Chapter 12: The Illusion of Certainty
A young man once found himself in the presence of J.P. Morgan. Eager to make use of the moment, he asked the great financier what the market would do next. Morgan's reply became legend: "Young man, I believe the market is going to fluctuate."
The joke works because it's true, and because everyone who hears it laughs, then promptly forgets it. Wall Street spends fortunes trying to predict what Morgan refused to guess. It builds models, hires analysts, and studies charts, all in pursuit of certainty in a system that offers none.
Every decade or two, something breaks. The mechanism is rarely mysterious: banks and investors operate on leverage, borrowed money that magnifies gains on the way up and losses on the way down. When confidence is high, leverage feels like a tool. When confidence cracks, it becomes a weapon turned against its user. The specific trigger changes each cycle—a housing bubble, a tech collapse, a pandemic—but the underlying vulnerability does not. The only certainty is that positive thinking has no place in a leveraged position.
Behind this instability lies a deeper problem: the market is not, and cannot be, perfectly rational. George Soros spent his career attacking the efficient market hypothesis, the theory that prices always reflect true value because investors act on complete information. His objection was simple: "Any theory that assumes perfect knowledge is attainable is flawed." Investors do not operate on facts. They operate on beliefs about facts, and those beliefs are frequently wrong.
Soros called this reflexivity. Flawed assumptions drive action, and those actions shape the outcomes they were meant to predict. A stock rallies on belief rather than fundamentals: enough people expect it to rise, and their buying makes the expectation real, at least for a while. GameStop's 2021 surge was reflexivity on full display, a story about a squeeze becoming the mechanism that produced one.
This is where opportunity lives. If markets were truly efficient, there would be nothing to exploit, no mispricing to correct, no gap between perception and reality to trade against. But because markets run on collective belief rather than pure information, they can remain wrong for a long time. Sometimes years. In The Big Short, the investors who bet against the housing market were right about the collapse, but the market took four years to agree with them. Being correct and being rewarded are not the same event, and the distance between them can bankrupt the impatient.
Howard Marks, in Mastering the Market Cycle, explains why markets overshoot rather than settle: "The swing back from a high or low almost never halts at the midpoint, regardless of how 'right' or 'appropriate' the midpoint may be." The same optimism that drove a recovery keeps pushing past fair value, into overpricing. The pendulum does not pause at center; it swings to the opposite extreme.
1929
On October 29, 1929, the bubble that had defined a decade popped in a single afternoon. What followed was a collapse of the American financial system and the beginning of the Great Depression. The men who built that bubble never saw it coming, not because they lacked intelligence, but because they had convinced themselves the old rules no longer applied.
The mechanism was leverage. In 1929, the instrument was the stock market; in 2008, it would be subprime mortgages. What made the 1920s different was how leverage escaped the confines of Wall Street and moved into American living rooms. Credit had already transformed consumer life: families bought cars, refrigerators, and radios by putting down a fraction of the price and financing the rest. It was only a matter of time before that same logic reached the stock market.
Charles Mitchell made sure it did. Through National City Bank and its securities affiliate, he extended credit aggressively to everyday investors and brokerage firms alike, making margin accounts available to nearly anyone. A buyer needed only 10 percent down to purchase a stock, borrowing the rest against its future value. Brokers' loans ballooned from roughly $1 billion at the start of the decade to $8.5 billion by October 1929, more than the entire money supply of the United States. Ordinary Americans watched their net worth climb on borrowed money, and mistook leverage for wealth.
Even the House of Morgan, a name synonymous with conservative banking, caught the fever. The firm whose founder had refused to guess where the market was going now assembled syndicates to fund speculative ventures on enormous leverage. Customers wanted in, and the bankers who once served as gatekeepers decided it was better to profit from the mania than resist it.
When the crash came, it exposed how little substance had supported the boom. As John Kenneth Galbraith wrote, the collapse "broke into an exceptionally fragile financial, banking, and general economic structure." The crash did not simply reflect economic weakness; it caused it. Andrew Ross Sorkin has described what followed: the sharp decline in asset prices left too little equity to cover the debt piled on top of it, and that shortfall gutted the credit markets. Lenders who once extended margin loans freely now found no collateral they could trust, and credit, the lifeblood of any economy, vanished. Americans stopped trusting banks altogether, pulling their savings out and hiding cash in mattresses. Mass unemployment followed, then shanty towns, then breadlines. The Great Depression had arrived.
In 1929, no institutional safeguards existed to slow the descent: no SEC, no insider trading law, no capital requirements forcing banks to hold reserves. It took the creation of the FDIC in 1933 and the SEC in 1934 to rebuild that missing scaffolding, which is part of why 2008, devastating as it was, never metastasized into a second Great Depression. But not all the lessons stuck. By the mid-1950s, the fear had already faded, and a new boom was underway. It would take less than a decade for the market to prove just how short that memory was, when a fresh euphoria gripped Wall Street and set up the break of 1962.
