Why did the global financial system nearly collapse in 2008?

The answer in one minute

American house prices rose for years, and the financial system built an enormous industry on the assumption that they would keep rising. Mortgages — including subprime loans to borrowers with weak credit — were bundled into securities, stamped safe by rating agencies, and sold to banks and investors worldwide.

The firms holding these securities were run on extraordinary amounts of borrowed money, much of it borrowed for days at a time. When house prices turned down in 2006–07, nobody could tell which institutions were sitting on fatal losses — so lenders pulled back from everyone. Bear Stearns failed in March 2008; Lehman Brothers followed in September, and credit froze across the globe.

Governments and central banks then intervened on a scale never seen in peacetime — rescues, guarantees, near-zero interest rates and money-printing — which stopped the collapse but not a deep global recession. The world still lives with the politics, the debts and the rules written in 2008’s shadow.

A boom built on houses

Between the late 1990s and 2006, the price of American homes did something it had never done in national records: it rose, steeply, almost everywhere, year after year. The S&P Case-Shiller national index climbed by more than 80 per cent between January 2000 and its peak in July 2006. In the hottest markets — Las Vegas, Miami, Phoenix — prices more than doubled.

A long boom changes what people believe. Households treated homes as investments that could not lose. Lenders treated rising prices as a safety net: if a borrower could not pay, the house could always be sold for more than the loan. And Wall Street treated American mortgages as a vast raw material — a stream of monthly payments that could be processed, packaged and sold to the world.

Cheap credit fed the machine. After the dot-com bust and the September 11 attacks, the Federal Reserve had held interest rates low, and money from fast-growing exporters such as China was pouring into American debt, pushing rates on safe assets lower. Investors everywhere were hunting for something that paid a little more while still looking safe. Housing seemed to be the answer.

Subprime: lending to those who could not easily pay

For most of the twentieth century, a mortgage was a loan a bank made and kept, so the bank cared intensely whether it would be repaid. By the mid-2000s, that link had dissolved. Most mortgages were sold on within weeks of being written, so the fees came from volume, not repayment. Lending standards slid accordingly: loans requiring little or no documented income, “teaser” rates that reset sharply upward after two or three years, mortgages for the full price of the house.

At the peak of the boom, roughly one in five new American mortgages was subprime, and many more sat in a grey zone of near-prime lending. This was not a hidden corner of the market. It was the market’s growth engine — and it depended entirely on house prices continuing to rise, because a struggling borrower’s only exit was to sell or refinance at a higher valuation.

The machine: turning mortgages into “safe” securities

The crucial invention — and the reason an American housing problem became a global banking problem — was securitisation.

Homebuyers
take out mortgages, promising decades of monthly payments
Mortgage lenders
make the loans — then sell most of them on rather than holding them
Investment banks
pool thousands of loans and slice the payment stream into tranches
Rating agencies
grade the slices — most were stamped triple-A, the safest grade
Investors worldwide
pension funds, banks and money funds buy the securities and collect the payments

Money flowed from stage 5 to stage 1: investors’ cash, passed back along the chain, ultimately funded new mortgages.

Risk flowed from stage 1 to stage 5: when homeowners stopped paying, the losses landed on whoever held the securities.

The mortgage-securitisation chain. Each stage passed the loans along for a fee; almost nobody in the middle kept the risk they created.

Investment banks bought mortgages by the tens of thousands and pooled them. The pool’s monthly payments were sliced into layers called tranches. The top tranche was paid first, so it looked extremely safe; the bottom tranches absorbed the first losses in exchange for higher returns. The logic sounded reasonable: American mortgage defaults had historically been rare and local, so a nationwide pool seemed diversified. Thousands of shaky loans, the theory went, could be refined into bonds as safe as government debt.

The theory had a flaw that would prove nearly fatal: it assumed house prices would never fall across the whole country at once. If they did, the “diversified” loans would all sour together — and the safety of every tranche, everywhere, would be in doubt simultaneously.

