Transcription
There is a sequence. It has run four complete times since 1900. Every single time it has run to its conclusion, what followed was a financial crisis of historic proportions. The kind that destroys retirement savings, collapses employment, wipes out decades of accumulated wealth, and reshapes the economic lives of ordinary families for a generation.
The sequence is not secret. It is not proprietary. It is documented in publicly available government data, in Federal Reserve publications, and in the historical records of every major financial crisis in the modern era. And right now, in 2026, the current data places us somewhere between stage two and stage three of that sequence.
Most people hearing that will feel one of two things. The first response is dismissal. Warnings like this have appeared before. The predicted crisis never arrived on the predicted schedule, and the person who issued the warning eventually stopped being invited to dinner parties.
The second response is a specific, quiet discomfort because somewhere in the data they have been half-watching, in the headlines they have been skimming, in the conversation they had with someone they trust, something already feels wrong, and they have not yet been able to name exactly what it is or precisely why it concerns them.
I want to introduce you to two men before we go into the sequence itself. Not because the history is interesting, though it is, but because these two men demonstrate in concrete, documented, financially precise terms what the sequence does and does not do for the people who understand it versus the people who do not. One of them made over $700 million in today's money from the worst financial crisis in American history. The other lost $150 million despite being the most celebrated monetary economist in the country. The difference between them was not intelligence. The difference was one thing, and I am going to show you exactly what it was.
Alfred Lee Lumis was 42 years old in 1929, an investment banker operating out of New York City. His firm, Bondite and Company, had spent the preceding five years financing approximately $1.6 to $6 billion in utility company mergers, expansions, and capital raises during the great industrial buildout of the Roaring Twenties. Lumis was not an outsider or a contrarian by temperament. He was embedded in the financial establishment of his era, but he had developed a habit over many years of tracking a specific category of data that most of his peers considered peripheral and unimportant. He tracked margin debt, the aggregate amount of money borrowed against stock portfolios across the entire market, a figure published regularly and available to anyone who wanted to find it. He had been watching that number for years, and he understood what it represented: the structural leverage underneath an asset market that had been rising for eight consecutive years.
In November 1928, margin debt peaked. The absolute level of borrowed money supporting stock positions reached its highest point and then began quietly and without fanfare to decline. Stock prices continued rising. The headlines continued to be bullish. The Dow Jones continued making new highs well into 1929, but the divergence had appeared. The leverage that had been fueling price appreciation was beginning to contract while prices continued to advance. Lumis recognized this divergence from historical patterns he had studied and he acted on it.
In February 1929, eight full months before Black Thursday, Alfred Lee Lumis and his partner Landon Thorne transferred every position they held into cash and United States Treasury bonds. Not a partial reduction, not a hedge, everything. When colleagues asked why, he described the margin debt divergence he had been watching. Most of them were politely dismissive. The market was still rising. The economy was still strong. The story of the Roaring Twenties was still being written in newspaper headlines every morning.
Eight months later, on October 24th, 1929, the market collapsed. Lumis was entirely out. His estimated profit in the years that followed, deploying cash into collapsed assets while panicked sellers needed liquidity at any price, was $50 million in nominal 1930s terms. Converted to today's money, that's over $700 million. He did not short the market. He did not use a sophisticated derivative instrument. He read a sequence that was written in publicly available data, recognized where in that sequence the market stood, and moved his money eight months before anyone else understood what was about to happen.
Now, meet Irving Fisher, Yale University economist, the most celebrated monetary economist in America in 1929, and arguably the most technically sophisticated analyst of monetary systems in the world at that time. He had written the textbooks that defined how the economics profession understood money, credit, and financial markets. He had built mathematical models of market behavior that were decades ahead of their time. He understood the mechanics of financial systems at a depth that virtually no one else alive could match.
On October 15th, 1929, nine days before Black Thursday, Irving Fisher stated publicly at a prominent business conference that stock prices had reached what looks like a permanently high plateau. This was not a careless remark. It was a considered analytical judgment from the world's foremost monetary economist based on a thorough examination of the available data through the lens of his theoretical framework. He lost an estimated $8 to $10 million in the crash. That's $150 million in today's money lost by the man who understood monetary economics better than virtually anyone on the planet.
The difference between Lumis and Fisher was not intelligence. Fisher was arguably the more intellectually formidable of the two. The difference was that Lumis was watching the sequence, and Fisher was watching the story. The story said the economy was fundamentally sound. The Roaring Twenties had delivered a 505% gain on the Dow Jones from 1921 to September 1929. Corporate earnings were strong. Industrial output was robust. The narrative logic of the story pointed to continued growth. The sequence pointed to a structural condition, leverage contracting beneath prices, that the story could not see and that Fisher's theoretical framework did not highlight as the primary warning signal.
