This is What “Always” Happens Before a Market Crash

Stage One: The Delusion That This Time the Rules Changed

The video opens with a jolt: nearly $2 trillion erased in a single day, a national exchange forced to halt trading, and then, almost absurdly, the market racing back toward record highs as if nothing happened. That snapback is the first clue. The speaker says crashes do not come out of nowhere, they begin with a familiar mental trick, the same one people use when they talk themselves into buying a car they cannot quite afford or eating the second slice of cake they know they do not need. There is a quiet voice saying, this is a bad idea, and a louder one saying, it will be fine just this once. His argument is that whole markets can fall into that same mood.

He grounds that idea with the Buffett indicator, which compares the total value of the US stock market to the size of the US economy. He says it has reached roughly 238%, an all time record, and reminds viewers that Warren Buffett once said that when it gets near 200%, you are playing with fire. That matters because the market is not just expensive by ordinary standards, it is more stretched than at the top of the dot com bubble. He adds another layer: Wall Street analysts are projecting around 25% annual long term earnings growth for America’s biggest companies, a level of optimism that exceeds what investors were pricing in right before the dot com bust.

Then he makes the abstract concrete. Nvidia and Broadcom are trading at well over 20 times sales, not profits, sales. That only makes sense if growth keeps roaring without slowing, which is exactly the kind of assumption that looks safe until it suddenly does not. What unsettles him most is not ignorance. He says everyone knows these numbers. Buffett is sitting on a mountain of cash. Comment sections are full of people agreeing there is an AI bubble. Yet the money still pours in.

That is why he calls this first stage delusion rather than stupidity. He has sat through dinners where someone explains in detail why an asset is overvalued, then immediately says they bought more anyway. Fear of missing out overwhelms discipline. Even after nearly 40 years of investing, he admits he feels that itch too. The point is not that high prices alone cause crashes. The real danger begins when investors stop caring whether value matters at all.

Stages Two and Three: Complexity Hides the Risk, Debt Loads the Gun

He shifts from expensive markets to a more dangerous problem: the moment when finance becomes so complicated that even the people paid to understand it stop asking questions. He compares it to skimming a phone contract until your brain gives up and you just tick the box. In past crashes, that blind surrender showed up in different costumes. In 1987 it was portfolio insurance, a supposedly clever system that was meant to make crashes impossible but instead helped trigger one. In 2008 it was mortgages sliced up and repackaged until buyers no longer knew what they owned.

For today, he points to what he calls circular financing in AI. He asks viewers to picture three companies in a loop: Nvidia designs chips, OpenAI runs large language models, Oracle provides the cloud infrastructure. Nvidia invests billions into OpenAI. OpenAI commits hundreds of billions to Oracle. Oracle then spends that money buying Nvidia chips. The same dollars travel around the circle and, each time they pass through a company, they can be recorded as fresh demand and growth. By some estimates, around a trillion dollars in deals is now looping among a small cluster of companies like this. From the outside it looks like multiple independent transactions. In reality, part of the boom may be the same money being counted three or four times.

He strengthens the warning with a historical echo. In 1999, Lucent Technologies became the most widely held stock in America by financing customers who could not afford its telecom equipment. Lucent lent them the money, booked the purchases as revenue, looked brilliant for about two years, and then collapsed 99% once those startups could not pay. The lesson is painful and simple: sales are not the same as durable cash generation if you are effectively funding your own demand.

Then comes the accelerant. Complexity alone does not guarantee disaster, he says. Debt is what turns a correction into something like 2008. He brings the focus to private credit, lending that happens outside traditional banks, through large funds making direct loans that are harder for ordinary investors to inspect. AI related deals made up around a third of all private credit issued in 2025, versus an average of 17% over the prior five years. That is not a subtle increase, it is a huge concentration of risk.

The numbers get sharper from there. Morgan Stanley estimates defaults in private lending could rise to 8% if AI spending does not translate into revenue quickly enough. UBS says a rapid and severe AI disruption could push that figure to 15%. One measure of private credit defaults has already reached 6%, close to pandemic era stress. His most unsettling point is the catch built into the loans. Much of the money was lent against companies, like software firms, that AI is supposed to replace. If AI fizzles, the spending was wasted and loans sour. If AI succeeds too quickly, the underlying businesses get crushed and loans sour anyway. It is a lose lose setup hidden in a corner most people never look at.

Stages Four and Five: The Rescue Fantasy, the Tremor, and What to Do Now

Instead of asking why investors are scared, he asks the more interesting question: why are they still so calm? His answer is that the market has been trained. Over roughly the last 15 years, every serious decline has seemed to summon a rescue, whether through rate cuts, emergency lending, bailouts, or policy reversals. Investors learned the pattern the way a child learns a household rule. Fall hard enough, and someone catches you. Wall Street has a name for one version of this, the Fed Put, the belief that the central bank steps in when things get ugly. He thinks it has spread beyond the Fed into a broader faith that somebody with a big enough checkbook will not let the system break.

That belief explains why obvious warning signs can coexist with rising prices. If you truly think losses will be socialized and danger will be smoothed over, then risk stops feeling real. He says the problem in modern bubbles is not that people are asleep. It is that they are fully awake, can see the warnings, and choose to ignore them because they assume rescue is guaranteed. In his framing, that is the psychological engine of the whole cycle.

Then he turns to June’s selloff as a possible tremor. It started with memory chips, when investors began questioning the AI spending supporting the trade. Micron dropped around 13%. Samsung and SK Hynix fell around 12%. South Korea’s Kospi Index plunged 10% in one day, fast enough to halt trading, and close to a trillion dollars vanished. For a brief stretch it felt like the floor was opening. Then the panic faded, and within three weeks the S&P 500 had climbed back to the edge of its record high.

That pattern reminds him of March 2008, when Bear Stearns imploded over a weekend and had to be sold to JP Morgan for a fraction of its earlier value. Markets shuddered, then rallied for months as investors convinced themselves Bear Stearns was just a one off. Six months later, Lehman Brothers collapsed and the global economy went with it. His point is careful but pointed: Bear Stearns was not the crash, it was the tremor. June may or may not be the same kind of warning, and he is explicit that no one knows for sure. But uncertainty is not an excuse to tune out. It is a reason to stay alert.

He ends with the investing posture this history has taught him. He is still buying every month because trying to call the exact top is one of the costliest mistakes investors make. But he is not crowding into the same AI names everyone else owns. With the market leaning heavily on about seven companies, he has spread his money wider. He also admits to holding a meaningful cash position, not to abandon investing, but so that if a real crash arrives, he can be the person calmly buying while everyone else panics. That is his version of surviving the pattern: stay in the game, do not become delusional, and keep enough flexibility to act when fear finally returns.