The History of the Credit Cycle: Why Booms Always End Badly

On the morning of September 24, 1720, Isaac Newton sold his remaining shares in the South Sea Company, crystallizing losses of approximately £20,000 — around £3 million in today’s money, and a meaningful fraction of his lifetime savings. Newton had actually been a savvy early investor, buying in at the beginning of the year and selling in April for a tidy profit. Then he watched friends and acquaintances get richer as the share price continued to climb, bought back in near the top, and rode the collapse all the way down. “I can calculate the movement of stars,” he reportedly said afterward, “but not the madness of men.” Newton was one of the most powerful analytical minds in human history. The credit cycle took him anyway.

The credit cycle — the rhythmic alternation between expansion and contraction in the availability of credit, the willingness to lend, and the price of risk — is the most persistent and most destructive pattern in all of economic history. It has occurred in every era, in every economic system that has used credit at scale, with variations in speed and amplitude but with a structural logic so consistent that calling it a cycle is almost an understatement. It’s more like a law.

The Mechanics of Expansion

Credit expansions don’t begin with fraud or irrationality. They begin with something real. The South Sea bubble started with a genuine commercial opportunity: monopoly rights to trade with Spanish South America. The Mississippi bubble, which gripped France simultaneously and collapsed in the same year, began with John Law’s genuinely innovative ideas about paper money and the potential of Louisiana territory. The 1990s technology boom was rooted in a real transformation of commerce and communication. The 2000s housing boom was anchored in a genuine secular decline in interest rates and a genuine increase in American homeownership. Every boom has a real kernel. That kernel is what makes it so seductive.

The sequence, once launched, is mechanical. Rising asset prices produce paper wealth. Paper wealth increases the collateral value of existing holdings. More collateral means more borrowing capacity. More borrowing drives more purchases. More purchases push prices higher still. At each step, the investment looks more validated: the asset has gone up, which seems to confirm the original thesis. Lenders who previously demanded caution begin to compete for business. Borrowers who previously showed restraint find that caution is leaving money on the table. Standards slip not because everyone becomes stupid, but because the incentive structure shifts: in a rising market, the risk of being too conservative — of missing gains your competitors are capturing — begins to outweigh the risk of being too aggressive.

Hyman Minsky spent a career explaining this dynamic, and his framework remains the most precise account of what happens next. He divided borrowers into three types: hedge borrowers, who can service both principal and interest from cash flows; speculative borrowers, who can service interest but must roll over principal; and Ponzi borrowers, whose cash flows cover neither interest nor principal and who depend entirely on continued asset appreciation to stay solvent. In an expansion, the composition of the borrowing pool shifts steadily from the first category toward the third. This shift is individually rational at every step. It is collectively catastrophic.

Why No One Stops It

The question that every history of financial crises implicitly raises, and that most accounts handle inadequately, is why the expansion is never stopped before it becomes dangerous. The answer is not that regulators are captured, executives are venal, or investors are deluded — though all three are often true at the margins. The answer is structural: the people positioned to stop the boom are the same people who profit from it, and the moment at which intervention would be most effective is the moment at which the case for intervention looks weakest.

Consider the epistemic problem facing a bank regulator in year three of a housing boom. Asset prices have risen substantially. Default rates remain near historical lows. Lending standards have loosened, but every loan made in the past three years has performed. The economy is growing. Employment is rising. Every data point in front of the regulator describes a healthy, expanding system. The warning signs are structural — ratios of debt to income, price-to-rent multiples, the percentage of originations that are Ponzi-category — but these structural measures are routinely discounted in expansions because the same people presenting them got the last expansion wrong too. The regulator who calls a top early looks like a fool for two years before they look like a prophet. The political economy of that situation reliably produces inaction.

The banks face the same problem from the inside. The executive who in 2005 told the CEO of a major investment bank that their mortgage portfolio was too large and their risk too concentrated was not wrong. He was also not there in 2007, because by 2005 the portfolio was generating record revenues and the executive who cut it back was leaving money on the table while competitors gained market share. Chuck Prince, then CEO of Citigroup, gave the most honest description of this dynamic ever offered by a senior banker: “As long as the music is playing, you’ve got to get up and dance.” This is not an admission of irresponsibility. It is an accurate description of what competitive markets do to institutional risk management when everyone else is dancing.

