1. The Thirty-Seven-Year Gap
In 1973, Fischer Black and Myron Scholes published a formula for pricing derivatives.
The Black-Scholes formula was not abstract mathematics. With this equation, option prices could be calculated with theoretical precision — which meant entirely new financial instruments could be created. Risks could be isolated and traded. A market worth tens of trillions of dollars was built on top of this formula — though the Black-Scholes equation alone did not create the derivatives market. Interest-rate liberalization in the 1970s, financial deregulation in the 1980s, and computerization in the 1990s all played their part. But without the mathematics, the market could not have grown this fast.
It was technology. Financial technology.
On July 21, 2010, President Barack Obama signed the Dodd-Frank Act into law. For the first time, comprehensive regulation began to apply to the OTC derivatives market.
From 1973 to 2010: thirty-seven years.
In Book 3, we looked inside this gap. When Michael Burry read subprime MBS prospectuses from start to finish — documents no one else had read — he found cracks in the system. But institutions had no ears to listen. Neither Brooksley Born's warning nor Michael Burry's bet could compel an institutional response.
The information existed. The signals existed. People were sounding alarms. Institutions failed to process the signals.
2. Why Thirty-Seven Years?
In Chapter 1, in ancient Rome, we saw incumbent capture. The Senate was both the beneficiary of the latifundia system and the decision-maker on its regulation.
In finance, the same structure repeats — only more sophisticated.
The 1998 counterattack by Greenspan, Rubin, and Levitt was not crude conflict of interest. They genuinely believed in market self-regulation. Adam Smith's laissez-faire ideology had been resurrected in the language of twentieth-century finance as the "efficient market hypothesis." The logic was structurally identical to the British factory owners in Chapter 2 who invoked "freedom of contract" to block child-labor regulation.
Ideology legitimizes interest. And ideology outlives interest.
In 2000, Congress passed the Commodity Futures Modernization Act (CFMA). The law explicitly exempted OTC derivatives from the regulatory jurisdiction of both the CFTC and the SEC — meaning credit default swaps (CDS), interest-rate swaps, and most other over-the-counter derivatives could be traded without federal oversight. It codified the regulatory vacuum into law. The bill was quietly inserted into an 11,000-page omnibus spending bill on the final day of the congressional session. There was no separate committee review, and most members who voted on it had not even read the provisions.4
Warnings had been issued. Brooksley Born's alarm (1998). The LTCM collapse (1998) — a hedge fund built by Nobel laureates came within inches of bankruptcy, forcing the Federal Reserve to intervene. The Orange County bankruptcy (1994) — a California local government invested in derivatives and lost approximately $1.7 billion.
Signals kept recurring. Institutions kept ignoring them.
There were structural reasons. Of the five mechanisms analyzed in Chapter 1, three were operating simultaneously.
Incumbent capture: The financial industry was one of the largest lobbying forces in America. The more regulation was relaxed, the greater the profits. More lobbying money bought more deregulation.
Ideological barriers: The efficient market hypothesis — "the market regulates itself" — was the language of mainstream economics. Anyone who called for regulation was dismissed as someone who "doesn't understand markets."
Information asymmetry: The structures of CDOs (collateralized debt obligations) and CDS (credit default swaps) were too complex for regulators to understand. The prospectuses that Michael Burry read — those two-inch-thick documents — regulators never touched. They had neither the personnel nor the incentive to read them.
3. It Takes Blood to Move
On September 15, 2008, Lehman Brothers filed for bankruptcy.
A 158-year-old investment bank collapsed overnight. The next day, AIG requested a government bailout. The following week, the global financial system came close to paralysis. Stock markets crashed, credit markets froze, and companies faced the prospect of being unable to make payroll.
The crisis was dramatic — sudden, visible, catastrophic. In the United States alone, between 2008 and 2012, approximately four million completed home foreclosures were recorded.5 Some tallies, counting foreclosure filings rather than completions, exceeded ten million. Losing a home was not a statistic. It meant a child's school, the faces of neighbors, an entire credit history — all gone at once.
Behind every lost home stood a lost job. From September 2008 through the end of 2009, more than eight million jobs vanished in the United States alone.6 Manufacturing workers, construction laborers, bank tellers — people who did not understand the architecture of derivatives, and never needed to. They did not know that the risks they would bear were being engineered dozens of floors above them.
