, who
- 1:11unpacks A viral thesis from the tech investor Chamath
- 1:14Pallihapatiya about how AI is just fundamentally breaking the
- 1:18valuation model of the whole stock market.
- 1:21Two totally different approaches, you know. 1 is
- 1:23rooted in historical data and the other is based entirely on
- 1:26futuristic technology, but they are arriving at the exact same
- 1:30terrifying number. It's wild that we are dealing
- 1:33with some heavy financial theory today, so a quick point of
- 1:36housekeeping before we pull this thread.
- 1:38All analysis is based on publicly available sources and
- 1:42may be subject to revisions or differing interpretations.
- 1:45All contents from this channel is for educational and research
- 1:48purposes only and does not constitute investment advice.
- 1:51Conduct your own due diligence before making investment
- 1:54decisions. Always super important to keep
- 1:56in mind, especially with stuff like this.
- 1:58Totally also to access more advanced researches,
- 2:02applications and our latest published personal Intelligent
- 2:05Multi agent Deep Value intelligence.
- 2:08Welcome to subscribe on Deep Values dot AI to track total
- 2:11market temperature. Welcome to search and download
- 2:14the iOS App Buffett Indicator from the Apple App Store.
- 2:18All right, so let's get into it. To understand how a 75% drop is
- 2:23even mechanically possible without the world ending, we
- 2:26first have to shatter a very comforting, very pervasive myth.
- 2:29OK, lay it on me. It's the myth that the stock
- 2:31market always goes up as long as you wait long enough.
- 2:34I mean, every standard financial planner, every retirement
- 2:37brochure shows you that beautiful chart, right?
- 2:40Just go steadily up and to the right.
- 2:42It looks like this guaranteed staircase to a comfortable
- 2:44retirement. You just close your eyes, dollar
- 2:46cost average and let time do the work.
- 2:48Right. But Michael Bury points out a
- 2:50massive flaw in how those charts are presented.
- 2:53They are almost always nominal charts.
- 2:55They simply track the numerical price of the index over time.
- 2:58Which doesn't tell the whole story.
- 3:00Exactly because nominal charts are completely misleading you as
- 3:04an investor. You don't actually care about
- 3:06the numerical value of your portfolio, you care about your
- 3:09purchasing power. What kind of life can that money
- 3:12actually buy you when you retire?
- 3:15Right, because inflation is always eating away at the
- 3:16background. Exactly, and when Bury adjusts
- 3:19the SP500 for inflation to show actual purchasing power, the
- 3:23historical chart looks radically different.
- 3:26I mean, in real terms, the market between 1930 and 1985 was
- 3:30effectively flat. Wait flat for like 55 years.
- 3:34Functionally, yes. Let's look at a specific window
- 3:37in that period. If you bought into the S&P 500
- 3:40in 1930 and held it for 25 long years, you weren't just treading
- 3:43water. Don't tell me you lost money.
- 3:45In terms of actual purchasing power, your ability to buy
- 3:47groceries or a house, you were down nearly 50% after 1/4 of a
- 3:51century. OK, let's unpack this because
- 3:54that goes against literally everything we are taught about
- 3:56passive investing. The golden rule drummed into our
- 3:59heads is time in the market beats timing the market.
- 4:04Yeah, that's the classic line. But you're telling me if I
- 4:06follow that rule in the late 60s, I wouldn't have even broken
- 4:10even on my purchasing power until the mid 90s?
- 4:13The historical data is incredibly clear on this.
- 4:16And it's not just ancient history from the Great
- 4:18Depression either. Look at the period from 2000 to
- 4:212010, another full decade where the market was down massively in
- 4:25real terms. Man, it's like running on a
- 4:27treadmill for decades. You're putting in all this
- 4:30effort, you're tying up your capital, you're doing exactly
- 4:34what the personal finance gurus told you to do, but you are
- 4:37actually going absolutely nowhere.
- 4:39Yeah, you're just sweating for nothing.
