The 2008 Financial Crisis, in Plain English

Most readers have heard the fragments. Subprime. Lehman. CDOs. Too big to fail. They've heard the names without ever being shown the chain of events that produced them. Either explanations stop at “it was a housing bubble” — true but useless — or they jump immediately into terminology so dense the reader gives up.

This article is the middle path. It walks through, in order: how a mortgage on a house in Phoenix became part of a bond on a European pension fund's books, why that arrangement fell apart all at once instead of gradually, what the U.S. government's response permanently changed about how money works, and why a nine-page PDF posted to a small mailing list two weeks after Lehman Brothers collapsed turned out to matter.

Why mortgage paper became irresistible

In the years after the dot-com bust, interest rates were low and big institutional investors — pension funds, insurance companies, sovereign wealth funds — were starving for returns. These investors have liabilities they have to meet (pensions to pay, insurance claims to settle), and they need bonds with reliable returns to fund those liabilities. Government bonds were paying very little. Anything paying more was attractive.

U.S. residential mortgages looked perfect for this. They had a long track record: defaults had been historically low, the bank could foreclose and recover most of the value if a borrower stopped paying, and U.S. house prices had not declined nationally since the Great Depression. Mortgage debt was as close to a sure thing as bonds offered.

There was also an unusual amount of mortgage paper to package. After the dot-com bust in 2000–2001, the Federal Reserve had cut interest rates from 6.5% to 1% and held them low through 2003. Cheap mortgages fueled a housing boom. As prices climbed, lenders extended credit to a wider pool of borrowers, including the subprime market — borrowers with weak credit and limited ability to make full monthly payments. By 2005, roughly one in five new mortgages was subprime.

This wasn't just a consumer belief that “house prices always go up.” It was embedded in the risk models the credit-rating agencies used. The models assumed that even if a few mortgages in a pool defaulted, defaults wouldn't all happen at the same time across different cities, different income brackets, and different loan types. That assumption — that defaults across U.S. mortgages were largely independent of each other — is the one that broke in 2007.

How a pile of mortgages becomes a AAA bond

Here is the mechanism. Strip out the jargon and it's surprisingly simple.

A bank, or a non-bank lender, writes mortgages. It can hold those mortgages on its own books, but it would much rather get its cash back so it can write new ones. So it bundles, say, two thousand mortgages together into a pool. Every month, those two thousand homeowners send their mortgage payments into the pool. That stream of payments becomes the basis for new bonds. The cash flow is sliced into layers — the industry calls them tranches, which is just French for “slices” — and each layer is sold to investors as a separate bond.

The top tranche gets paid first, every month, in full. The middle tranches get paid next. The bottom tranche gets whatever's left. If a homeowner defaults, that loss falls first on the bottom tranche. The bottom tranche has to be wiped out completely before the middle tranche takes any losses, and the middle tranches have to be wiped out before the top tranche feels anything.

In exchange for absorbing the first losses, the bottom tranche pays a much higher return — that's the trade-off that makes someone willing to buy it. Hedge funds and other risk-tolerant investors typically bought the bottom tranches; pension funds and insurance companies bought the top.

That math is the whole trick. Because the top tranche only takes losses if a huge number of homeowners default at once, the credit-rating agencies (Moody's, S&P, Fitch) rated the top tranche AAA — the same rating they gave U.S. Treasuries. A pool of mortgages, any one of which could default, had been turned into something pension funds could legally buy. The structure has a name: mortgage-backed security, or MBS.

The next move was even more aggressive. Banks took the BBB-rated middle tranches from many different mortgage-backed securities and pooled them together. They sliced the new pool the same way as before, and the top tranche of this pool was rated AAA again. A bond called a CDO (Collateralized Debt Obligation) had been engineered out of leftover BBB inputs.

Variants of this trick kept getting built on top of variants. By late 2006, the financial system had built hundreds of billions of dollars of leveraged exposure to a single assumption: that U.S. mortgage defaults would stay independent of each other.