Later, in Chapter 9, we will see algorithms making hiring and firing decisions. A layoff notice arriving on a smartphone just before lunch. A counselor with twenty years of service losing her job to a single word: "efficiency." There too, the people who bear the harm do not understand the structure behind the decision. Those who design the system and those who absorb its consequences never meet. The pattern repeats.
Five weeks later, on October 23, Alan Greenspan sat before the House Committee on Oversight and Government Reform. He was the man who, as Fed chairman for eighteen years, had shaped the rules of the financial world. Chairman Henry Waxman asked: "You were the greatest advocate for the position that markets could self-regulate. My question for you is simple. Were you wrong?"
Greenspan answered: "Yes, I found a flaw. I don't know how significant or permanent it is, but I've been very distressed by that fact."7
Waxman pressed: "In other words, you found that your view of the world, your ideology, was not right?"
"Precisely. That's precisely the reason I was shocked, because I have been going for forty years or more with very considerable evidence that it was working exceptionally well."
It was the moment an ideology held for forty years officially collapsed in a congressional hearing room. He called it "a once-in-a-century credit tsunami." Arthur Levitt, the former SEC chairman who had joined Greenspan in blocking Born's warning, later admitted as well: "I was dead wrong. Alan Greenspan was dead wrong. Bob Rubin was dead wrong."8
That was the official reckoning. It came ten years after those spring nights in Washington in 1998, when Brooksley Born had sat writing her warning. For the four million households that lost their homes during those ten years, the admission offered no consolation.
Twenty-two months later, in July 2010, the Dodd-Frank Act was passed. All 2,300 pages of it. OTC derivatives regulation, the Volcker Rule (restricting banks' proprietary trading), and the creation of the CFPB (Consumer Financial Protection Bureau). Institutions that had not moved for thirty-seven years moved in twenty-two months.
It takes blood to move. But once blood is drawn, institutions move fast.
This is the pattern. When the Dow plunged 22.6 percent in a single day on Black Monday in 1987, circuit breakers — automatic trading halts — were introduced within months. When the Flash Crash of 2010 sent the Dow plummeting 998 points in nine minutes, improved circuit breakers were implemented that same year.
Before crisis: thirty-seven years. After crisis: twenty-two months, or a few.
4. Ten Days of Collapse
November 2, 2022. CoinDesk published a report revealing that the balance sheet of Alameda Research, an affiliate of the cryptocurrency exchange FTX, was composed primarily of FTT — a token issued by FTX itself.9
On November 6, Binance CEO Changpeng Zhao (CZ) announced on Twitter that Binance would liquidate its entire FTT holdings, worth approximately $580 million.
On November 8, FTX effectively halted withdrawals. Over $6 billion in withdrawal requests had flooded in within seventy-two hours, and the system could not keep up. The same day, Binance signed a memorandum of understanding (MOU) to acquire FTX, but withdrew after due diligence just one day later, on November 9.
On November 11, FTX and roughly 130 affiliated companies filed for Chapter 11 bankruptcy protection in a Delaware federal court. Sam Bankman-Fried resigned as CEO, and restructuring specialist John J. Ray III was appointed as his successor.
Ten days. A company valued at $32 billion vanished in ten days. More than $8 billion in customer funds had been siphoned to affiliates. It was fraud. But what made the fraud possible was a regulatory vacuum.
The irony ran deeper. Just two months before the collapse, in September 2022, Bankman-Fried had publicly declared that "stricter regulation could prevent the next crypto crisis." He appeared before Congress to advocate for CFTC oversight of cryptocurrency — the picture of an industry leader championing sensible regulation.
Immediately after FTX collapsed, Vox reporter Kelsey Piper sent Bankman-Fried a direct message on Twitter. She asked whether his advocacy for regulation had been sincere.
"Yeah, just PR."10
In the same conversation, he added: "F\\\* regulators. They make everything worse."11 The gap between his face at the congressional witness table and his face in private messages — that was the structural limit of industry self-regulation as a concept.
On October 3, 2022, FSOC (Financial Stability Oversight Council) had flagged the "absence of direct federal oversight of the crypto spot market" as the most critical gap in its official report.12 Thirty-nine days after that report was published, FTX collapsed.
The warning was precise. Congress was nowhere.
At the end of Chapter 12 in Book 3, we returned to the story of the Medici bank's branch managers. For six hundred years, the core question of finance had been the same: "Can this person repay?" On Wall Street in 2005, the people asking that question had disappeared. That was the real cause of the 2008 crisis.
FTX was no different. The fundamental question had vanished: "Do this exchange's customer funds belong to the customers?" With no regulation in place, no one was asking.