Lumis made $700 million. Fisher lost $150 million. Same market, same crash, different tools for reading what was actually happening.
My name is Ban. I have been tracking this sequence for four months, mapping the patterns present in the 1929 data, the 2000 data, and the 2008 data against current readings from the Federal Reserve, FINRA's monthly margin debt publications, the FRED database, and the US Treasury's fiscal data portal. What I am about to show you is not a prediction, and I want to be precise about that distinction. Markets do not operate on fixed schedules. The sequence does not come with a countdown timer. What it provides is a framework for understanding structural conditions and a historically documented set of signals that have appeared in the data before every major financial crisis in the last 100 years.
The Federal Reserve's own researchers published a paper documenting one of these signals in 2006, two years before the 2008 crisis, which makes it entirely free of hindsight bias. I will show you that paper and what it means for the current data. The sequence has five stages. Knowing which stage you are in is the operational difference between being Alfred Lee Lumis and being Irving Fisher, between making $700 million and losing $150 million. Between being prepared and being the person who gets liquidated.
Stage one is institutional divergence. Smart money begins reducing exposure while public sentiment remains bullish and prices continue rising. The specific, trackable signal is this: aggregate margin debt begins declining while stock prices continue making new highs or holding near them. These two data series, which normally move in the same direction because leverage fuels price appreciation, begin moving in opposite directions. The divergence is the signal. Most retail investors do not track margin debt. They track price. Price is still up. Everything looks fine from the direction they are looking. Lumis was looking at the divergence. He saw stage one begin in November 1928. He acted in February 1929. He was three months into stage one when he moved.
Stage two is liquidity tightening. Credit conditions begin tightening. Banks call in margin loans more aggressively. The Federal Reserve raises interest rates to address speculative excess. In August 1929, the Fed raised its discount rate to 6%. This was not front-page news. It was a technical policy adjustment, but it raised the cost of holding leveraged positions and reduced the supply of new credit available to fuel further price appreciation. Stage two does not produce a crash. It removes the fuel that has been keeping prices elevated and creates the structural conditions under which the stage four catalyst can cause maximum damage. An eight-year bull market built on expanding leverage does not unwind gently when the leverage stops expanding.
Stage three is volatility escalation. The market becomes erratic. Large single-day moves begin appearing with increasing frequency in both directions. In September 1929, the Dow Jones had 11 separate sessions with moves of 1% or more, an unusual concentration of daily volatility that, in retrospect, was a clear sign of structural instability. Investors at the time interpreted this as healthy market activity, normal choppiness before the next leg higher. The sequence said the structural conditions for a major crisis were now fully in place, and the catalyst could arrive at any time. Stay with me because this is where the pattern gets most important to understand clearly.
Stage four is the catalyst event. A single session with a decline of 5% or more on above-average volume. In 1929, this was October 24th, Black Thursday. In 2000, it was the NASDAQ's collapse beginning in March. In 2008, it was the sequence of events in September and October, culminating in the post-Lehman freezing of global credit markets. Stage four is what most people identify as the beginning of the crisis. But from the perspective of the sequence, stage four is the end of the preparation window. By stage four, Lumis had been out for eight months. The 847 families documented in the Brookings Institution's 1934 Wealth Survey who came through the crash not just intact but significantly wealthier had made their moves in stage one or stage two. By stage four, the window was closed, and all that remained was watching from a position of safety while 12 million people lost their jobs and 9,000 banks failed.
Stage five is redistribution. Cash-holding families and investors deploy their liquidity into collapsed assets at prices that will not be seen again for decades. This is the stage that transforms crisis protection into actual wealth creation. James Mercer, 38 years old, an oil field roughneck from Midland, Texas, not a sophisticated investor, not a financial professional, used $80 in cash in 1931 to lease mineral rights on 40 acres at $2 per acre. The pre-depression market price had been $15 to $20 per acre. Galbraith had $80 of cash when everyone around him needed cash desperately enough to sell valuable assets at any price they could get. In 1939, an adjacent Perian Basin oil discovery turned those mineral rights into an offer of $4,200. His $80 became $4,200, a 5,150% return in eight years. Not because he was a sophisticated investor, because he had cash in stage five when everything around him was priced at stage five values.