The crack: when the music stopped

Starting in mid-2004, as inflation rose, the Federal Reserve began raising rates again — from 1% to 5.25% by mid-2006. That was a problem for the subprime market specifically. Most subprime mortgages had been written as adjustable-rate loans with low “teaser” rates that reset to much higher rates after two or three years. As Fed rates climbed, those resets came in higher. Monthly payments on subprime mortgages jumped — sometimes by 30 to 50 percent. Borrowers who could only afford the teaser rate started missing payments when the resets hit.

With marginal buyers dropping out, U.S. house prices peaked in mid-2006 and started falling. At first, slowly. By 2007, subprime defaults were rising fast.

And because pools across the system held the same kinds of mortgages, defaults rose everywhere at once. The bottom tranches got wiped out together. Losses started climbing into the middle tranches and the supposedly AAA tranches above them. The math trick had turned a pile of mortgages into a “safe” bond. It depended on defaults staying independent — on Phoenix and Miami and Las Vegas not all failing in the same year. They were now all failing in the same year.

In April 2007, New Century Financial, one of the biggest subprime mortgage originators, filed for bankruptcy. In June 2007, Bear Stearns had to bail out two of its own hedge funds, which had heavy CDO exposure. The funds were eventually liquidated. By late 2007, the market had stopped believing in the AAA ratings.

In September 2007, the British bank Northern Rock experienced a depositor run; footage of queues outside branches went global. It was the first run on a British bank since Overend, Gurney & Co. in 1866.

By the end of 2007, the crisis was already systemic. Most people just hadn't named it yet.

The cascade: when the dominoes fell

The chronology of the months that followed reads like a sequence of dominoes too big to ignore.

March 16, 2008 — Bear Stearns. The investment bank's overnight funding evaporated over a single weekend. Treasury and Federal Reserve officials engineered an emergency sale to JPMorgan Chase at $2 per share (later renegotiated to $10), backstopped by $29 billion in Federal Reserve financing. It was the first time since the Great Depression that the Federal Reserve had used its emergency lending powers to support a non-bank firm.

September 7, 2008 — Fannie Mae and Freddie Mac. The two government-sponsored mortgage giants bought mortgages from banks, repackaged them into bonds, and between them backed roughly half of the U.S. residential mortgage market. The U.S. government placed both into conservatorship — direct government control.

September 15, 2008 — Lehman Brothers. Lehman filed for Chapter 11 bankruptcy with $691 billion in assets. Over the preceding weekend, Treasury Secretary Henry Paulson, Fed Chair Ben Bernanke, and NY Fed President Timothy Geithner had assembled the heads of every major Wall Street firm to find a buyer. Bank of America, initially considering Lehman, pivoted to acquiring Merrill Lynch instead. Barclays couldn't get UK regulator approval in time. Treasury and the Fed declined to provide the financing that had backstopped the Bear Stearns deal. By Sunday night there was no buyer. By Monday morning Lehman was bankrupt.

September 16, 2008 — AIG. Twenty-four hours after letting Lehman fail, the Federal Reserve provided an $85 billion emergency loan to insurance giant AIG to prevent its collapse. AIG had sold a kind of insurance — called a credit default swap — that paid out if those AAA tranches lost value. Lehman's failure forced banks to mark down those tranches, which meant AIG suddenly owed enormous payouts. If AIG had failed, every bank holding those tranches would have had to write them down to zero overnight. Total support for AIG eventually reached $182 billion.

The two-day shock. The day AIG was rescued, a money-market fund called the Reserve Primary Fund “broke the buck.” Money-market funds are designed to be as safe as cash — a dollar in, a dollar out, always. When Reserve Primary's share price fell below $1.00, that promise broke for the first time in 14 years. The fund had held $785 million in Lehman commercial paper that became worthless overnight. Within two weeks, money-market funds across the industry saw roughly $400 billion in net outflows. The commercial paper market — which ordinary American companies use to fund payroll and operating expenses — seized up. The system that everyone treated as “essentially cash” was suddenly not.