1962
At 12:41 on Tuesday, May 29, 1962, the ticker tape at the New York Stock Exchange ran fifty-six minutes behind the trades it was meant to report. Brokers watched a machine tell them a selling panic was still underway, while on the floor beneath them, a buying panic had already taken hold. Telephone stock had traded at 101¼ a few minutes earlier, in the dollars-and-fractions pricing of the era; it was now changing hands at 103½. The machine meant to record reality could no longer keep pace with it.
The chaos had been building for a day. On Monday, the Dow had dropped 34.95 points, the second-worst single-day point loss in its history, trailing only the 38.33-point collapse of October 28, 1929. It came in the middle of an excellent business year, which left Wall Street's pundits inventing reasons for what had just happened. The truth was simpler and less comfortable: after twelve years of nearly uninterrupted gains, brokers and investors alike had grown convinced they were good at this. The market was about to argue otherwise.
What made 1962 genuinely dangerous, though, wasn't the margin selling that had gutted 1929. Federal credit regulations had made a repeat of that particular catastrophe nearly impossible; the exchange later estimated 91,700 margin calls went out that week, many by telegram in the dead of night, but the system held. The newer fear was mutual funds, which had grown from almost nothing in 1929 to twenty-three billion dollars in assets by 1962. If enough small investors demanded redemptions at once, the funds would be forced to dump stock to cover them, driving prices lower, triggering more redemptions, in a descending spiral nobody had ever seen tested at this scale.
Tuesday brought the reckoning. Rumors of firms on the verge of failure spread through the financial district, none of them true, none of them harmless for spreading anyway. The seventeenth-century trader Joseph de la Vega had diagnosed this centuries earlier: "The expectation of an event creates a much deeper impression than the event itself." Orders vanished into snowdrifts of paper on the exchange floor. Some brokers, unable to reach their partners or get a reliable quote, gave up and went out for a drink, unprofessional conduct that likely saved their clients money they'd have lost trying to trade blind.
Then, just before noon, with the Dow down twenty-three points and hovering at 553.75, the market reversed. Not gradually. Violently. General Motors ran from 45½ to 50 in under two hours. Standard Oil of New Jersey climbed from 46¾ to 51. It was the kind of rebound that invited comparison to October 30, 1929, when the market staged an almost identical bounce immediately before the slide that produced the Depression.
The mutual funds, it turned out, had behaved nothing like the villains everyone feared. Exchange data later showed they'd bought 530,000 more shares than they sold on Monday and sold 375,000 on the rebound Thursday, a pattern that stabilized prices rather than amplifying the panic. Nobody credits this to public-spiritedness. Fund managers were buying bargains and selling profits, the same instincts that drive any trader, and this time those instincts happened to calm the storm rather than feed it.
As for whether it could happen again, one Wall Street veteran put it plainly: of course it can. People grow careful for a year or two, then build toward another speculative run and another crash, and the cycle repeats itself until, as he said, "God makes people less greedy." Nothing about 1962 suggests he was wrong.
1987
In 1987, Michael Lewis was working on Wall Street at Salomon Brothers, hovering around the fortieth floor of One New York Plaza, when the market lost 22.61 percent of its value in a single day. Nobody on that floor had predicted it, and nobody, even hours later, could say why it was happening. Some brokers, rather than execute the sell orders panicking customers demanded, stopped answering their phones. If proof existed that Wall Street's insiders understood no more about the market's next move than anyone else, this was it.
The mechanism had a name, and it had been sold to them as a safeguard. Portfolio insurance was a strategy worked out by finance academics at Berkeley in the late 1970s and marketed hard through the decade that followed: software that promised to cap a fund's downside without requiring it to sell anything in advance. As prices fell, the program would sell stock index futures, offsetting losses in the portfolio underneath. As prices recovered, it would buy back in. A fund could stay fully invested and still, in theory, be protected. By the autumn of 1987, something on the order of $80 billion was managed under some version of it.
The flaw was not in the mathematics. It was in the assumption that you were the only one running it. One fund selling futures into a decline is a rounding error. Tens of billions of dollars selling on the same trigger at the same moment is the decline. Program trading had tied the futures pits in Chicago to the stock market in New York closely enough that selling in one dragged the other down, which tripped the next round of selling, which dragged it lower again. Every participant was following a strategy that was individually prudent and collectively suicidal.
On October 19, the loop closed. The Dow's 508-point fall was the visible number. Underneath it, S&P 500 futures dropped nearly 29 percent, well past the decline in the stocks they were meant to track, and the gap between the two markets grew wide enough that the arbitrageurs who normally close it stood back rather than step in. The exchange's automated order system buckled under volume it had never been built to carry. Prices on the tape stopped describing prices in the market, the same blindness that had gripped the floor in 1962, arriving faster and with more money behind it.