- 4:40It really forces you to ask yourself a very uncomfortable
- 4:43question, like, do you realistically have 30 or 40
- 4:47years to just wait around for your portfolio to break even?
- 4:50Most people simply don't have that kind of timeline, and this
- 4:53brings us to the core mechanism of why these massive multi
- 4:57decade flatlines and crashes happen in the first place.
- 4:59Which is what exact? Well, most retail investors
- 5:02operate under the assumption that a company's stock price is
- 5:05just a direct reflection of its earnings.
- 5:08If a company makes more profit, the stock goes up.
- 5:10If they lose money, the stock goes down.
- 5:12I mean earnings equal price. It's logical.
- 5:15Logical, yeah, but incomplete. It's missing the most critical
- 5:18variable in the entire financial system.
- 5:21Price actually equals earnings multiplied by what we call the
- 5:24multiple or the price to earnings ratio.
- 5:26The famous PE ratio. Right.
- 5:28Investors spend all their time obsessing over the earnings part
- 5:32of that equation. You know, did Apple sell more
- 5:35iPhones this quarter? But Michael Bury's core thesis
- 5:39is that massive market crashes are almost entirely driven by
- 5:43multiple compression. I want to make sure we're
- 5:45totally clear on this concept, because it's the linchpin of a
- 5:48whole 75% drop theory. So let's use an analogy.
- 5:50Sure, I'll go. It's like paying for a designer
- 5:52label. The earnings represent the
- 5:54actual physical quality of the shirt.
- 5:57You know, the cotton, the stitching, how well it keeps you
- 5:58warm. But the multiple.
- 6:00That's the designer tax. I like that.
- 6:02Right. It's the massive premium you
- 6:04agree to pay just because of the brand name stitched onto the
- 6:07collar. So if the culture suddenly
- 6:09shifts and the market decides that designer label isn't cool
- 6:13anymore, the price of the shirt is going to plummet.
- 6:15Even though it's the exact same shirt.
- 6:17Exactly. Yeah, the physical quality of
- 6:19the shirt hasn't changed one bit, but the premium just
- 6:22vanishes. That captures the psychology
- 6:24perfectly, and we have seen the financial equivalent of that
- 6:28designer label losing its premium.
- 6:31Look at the bear market from 2000 to 2009.
- 6:34The Lost Decade. Yeah, over that decade,
- 6:37corporate earnings actually grew by 46%.
- 6:40The underlying companies in the S&P 500 were making
- 6:42significantly more money at the end of the decade than at the
- 6:45beginning. But portfolios still got wiped
- 6:47out. Exactly why?
- 6:49Because there was a 98% compression in the PE multiple.
- 6:53The earnings grew robustly, but the market's willingness to pay
- 6:56a premium collapsed, dragging the entire market down with it.
- 7:00Wait, a 70 97 drop just from the multiple shrinking?
- 7:04That assumes the underlying earnings of the 500 don't grow
- 7:07enough to offset the multiple compression Are burying gammon
- 7:11saying cororate innovation just stops.
- 7:13Not at all. As the 2000s showed us, earnings
- 7:16can grow and you still lose your shirt if the multiple compresses
- 7:18hard enough. Bury runs the math on where we
- 7:21are sitting right now, and the numbers are just incredibly
- 7:24sobering. How bad are we talking?
- 7:25Well, he calculates that if the market's multiple just reverts
- 7:29to the average we've seen since 1990.
- 7:31So we aren't even talking about a historic crash, just returning
- 7:34to the modern everyday average. The S&P 500 drops 32%.
- 7:38A 32% haircut just to get back to normal?
- 7:41That's rough. Now if we hit the multiple
- 7:43through we saw post 1980 which was a PE of about 11, the market
- 7:47dropped 72% and if we revert to the all time historic trough
- 7:51going back to 1921 which was a multiple of 9.25, the S&P 500
- 7:55crashes by 77%. You really have to let that
- 7:58number sink in. Is 77% crash.