Thomas Mercer owned a hardware store in Galina, Illinois. He is documented by name in the WPA Federal Writers Project oral history interviews recorded in 1938. He had $2,400 in savings in 1929. He read a Chicago Tribune article describing banks calling in margin loans and making credit harder to access. He didn't fully understand the mechanism, but he understood enough to feel uneasy and to act on that unease. He converted $1,800 of his $2,400 into postal savings certificates and kept $600 in physical cash at home. His neighbor, William Grady, a school teacher with $2,100 deposited at First State Bank of Galina, lost everything when that bank failed in November 1930. Thomas Mercer emerged from the depression with $1,800 intact, plus two years of compounded interest. He had stumbled into stage two preparation through practical common sense and one newspaper article without ever knowing the word "sequence" or the name "Lumis" or the concept of "institutional divergence." He had been watching one signal, imperfectly understood, and it was enough.
Irving Fisher had built the most sophisticated monetary analysis framework of his generation. Thomas Mercer had read a newspaper article and felt uneasy. Thomas Mercer came through intact. Fisher lost $150 million in today's money. The sequence rewards correct positioning. It does not reward analytical sophistication for its own sake.
Now, here is what the current data looks like mapped against the five stages. And I want to be specific about what is confirmed in the data, what is approaching confirmation, and what has not yet arrived.
Stage one signal: margin debt divergence. US NYSE margin debt peaked in late 2021 and declined substantially through 2022. While equity prices initially continued near highs before falling, it recovered through 2023 and 2024 alongside recovering prices. The classic stage one pattern was clearly present in 2022. The current reading is less clean than the 1929 signal, but the FINRA monthly data shows elevated margin debt volatility relative to price, consistent with a market that has experienced stage one conditions and is working through their consequences. FINRA publishes this data monthly at finra.org. It is free. It takes five minutes to check.
Stage two signal: liquidity tightening. The Federal Reserve raised its benchmark rate from 0.25% in March 2022 to 5.5% by July 2023. That is a 5.25 percentage point increase in 16 months, the most aggressive tightening cycle since 1980. Rates were then held at 5.25% to 5.5% for the longest sustained period since 2007. The Federal Reserve's own senior loan officer survey showed bank lending standards tightening significantly through 2022 and 2023. Stage two readings are confirmed. The tightening happened. Its full effect on leveraged positions and credit-dependent asset prices works through the system on a lag measured in months to years, not days.
Stage three signal: volatility escalation. Single-day moves of 2% or more in the S&P 500 have become more frequent than in the 2019 or 2021 periods. Corporate credit spreads have been widening. The VIX has shown elevated readings on a more consistent basis. These readings are consistent with stage three conditions. The data does not yet provide the kind of clean confirmation that the September 1929 pattern provided in retrospect, but the structural fingerprint is present.
Stage four has not happened. Stage five has not happened. The preparation window remains open. It is narrower than it was in early 2022 and it will not stay open indefinitely, but it is open.
There are three specific data signals that I track for this sequence that any person watching this can track themselves for free without paying for any service or subscription, using data published by the Federal Reserve, FINRA, and the US Treasury.
The first is NYSE margin debt, published monthly by FINRA at finra.org. When margin debt falls more than 15% from its most recent peak while stock prices remain within 5% of all-time highs, that is the stage one divergence signal. That single indicator preceded the 1929 crash, the 2000 dot-com collapse, and the 2008 financial crisis. Check it once a month.
The second is the Federal Reserve's H.4.1 statistical release, published every Thursday at federalreserve.gov. It shows in real time what the Fed is doing with monetary liquidity, whether it is expanding or contracting the monetary base. Consistent contraction over multiple consecutive weeks means the fuel supply to leveraged asset markets is tightening. This is the modern equivalent of the Fed raising rates to 6% in August 1929. Available every Thursday, free.
The third signal is the one documented by the Federal Reserve's own researchers before the 2008 financial crisis. Arturo Estrella and Mary Truby of the Federal Reserve Bank of New York published their analysis in the Current Issues in Economics and Finance journal in July and August of 2006, two full years before Lehman Brothers collapsed in September 2008. Their paper examined the yield curve as a leading recession indicator and documented its predictive record going back to 1955. The finding was unambiguous: every recession since 1955 has been preceded by a yield curve inversion, in which the two-year US Treasury yield exceeds the ten-year yield. Zero false positives in 70 years of data.
The current inversion began in July 2022. As of early 2026, it has been running for approximately 42 months, the longest sustained inversion in the modern data series by a significant margin. But here is the critical point that almost every financial commentator misses when they discuss the yield curve: the inversion itself is not the action signal. The uninversion is. When the yield curve crosses from negative territory back to zero after a sustained inversion, when it appears to be returning to normal, that crossing has historically been the final warning before recession arrives. Every instance of uninversion following a sustained inversion in the post-1955 record followed recession within three to six months. The apparent return to normal is the danger signal, not the safety signal. Watch FRED T10Y2 when it crosses back through zero after this record 42-month inversion. That is the stage three to stage four transition signal. Not the inversion, the uninversion.