That crash, Lewis later wrote, "set in motion a process, weak at first, stronger over the years, that has ended with computers entirely replacing the people."
2008
Timothy Geithner had spent his entire career behind the scenes. He had dreaded public speaking since high school, and in February 2009, before a national television audience and with unemployment in Elkhart, Indiana, having quadrupled in a year, he had to stand at a podium and convince the country that its financial system would survive. He swayed like a passenger on an unsteady ship. He peered around the teleprompter and looked, in one commentator's words, like a shoplifter. Within days, an actor on Saturday Night Live opened the show by announcing that the new Treasury Secretary's plan was to hand $420 billion to whichever caller had a better idea.
But the speech was never the real test. The real test had begun months earlier, in the rubble of a collapse that had already erased $15 trillion in household wealth, thrown 9 million people out of work, and pushed stock markets to less than half their 2007 peak. Geithner counted five institutions still capable of detonating a second Lehman: Fannie Mae, Freddie Mac, AIG, Citigroup, and Bank of America. Each had already been rescued once. Each was in trouble again. His aides called Citi and Bank of America the "Financial Death Stars," and the math behind them was staggering: the banking sector alone might need $684 billion to survive an extreme scenario, against roughly $300 billion left in emergency funds.
Every instinct in Washington pushed toward punishment: nationalize the reckless banks, force bondholders to eat losses, make Wall Street pay for what it broke. Geithner understood the appeal. He called it the Old Testament view: the corrupt should suffer, the irresponsible should not be rescued. But he had watched panic move the way a contagion moves, and he knew that half-measures, however satisfying, would only widen it. "Every financial crisis is a crisis of confidence," he would later write. "A financial crisis is a bank run writ large, a run on an entire financial system. People lose confidence that their money is safe—whether they're stockholders or bondholders, institutional investors or elderly widows—so they rush to pull it out of the system, which makes the money remaining in the system even less safe, which makes everyone even less confident."
Confidence, not capital, was the real currency in a panic, a fact embedded in the word credit itself, which descends from the Latin for believe. Roosevelt had understood this when deposit insurance took effect in 1934: an insured depositor has no reason to start a run. That scaffolding was still standing in 2009, which is why the ground beneath Geithner was firmer than the ground beneath Hoover. His team built something similar for the giants of the new crisis, a "stress test" that sounded clinical but functioned as a verdict. Regulators would examine every major bank's books and calculate what it would need to survive a genuine depression. Firms that could raise the capital privately would do so; firms that couldn't would receive it from the government, whether the public liked it or not.
President Obama wanted speed. "We need to rip the Band-Aid off," he told Geithner. "I want to do this right, and get it over with." Geithner agreed, with a condition: "We do have to rip the Band-Aid off. But we have to make sure we don't break the financial system." Larry Summers put it to the President directly: a plan that shattered a fragile system and crashed the stock market would defeat its own purpose. The goal was never to shield Wall Street from consequence. It was to keep the damage from reaching Main Street, where laid-off workers, foreclosed families, and drained pension funds would pay the price.
Joseph Stiglitz, the former World Bank chief economist, dismissed the plan bluntly: "banks win, investors win, and taxpayers lose." Geithner never fully disputed the optics. He argued instead that justice and stability sometimes pull in opposite directions during a real emergency, and that the moral choice is not always the vengeful one. As Obama later put it, you don't refuse to send fire trucks into a burning neighborhood just to make a point about smoking in bed. The arsonists might escape their full punishment. The alternative, a second Great Depression, would punish everyone else far more.
None of it erased the damage already done, the homes lost, the jobs gone, the retirement accounts gutted by people who had never gambled a dollar on a subprime bond. But it confirmed the wager at the center of the approach: in a crisis built on lost faith, the only way out is to act with enough force to make that faith credible again, even if the world watching mistakes calm nerve for a bad performance.
2022
By November 2022, the wreckage was everywhere. FTX had collapsed. So had Three Arrows Capital, Luna, and Celsius. Billions in retail and institutional money vanished almost overnight. To anyone who had studied financial history, none of it should have been surprising: bank runs happen, and they always have.
Timothy Geithner had written with confidence just a few years earlier: "So far, the prophets of doom who have predicted runaway inflation, runaway interest rates, a double-dip recession, a collapse in demand for U.S. government securities, and other horrors for America have been false prophets." For over a decade, he was right. Then came 2022, which tested the first item on his list.