- 8:02And again, that's not because companies suddenly forget how to
- 8:05make money or because the economy physically collapses.
- 8:08It's literally just the multiple shrinking back to historical
- 8:11lows. Exactly.
- 8:12Just the premium evaporating. So the obvious question, the one
- 8:15you have to be asking right now, is why would the multiple
- 8:18compress now? Like what is the catalyst that
- 8:21forces the market to suddenly RIP that designer label off the
- 8:24shirt? George Gammon's breakdown points
- 8:27to the first major catalyst, which is demographics, and
- 8:30specifically it's tied to the hidden mechanics of passive
- 8:34investing. Passive investing.
- 8:35For years, we've had a system where young workers
- 8:38automatically blindly pour about 10% of their paychecks into S&P
- 8:42500 index funds. The classic 401K autopilot.
- 8:45Yeah, every two weeks, algorithms just buy the S&P 500
- 8:48regardless of the price. It creates this artificial
- 8:50constant demand. Demand that props up the
- 8:52multiple, but we are currently crossing a historic demographic
- 8:56inversion. We have reached a point where
- 8:59for every young worker putting $1.00 into the market, there are
- 9:02retiring baby boomers pulling $3 out to fund their retirements.
- 9:06Wow, a three to one ratio of money leaving versus money
- 9:11entering. Historically, there was always
- 9:13more fresh capital going in than coming out, which naturally
- 9:16pushed the market in the multiple steadily higher.
- 9:18When you have $3 exiting for every $1.00 entering, the math
- 9:22reverses. That makes total sense.
- 9:24Yeah, it creates A structural, unavoidable downward pressure on
- 9:28the multiple let's. Translate some of the jargon
- 9:30from the source material here because Gammon references a
- 9:32recent -92,000 non farm payroll print.
- 9:37Right, the jobs report. Yeah, a negative payroll print
- 9:40basically means the economy is actively shedding jobs.
- 9:43Mechanically, what happens to the market when job losses start
- 9:46piling up? Those automatic 4 O 1K deposits
- 9:49vanish instantly. If someone loses their job, they
- 9:52stop contributing to their retirement account, Which?
- 9:54Kills the automatic demand. Exactly.
- 9:57Gammon notes that if unemployment climbs to somewhere
- 9:59around 5.5 to 6%, those passive inflows completely stall out and
- 10:04turn into net outflows. People aren't just stopping
- 10:07their contributions, they are actively withdrawing cash to
- 10:10survive. Because they have to pay rent.
- 10:12Right. And when that happens, the
- 10:14multiple compresses aggressively because the structural automated
- 10:18bid that was holding the market up just disappears entirely.
- 10:22So the demographic shift and the reliance on passive flows that
- 10:26kind of pulls the floor out from underneath these companies
- 10:29slowly over a decade. But what happens if the
- 10:32companies themselves lose their inherent value overnight?
- 10:34Yeah, that's the real wild card. Because that's the second, much
- 10:37faster threat we need to look at.
- 10:39Here's what gets really interesting.
- 10:41This brings us to Terry Chen's analysis of Chamath's viral
- 10:44thesis about artificial intelligence.
- 10:46Right. If demographics are a slow
- 10:48moving glacier, AI is a sudden seismic shock.
- 10:51Chamath's thesis revolves around this fundamental concept in
- 10:54finance called terminal value. Terminal.
- 10:56Yeah. So when you buy a stock, a
- 10:58massive chump of the price you pay today, sometimes like 60 to
- 11:0280% of it, is based on the assumption that the company will
- 11:05continue to exist and generate cash flow far, far into the
- 11:09future. We're talking about the concept
- 11:11of a Moat. Investors have historically
- 11:14given tech companies like Apple or Microsoft these massive 50X
- 11:18to 100X multiples because they assume those companies have
- 11:22permanent competitive advantages.