October 3, 2008 — TARP signed. Congress passed the Troubled Asset Relief Program, authorizing the Treasury to spend up to $700 billion to stabilize the financial system. The bill had failed in the House on September 29, sending the Dow Jones down 778 points in a single day. It passed in revised form four days later.

November 25, 2008 — QE1. The Federal Reserve announced its first asset-purchase program: the Fed would buy mortgage-backed securities and Treasury bonds directly, creating new bank reserves to pay for them. The first round of quantitative easing had begun.

The response: what actually changed

Two things happened in parallel during the rescue, and they had very different long-term consequences.

The first was TARP itself. Of the $700 billion authorized, about $443 billion was actually disbursed: ownership stakes in major banks (the government got shares in exchange for the cash), support for AIG and the auto industry, and a smaller mortgage-assistance program. Eventually, the Treasury got all of that money back — and a small profit on top. This is the source of the often-repeated claim that “TARP made money.”

The claim is technically correct. It is also misleading.

The second thing that happened, in parallel with TARP, was the Federal Reserve's response: emergency lending facilities for every funding market that had seized (eight or so programs, collectively dubbed the “alphabet soup”), and the start of quantitative easing. All those bond purchases added up. The Fed's balance sheet — the total assets it now held — went from approximately $900 billion before the crisis to approximately $4.5 trillion by 2015. That expansion did not reverse.

TARP was the visible, vote-able portion of the rescue. The Fed-side response was many times larger and didn't require congressional authorization. TARP made money. The larger system the rescue established, didn't.

The pattern that didn't end

The 2008 response wasn't a one-off. It was a template.

QE1 (2008–2010) was followed by QE2 (2010–2011), QE3 (2012–2014), and the COVID-era expansion (2020–2022). The triggers varied — financial crisis, slow recovery, pandemic — but the response was always the same: the Federal Reserve expanded its balance sheet by buying assets, and each expansion stuck. Each round was framed as exceptional. Each became part of the baseline.

This is the mechanism behind the price drift the clock is measuring. As the previous article on where new money comes from walked through, banks create deposits when they lend; the Federal Reserve creates reserves when it buys assets. Both expand the money supply. The 2008 response normalized doing the second of those at scale, without expecting to reverse it.

The backdrop for Bitcoin

Step back to October 31, 2008. Lehman had failed six weeks earlier. TARP had been signed three weeks earlier. QE1 was four weeks away. On that day, a person or group writing under the name Satoshi Nakamoto posted a nine-page PDF to a cryptography mailing list. The PDF described a way for two people to send money to each other without going through a bank.

Three months later, on January 3, 2009, the first Bitcoin block was mined. Embedded in that block was a piece of text: “The Times 03/Jan/2009 Chancellor on brink of second bailout for banks.” It referenced the front page of that day's London Times, which reported on the British government's consideration of a second round of bank rescues.

The reference wasn't decorative. The system Bitcoin's design was negating — fixed supply against an unbounded one, no central authority against the Federal Reserve, no bailouts against TARP — was the system the post-Lehman response had just established as standard.

What this means for the clock

Every basket item on Satoshi's Clock is priced against a dollar whose backing has changed: the Federal Reserve's balance sheet has expanded roughly five times since 2008, and the broader money supply (M2) has more than doubled. The purchasing-power erosion the clock measures is downstream of that expansion.

A month of rent costs much more in 2026 than it did in 2008. A dozen eggs costs much more. A gallon of gasoline costs much more. The dollar amounts on the left side of every clock card grew because the supply of dollars grew. The supply of dollars grew because the response to 2008 — and to every crisis since — included expansion of the money supply as a routine policy tool.

The 2008 crisis itself was bad. The template the response established is the part that has shaped every monetary decision in the years since, and that template is what the clock counts against.