The setup had been building since the pandemic. COVID-19 triggered a global recession, and though its direct effects faded by 2022, it left behind an inflation surge and a supply chain crisis that took years to unwind. Russia's invasion of Ukraine in February strained an already unstable global market. By the time the Federal Reserve moved to raise interest rates in response to inflation, investor confidence was already cracking, because markets thrive on cheap money. Low rates let companies borrow, spend, and grow; raise them, and that growth engine stalls.
None of this was new. In 1929, a banker observed the same pattern with almost identical language: "I have heard thousands of reports of merchants, farmers, and men and women in all walks of life literally giving up their businesses to invest in the stock market. Most of them will, by necessity, have to go back to earning their living in normal ways." Crypto gave that story a new stage.
What made 2022 distinct wasn't the panic, but the absence of a safety net. In 2009, when the banks nearly collapsed, the government stepped in and absorbed the risk itself. That bailout didn't eliminate danger from the system; it relocated it. The too-big-to-fail banks earned that label because the government agreed to fail in their place if it ever came to that. Crypto has no such arrangement. When FTX and Celsius went down, there was no institution large enough, or willing enough, to catch the fall.
That absence is the design, not a defect in it. Crypto's premise is that no third party should have to hold your money for you: self-custody means the risk is yours alone, but so is the control.
Except that 2022 did not test that premise. It tested its abandonment. FTX and Celsius were failures of custody, not of self-custody. Their victims had handed their coins to an intermediary and then discovered it held no reserves, published no audited accounts, and answered to no regulator—a bank rebuilt without any of the guarantees that make a bank survivable. The lesson is not that crypto failed where banking succeeds. It is that banking's stability is borrowed, propped up by government guarantees, while crypto's instability is the honest price of independence. But that price only buys what it promises if you hold your own keys. Most of the people wiped out in 2022 had not opted out of the system. They had only changed its address.
This Time Is Different
On September 5, 1929, Irving Fisher told the world not to worry. The Los Angeles Times called him one of the nation's leading economists, and what people loved most about him was not his expertise but his certainty. Answering a rival forecaster who had predicted a coming collapse, Fisher allowed that there might be "a recession in stock prices, but not anything in the nature of a crash," and pointed to mass production and invention as proof that prosperity was structural now, not cyclical. Seven weeks later, the market began its collapse into the Great Depression.
Fisher wasn't a fool. He was a victim of a pattern old enough to have a name: the this-time-is-different syndrome. Economists Carmen Reinhart and Kenneth Rogoff, surveying eight centuries of financial crises, found the same mechanism at the root of nearly every one. Confidence is fragile, and short-term debt makes it lethal. Governments and banks can roll along for years looking invincible, and then, without warning, confidence collapses, lenders vanish, and the crisis arrives all at once. It happened to the United Kingdom's Northern Rock in September 2007, when depositors, unconvinced by a partial government insurance scheme, formed lines outside branches until the Treasury guaranteed their deposits in full. The bank was nationalized five months later.
Reinhart and Rogoff also dismantle the idea that crises can be engineered away. A lender of last resort doesn't eliminate risk; it invites banks to take on more of it, until even the rescuer goes broke. "Complete insurance against crises is neither feasible nor desirable," they write. The only defense is discipline: a government that runs surpluses, keeps debt low, and borrows long is largely immune to crises of confidence. One that does the opposite is a target waiting for a trigger.
What's remarkable isn't that crises happen. It's how quickly everyone forgets they ever did. Galbraith blamed "the extreme brevity of the financial memory." Disaster is absorbed, then dismissed, and within roughly twenty years, the time it takes for one calamity to fade and a new generation to grow confident in its own genius, the same idea resurfaces dressed as innovation. Wealth breeds a private myth of unique acumen; poverty breeds a public myth that the wealthy must possess one. Together, Galbraith argued, they form "an efficient design for concealing what, with the passage of time, will be revealed as self- and general delusion."
Crypto proved the pattern needed no new decade, only a new asset. In 2021, a year before the collapse, Dan Held, then Kraken's director of growth marketing, was building the loudest case in the industry for what he called the Bitcoin "supercycle": the theory that the old four-year boom-bust rhythm was finally broken, replaced by a permanent, adoption-driven surge toward prices as high as five hundred thousand dollars. Reckless money printing, collapsing trust in institutions, and infrastructure finally ready for mass adoption had converged, he argued, into "a perfect storm" unlike anything crypto had seen before. The market had heard that story before, in different words, in every decade for a century. Within a year the supercycle was in ruins.
Anyone who questions the story pays for it immediately, not eventually. In March 1929, Paul Warburg, a founding architect of the Federal Reserve and among the most respected bankers in the country, warned that "unrestrained speculation" would end in collapse and depression. The response was contempt: he was branded obsolete and accused of "sandbagging American prosperity." The critic is never right on time. Vindication arrives only after the crash, and by then no one is applauding.