5. Regulatory Time Resolution
In Chapter 3 of Book 3, we observed a credit review committee at a savings bank on the outskirts of Gyeonggi Province (경기도, the region surrounding Seoul). It was the early 2010s, the heart of a Korean financial crisis triggered by a cascade of savings-bank failures as real-estate project-financing (PF) loans turned sour.
BIS ratio. PF delinquency rate. Quarterly reports. The committee chairman scans the documents spread across the table. The Yongin construction project's site completion rate is listed at 93 percent. But the fact that the developer has stopped returning calls — that never shows up in the paperwork. You learn that only by visiting the site in person.
This is the problem of regulatory time resolution.
Reality moves daily. Regulation measures quarterly. In the gap between the two, crises grow.
LTCM's leverage ratio in 1998 fell within quarterly regulatory limits. In real time, the picture was different. Lehman's capital ratios in 2008 looked fine on paper. But while the mortgage market was collapsing day by day, the next quarterly report was still a month away. CEO Richard Fuld told shareholders in April 2008 that "the worst is behind us." Five months later, the company was bankrupt. Chief risk officer Madelyn Antoncic had recommended reducing mortgage exposure but was ignored, and Senior Vice President Matthew Lee, who flagged accounting irregularities, was fired days after sending his letter. The numbers on paper were healthy. Reality was crumbling.
AI could change this equation. Real-time cross-verification of construction-site drone footage, materials-delivery data, and concrete-pouring records could detect the signal — "the reported completion rate doesn't match the site" — on a weekly basis rather than at quarter's end.
But implementing such an AI monitoring system requires: a legal basis for regulation, budget allocation, system development, a pilot program, and full-scale deployment. A minimum of three to five years.
In the interim, the next crisis grows. The paradox of speed.
6. From Thirty-Seven Years to Six Months
The gap is compressing.
From Black-Scholes to Dodd-Frank: thirty-seven years. From Bitcoin to the GENIUS Act: seventeen years.13 The EU AI Act took effect thirty-three months after the appearance of ChatGPT. China issued generative-AI regulations in eight and a half months.
Thirty-seven years → seventeen years → thirty-three months → eight and a half months.
But the pace of technological change is accelerating alongside it. The Black-Scholes formula appeared in 1973, and the derivatives market took a decade to mature. ChatGPT launched in November 2022, and by early 2023, one hundred million people worldwide were using it. AI models turn over generations on a six-month cycle.
While regulators draft the EU AI Act's detailed guidelines, the AI systems they aim to regulate have already changed three times over.
The absolute duration of the gap has shrunk. But the relative distance between technology and regulation remains as large as ever — or larger.
This is the central tension that Book 4 confronts. During the thirty-seven-year gap in financial regulation, tens of thousands of child factory workers were sacrificed. What is being sacrificed in the AI-era gap? The fact that the harm may not look dramatic enough to trigger a response is precisely what makes it more dangerous.
7. Why We Cannot Wait for Crisis
In Chapter 2, we analyzed the conditions under which the British Factory Acts finally worked. There were three: the formalization of evidence, social learning, and the design of enforcement mechanisms.
Financial regulation followed the same pattern. Brooksley Born's warning (formalization of evidence) came first. LTCM and Enron (attempts at social learning) followed. But before enforcement mechanisms could be designed, the crisis arrived.
Britain endured sixty-four years during which tens of thousands of children suffered. Finance endured a thirty-seven-year gap that ended with a global financial crisis. What is the scale of crisis we must endure in the age of AI?
Algorithms make credit decisions for millions, determine hiring outcomes, intervene in medical diagnoses, and shape democratic elections. When these systems are found to be systematically biased — like Lehman in 2008 — will institutions move then?
By then, decades may already have passed.
In the next two chapters, we examine two experiments unfolding in real time. Chapter 4 takes us to Brussels: the European Union, which produced the world's first comprehensive AI regulation after thirty-five months of negotiation. Chapter 5 takes us to Beijing: China, which issued regulations in eight and a half months. The two experiments expose the tradeoff between speed and legitimacy in its starkest form.
With the lessons of thirty-seven years of financial regulation laid before them, two systems chose different answers. If Greenspan confessed during the financial crisis that "I found a flaw in the model that I perceived is the critical functioning structure that defines how the world works" — a belief he had held for forty years — what flaw will AI-era legislators discover? And will that discovery, once again, come only after crisis? Chapter 4 is the story of people who tried to answer that question first.