The Minsky Moment and Its Aftermath

The turn always comes from somewhere unexpected — or rather, the proximate trigger is always unexpected even though the underlying fragility was not. The Dutch Tulip Mania of 1637 turned on a single failed auction. The Overend, Gurney & Company collapse of 1866 — then the largest bank failure in British history — turned on a handful of bad railroad loans. The 1929 crash turned on a rate decision. The 2008 collapse turned on a pool of subprime mortgages in Arizona and California that turned out to be worthless.

What the trigger reveals is not a cause but a condition: a system that had become so fragile, so dependent on continued appreciation, so saturated with Ponzi borrowers, that any sufficiently large shock would detonate it. The trigger is irrelevant in the way that the specific spark is irrelevant in a room full of gunpowder. The relevant question is how the gunpowder accumulated.

Once the turn comes, the same mechanisms that produced the expansion go into reverse. Falling asset prices reduce collateral values. Reduced collateral forces margin calls and deleveraging. Deleveraging means selling assets. Asset sales push prices lower still. Lenders who competed aggressively for business begin demanding repayment. Borrowers who depended on rolling over debt find the window has closed. The composition of defaults shifts upward through the borrowing hierarchy: first the Ponzi borrowers, then the speculative borrowers, finally even some of the hedge borrowers who find that their supposedly safe cash flows have dried up in the general contraction. Irving Fisher called this dynamic debt deflation, and his 1933 paper describing it remains one of the most accurate and underread documents in economic history.

The Policy Response Paradox

Every major credit contraction produces demands for policy intervention, and every major intervention produces controversy. The debate is usually framed as a moral question — whether failing institutions deserve to be rescued — but the functional question is different: whether allowing the collapse to complete itself will destroy economic capacity that took decades to build and will take decades to replace.

The historical record on this question is reasonably clear. Countries that intervened aggressively to stop deflationary spirals — Sweden in 1992, the United States in 2009 — recovered faster and suffered smaller permanent output losses than countries that allowed the contraction to run its course. The American experience of the 1930s, during which the Federal Reserve contracted the money supply in the middle of a banking crisis and presided over the failure of thousands of solvent banks along with insolvent ones, produced a decade of depression that intervention would very likely have shortened. This is not a comfortable conclusion for those who believe that boom-bust cycles are morally self-correcting — that the pain of the contraction is a necessary punishment for the excesses of the expansion. The evidence doesn’t support that view. The pain falls disproportionately on people who had nothing to do with the excesses.

The moral hazard problem is real: if creditors and shareholders are always rescued, the incentive to manage risk prudently disappears. But moral hazard is a long-run problem. Debt deflation is a short-run catastrophe. The policy paradox of financial crises is that the short-run correct response — aggressive intervention — makes the long-run problem worse. There is no clean solution. Anyone who tells you otherwise is selling something.

What the Cycle Teaches

Isaac Newton’s £20,000 loss is instructive not because Newton was foolish but because he was extraordinarily smart and lost anyway. The credit cycle is not a story about greed or stupidity. It is a story about the structural properties of credit-based economies: the way that leverage amplifies both gains and losses, the way that competitive pressure erodes risk management, the way that paper wealth generated in an expansion feels as real as wealth created by production until the moment it doesn’t.

The cycle will continue. This is not pessimism; it is realism about incentive structures. Credit expansions produce paper profits that are distributed in real time to people with the power to demand more of them. Credit contractions produce losses that are distributed to people without the political power to prevent them. The regulatory tightening that follows each crisis is gradually unwound over the following decade by the political pressure of an industry with very strong incentives to expand and a public with very weak incentives to pay attention until the next disaster.

The honest policy agenda is not to prevent the cycle — that would require preventing credit itself, which would mean preventing most of the economic growth credit has historically enabled. The honest agenda is to reduce amplitude: more capital, better counter-cyclical buffers, cleaner resolution mechanisms for failed institutions, and an unsentimental willingness to tighten standards when the music is playing and everyone wants to dance. None of that is exciting. None of it will be popular when the next boom is underway. But it is, as Newton might have put it, the calculus of the situation — even if the madness of men repeatedly makes us forget the answer.