So, what does the Lumis position look like today in 2026? Expressed in instruments that exist right now. The iShares Short Treasury Bond ETF, ticker SHV, is the modern equivalent of what Lumis moved into in February 1929. The shortest available duration in US government paper through a single, liquid ETF. Near zero default risk, fully convertible to cash within one standard trading session in any market condition, including severe credit stress. This is not an investment that will make you wealthy. It is a position that will still be there in its entirety when stage four arrives and that can be deployed in stage five at the prices that stage five produces. Lumis' treasury bonds in February 1929 did not make him wealthy. The cash they represented in 2030, 2031, and 2032, deployed into collapsed assets, is what made him $700 million.
SPDR Gold Shares, ticker GLD, at five to eight percent of total portfolio value. Lumis held physical gold in 1928 as a secondary store of value maintained outside any institutional balance sheet. GLD provides this function in modern form, with the Perth Mint Physical Gold ETF, ticker AAU, as an alternative that holds physical gold in an Australian government vault entirely outside the US banking system. Five to eight percent is insurance. It is not a position. The distinction matters because insurance has a defined, limited cost and a defined protective function. A position is something you size for return. Gold at five to eight percent is the former.
For stage five preparation, the moment Lumis deployed cash into collapsed assets, the moment Galbraith leased mineral rights at $2 an acre, the Vanguard Real Estate ETF, ticker VNQ, falling below $70, represents approximately a 40% decline from its 2021 high of approximately $115. Historical data covering every REIT index decline of 40% or more since 1993 shows full recovery within five years in every single instance. When VNQ hits that level, it is the Perian Basin mineral lease opportunity. It is stage five. The entry point is defined. You either have the cash to act on it, or you do not. Lumis had the cash because he moved in stage one. Galbraith had the $80 because he had not borrowed against his savings and had not put everything into a bank that failed.
Let me bring this back to where we started. Alfred Lee Lumis sits in his private office on October 29th, 1929, Black Tuesday. One floor below him on the New York Stock Exchange trading floor, Harold Biggs is watching the ticker tape run 68 minutes behind a market losing $14 billion in a single session, listening to the sound of a waterfall become a roar that presses against the walls. Lumis' partner, Landon Thorne, is reading the morning paper. Their positions have been entirely in cash and treasury bonds for eight full months. Thorne feels the floor of the office vibrate slightly from the volume of noise coming up from the trading floor below. He turns the page.
Irving Fisher, that same morning, is watching $150 million in today's money evaporate from a portfolio managed by the most sophisticated monetary economist in America. He had the better analytical framework for understanding monetary systems in the abstract. He had the worst tool for reading where the sequence stood in the specific. He watched the story. The story said "permanently high plateau." The sequence said "stage four has arrived."
Thomas Mercer opens his hardware store in Galina, Illinois, on a Tuesday morning in November 1930, the morning after First State Bank of Galina failed and his neighbor William Grady lost $2,100. Thomas has $1,800 in postal savings certificates and $600 in cash at home. He did not understand the yield curve. He did not track margin debt. He read one newspaper article, felt genuinely uneasy, and made one practical decision in response to that unease. He came through intact not because he was sophisticated, but because he acted on a signal at a time when most people around him were still reading the story.
847 families, the Brookings Institution's best estimate from their comprehensive 1934 survey of American income and wealth, came through the worst financial crisis in American history not just intact, but significantly wealthier. They were not all Alfred Lee Lumis. Most of them were closer to Thomas Mercer: hardware store owners, grain elevator operators, postal savings certificate holders, people who had made one or two practical decisions at the right moment, sometimes without fully understanding why those decisions were correct, and who found themselves in stage five with cash and productive assets while 12 million people stood in unemployment lines.
The sequence is running. Stage two is confirmed in the historical data. Stage three readings are present and consistent. The window is not closed. What it requires is not genius or sophisticated analysis. It requires looking at the right data instead of the right story. Lumis looked at margin debt divergence in November 1928. Thomas Mercer read a newspaper article about margin loans in 1929. Both came through. Fisher had the better analysis and worse signals. He did not.
If this gave you a framework you did not have before, share it with someone who is still reading the story and has not yet looked at the sequence. Not after stage four. Now, this is Ban. The sequence is.