- 11:24Their moats are so deep, their brand loyalty is so strong,
- 11:27nobody can ever cross them. You were willing to pay a high
- 11:29price today because you assume Google is still going to be
- 11:32dominating search in 2050. But AI fundamentally breaks that
- 11:36assumption. AI drops the cost of innovation
- 11:38and disruption to near 0. It speeds up the development
- 11:42cycle so intensely that traditional modes just
- 11:45evaporate. Because anyone can code now.
- 11:46Right. Chamath argues that you can no
- 11:48longer assume a software company will be dominant in 10 or 20
- 11:51years because a handful of engineers in a garage with
- 11:54powerful AI tools could replicate their entire core
- 11:57business model in a weekend. That is wild.
- 12:00Let's get into the actual psychology of the math here,
- 12:02though, yeah? How does a disappearing Moat
- 12:05actually force the stock market to crash?
- 12:07It all comes down to the free cash flow multiple.
- 12:10Let's establish the baseline. Right now the 10 year U.S.
- 12:13Treasury yield is hovering around 4.5%.
- 12:16That is the risk free rate. OK, you can loan your money to
- 12:19the US government and get a guaranteed 4.5% return.
- 12:23So if you want investors to take on the risk of buying stocks
- 12:26instead of safe government bonds, you have to offer them a
- 12:28higher expected return, historically around 9%.
- 12:31Right, you need a premium for taking a risk.
- 12:33Nobody is investing in the stock market for charity.
- 12:36Exactly. So if you run the valuation math
- 12:38on a normal, stable company that isn't growing at lightning speed
- 12:41but will predictably exist for a long time, investors will
- 12:44typically pay about an 11X multiple on its free cash flow.
- 12:48OK. And 11X multiple.
- 12:50But here is Chamoth's brilliant insight What happens to that
- 12:54multiple if AI drastically shortens a company's lifespan?
- 12:58Because they get disrupted into oblivion by a faster, cheaper AI
- 13:02competitor. Yes, Shemoth calculates that if
- 13:06AI introduces just a 20% probability of completely
- 13:09destroying a company's business model every single year, that
- 13:12company's expected lifespan mathematically drops to just
- 13:14five years. Only five years.
- 13:16And if a company is only expected to survive for five
- 13:19years, investors are absolutely not going to pay an 11X
- 13:22multiple. They logically cannot recoup
- 13:25their initial investment in that short time frame without massive
- 13:28dividend yields, which modern tech companies rarely pay right.
- 13:32They just do buybacks. Exactly.
- 13:34So to account for the risk of the company dying in five years,
- 13:37investors will demand a much lower upfront price.
- 13:40They will only pay a 3.9 X multiple on that free cash flow.
- 13:44OK. Let's apply this 3.9 X multiple
- 13:46to the real world. If we look at the entire S&P 500
- 13:50today. Well, the current S&P 500 is
- 13:52valued at roughly A21X free cash flow multiple.
- 13:55That gives the index a total market capitalization of about
- 13:57$58 trillion. 58 trillion, OK. But if the threat of AI forces
- 14:01that multiple to compress down to 3.9 X because investors
- 14:05fundamentally stop believing these companies have a long term
- 14:07future. Hold on, let me do the mental
- 14:09math on that 58 trillion. Divided down by that ratio, that
- 14:14drops the market cap to roughly $14 trillion.
- 14:17A 75% evaporation of wealth. That is wow, that is exactly the
- 14:22same number. Michael Bury's looking at 100
- 14:25year historical charts of multiple reversions and gets a
- 14:2877% crash. Chamath is looking at AI
- 14:32shortening corporate lifespans, running the free cash flow math
- 14:36and gets a 75% crash. They're arriving at the exact
- 14:40same catastrophic destination from completely different
- 14:43starting points. And we have actually seen
- 14:45localized micro versions of this terminal value collapse play out
- 14:49in recent history. Think about what digital
- 14:51advertising did to the print media industry between 2005 and
- 14:542000. 15 Yeah, the traditional newspapers and legacy magazines.
- 14:57They're EBITDA multiples, which is essentially a metric of their
- 15:00operating profit compressed from 15X down to 2X almost overnight.
- 15:04They were still making money in 2008, but the the market
- 15:06realized their moats were permanently gone.
- 15:08The Internet had killed their terminal value.
- 15:10The market just priced them for death.
- 15:12Exactly. We saw the exact same mechanism
- 15:14when Amazon crushed traditional retail multiples, or when Uber
- 15:18completely destroyed the terminal value of taxi
- 15:20medallions. That's a great example.
- 15:22Yeah, the taxi industry didn't die the day Uber launched, but
- 15:25the moment the market realized the long term value of a
- 15:28medallion was zero, the pricing collapsed instantly.
- 15:32Chamath is simply arguing that AI is going to do to the entire
- 15:35S&P 500 what Uber did to taxi medallions.
- 15:38So what does this all mean for you listening right now if you
- 15:41have your retirement tied up in index funds, is it all just doom
- 15:45and gloom? Because frankly, if I'm a
- 15:47massive tech company with billions in the bank, I'm not
- 15:49just going to roll over and let a startup destroy me.
- 15:52There have to be holes in this theory.
- 15:53Oh, for sure. Terry Chen provides some
- 15:56essential counter arguments in his analysis.
- 15:58First, he points out that a company's Moat isn't just made
- 16:01of software code that an AI can rewrite.
- 16:03Real moats often include things that are incredibly hard for AI
- 16:06to disrupt quickly. Like physical infrastructure or
- 16:09regulatory capture? Exactly what physical government
- 16:13licenses, deeply entrenched regulatory approvals or massive
- 16:17user switching costs? Moving a Fortune 500 companies
- 16:21entire database off a legacy system to a new AI platform
- 16:25isn't just about having better tech.
- 16:27No, it's a huge headache. Right.
- 16:28It requires immense logistical nightmares and security audits
- 16:31that essentially protect the incumbent companies.
- 16:34And what are the earnings side of the equation?
- 16:36We've been so focused on the multiple compressing, but
- 16:38couldn't AI make these companies so intensely productive that
- 16:42their actual earnings absolutely explode?
- 16:44That's a huge. Point like if a company can
- 16:46replace half its white collar workforce with an AI agent, its
- 16:50profit margins could skyrocket which might offset the multiple
- 16:53drop. Yeah, that's a very valid
- 16:55macroeconomic counterbalance. Chin also points out that if the
- 16:59S&P 500 actually started to plummet by 50 or 70%, the
- 17:02Federal Reserve would not just sit there, They would likely
- 17:04panic and drop interest rates straight back to 0.
- 17:07And as we established earlier, the valuation model is based on
- 17:10that risk free treasury rate. If rates go to 0, borrowing
- 17:14becomes free and terminal values mathematically inflate backup
- 17:18cushioning the blow. But even with those Silver
- 17:20Linings, both sources agree that the landscape of investing is
- 17:23fundamentally shifting. AI's ability to disrupt the
- 17:26digital world is essentially limitless right now, but its
- 17:30ability to disrupt the physical world is still incredibly far
- 17:34off, and at single insight dictates where smart capital is
- 17:37preparing to flow. Capital behaves like water,
- 17:41right? It always flows to the path of
- 17:42least resistance, and more importantly, it flows to where
- 17:45it's safe. Always.
- 17:46AI can generate 1,000,000 lines of code at 0 marginal cost, but
- 17:49the marginal cost of creating a new acre of fertile farmland or
- 17:52physical gallon of freshwater, or a functional toll road is
- 17:55still Infinity. Capital is fleeing to absolute
- 17:58scarcity. Both of our sources point to
- 18:00very specific safe havens. Because of this they are looking
- 18:03at tangible AI proof assets like what?
- 18:06Things like energy, infrastructure, farmland,
- 18:09physical water rights, gold and short term U.S.
- 18:12Treasury bills. Which brings us to perhaps the
- 18:15most compelling real world evidence that the absolute
- 18:18heavyweights of investing see this exact shift coming.
- 18:22Look at what Warren Buffett is doing right now.
- 18:24He is the ultimate buy and hold guy, the investor who famously
- 18:28says his favorite holding period is forever.
- 18:30Right now, Warren Buffett is sitting on a historic,
- 18:34unprecedented cash pile of roughly $370 billion. 370
- 18:40billion. Yes, he has completely stopped
- 18:42buying the dip. He is aggressively selling off
- 18:44massive stakes in tech and banking and he's parking that
- 18:47money specifically in short term T-bills.
- 18:51Contrary to his folksy, grandfatherly image of his
- 18:53buying good companies and waiting, he is actively time in
- 18:56the market. He is pulling his chips off the
- 18:58table because he sees the underlying math changing.
- 19:01He is hoarding dry powder. He knows that whether it's the
- 19:04demographic squeeze of retiring boomers or the AI driven
- 19:07destruction of terminal value, the multiples cannot stay.
- 19:10This elevated waited forever. So he's just waiting it out.
- 19:13Right when the compression happens, the capital parked in
- 19:16safe physical assets, or in his case cash, will be the only
- 19:20capital left in a position to buy up the pieces at a massive
- 19:24discount. Let's bring this all together.
- 19:26What we've uncovered today is that the structural foundation
- 19:29of the stock market is under severe pressure from 2 entirely
- 19:33different directions. Yeah, it's perfect, Storm.
- 19:35On one side, you have the slow, inevitable math of demographics.
- 19:39An aging population pulling more money out of passive index funds
- 19:43than the younger generation is putting in, removing the
- 19:46structural bid that holds up the market multiple.
- 19:48And then on the other side. You have the lightning fast
- 19:51technological shock of AI which threatens to drop the lifespan
- 19:55of modern companies from decades down to mere years, destroying
- 19:59their terminal value in the process.
- 20:00Either one of these forces alone could cause massive multiple
- 20:03compression. Together, they explain why
- 20:06Titans like Burry and Chamoth are modeling a potential 75%
- 20:10evaporation of market wealth. As we wrap up, I want to leave
- 20:12you with one final, deeply ironic paradox from Chamoth's
- 20:16thesis to Mull over. OK, let's hear it.
- 20:19We know that building these revolutionary AI systems
- 20:22requires massive long term capital investments.
- 20:26We're talking about billions of dollars to build chip factories,
- 20:29data centers and the nuclear power grids required to run
- 20:33them. These are infrastructure
- 20:34investments that take 10 to 15 years to show a return.
- 20:37Right. You cannot build a billion
- 20:39dollar data center overnight, no matter how smart the AI is.
- 20:43But here's the catch. If AI actually succeeds in
- 20:46compressing corporate lifespans to just five years, it forces
- 20:50the entire stock market to adopt A ruthless short term mindset.
- 20:55Investors and CF OS will only care about the next 18 months
- 20:58because anything beyond that is completely unpredictable.
- 21:01Meaning nobody will want to fund a 15 year infrastructure
- 21:03project. Exactly.
- 21:05By accelerating the pace of disruption, AI might ultimately
- 21:08destroy the very long term capital markets it relies on to
- 21:11fund its own physical evolution. It could trap itself in a
- 21:15paradox of its own making, starved of the capital it needs
- 21:18because it made the future too uncertain to invest in.
- 21:21Wow, it's like building an engine.
- 21:23So Powerful tears the car apart before you can even get on the
- 21:25highway. Yeah, that's exactly it.
- 21:27That is a fascinating thought to leave on.
- 21:29If you like to access wider topics, including life and
- 21:32investing philosophy, please subscribe to our sister channel
- 21:35at Deep Values. Don't forget to like this video.
- 21:37Much appreciated. We'll catch you on the next deep
- 21:40dive.