Engineering piled on engineering. Tranches that proved hard to sell were re-bundled into collateralised debt obligations (CDOs), and sliced again. Insurance-like contracts called credit default swaps let firms bet on, or insure against, these securities’ failure — most famously at the insurer AIG, which guaranteed vast quantities of mortgage risk while setting aside almost nothing in case the guarantees were called.

Triple-A: the seal that made it possible

None of this could have scaled without the credit rating agencies. Pension funds, insurers, banks and money market funds are often required — by regulation or their own rules — to hold highly rated assets. A triple-A stamp was the passport that let mortgage risk into the world’s most conservative portfolios.

The agencies supplied those stamps industrially. The Financial Crisis Inquiry Commission found that Moody’s alone rated nearly 45,000 mortgage-related securities triple-A between 2000 and 2007(FCIC Report, 2011). The agencies were paid by the very banks whose products they graded, and their models leaned on years of data from a market that had known only rising prices.

When the downgrades came, they did not merely reprice some bonds. They broke the filing system the entire financial world used to sort safe from unsafe.

Leverage: the amplifier

Modest leverage — borrow $2 for every $1 of your own

$100 of assets = $33 own money + $67 borrowed. If assets fall 4%, the firm loses $4 — painful, but it survives with $29 of its own money left.

Extreme leverage — borrow $32 for every $1 of your own

$100 of assets = $3 own money + $97 borrowed. The same 4% fall loses $4 — more than all the firm’s own money. It is insolvent.

Now compare: the same small loss, drawn to scale

This is a 4% fall in asset values — wider than the second firm’s entire slice of own money. That is why, in 2008, small declines in housing assets could erase firms that had stood for a century.

the firm’s own money (equity) borrowed money (debt) a 4% loss

Two balance sheets holding the same $100 of assets. The same small loss is survivable for one and fatal for the other.

By 2007, the five big American investment banks were operating, by one measure in the official inquiry, with leverage as high as 40 to 1(FCIC Report, 2011) — a few dollars of their own capital behind every hundred dollars of assets. European banks were often no more conservative. At those ratios, a fall of a few per cent in asset values does not dent a firm; it erases it.

Borrowed overnight, owed for decades

Leverage decided whether firms could survive losses. A second fragility decided how fast they would die: the money was borrowed short.

Investment banks and many others funded portfolios of thirty-year mortgages with borrowing that matured in days — commercial paper and “repo” loans that had to be renewed constantly, secured against the very mortgage securities whose value was now in question. This was a banking system in economic substance but not in legal form: a shadow banking system, with no deposit insurance and no automatic access to central-bank lending. It was vulnerable to exactly what destroyed banks in the 1930s — a run. Not queues of depositors this time, but institutions declining, with a keystroke, to renew yesterday’s loan.

The turn

National house prices peaked in mid-2006 and began to slide. Teaser rates reset; borrowers who had planned to refinance found they no longer could; defaults climbed. By mid-2007 the supposedly safe securities were falling in value, and the rating agencies began downgrading them by the hundreds.

The deeper poison was uncertainty. Mortgage risk had been sliced, re-bundled and scattered so thoroughly that no one — not regulators, not the banks themselves — could say where the losses would land. Every counterparty became a suspect. In August 2007, the French bank BNP Paribas froze three funds, saying parts of the market had become impossible to value, and the interbank lending market seized for the first time. In Britain, Northern Rock — a mortgage lender wholly dependent on market funding — suffered the country’s first run on a major bank since the Victorian era and was nationalised.

2008: the year of falling giants

How the crisis unfolded

2007
9 August 2007
The freeze begins
BNP Paribas halts withdrawals from three funds holding U.S. mortgage securities; overnight lending between banks abruptly tightens and central banks inject emergency funds.
September 2007
Run on Northern Rock
Depositors queue outside the British mortgage lender — the first run on a major U.K. bank in over a century. It is nationalised in February 2008.
2008
16 March 2008
Bear Stearns falls
The fifth-largest U.S. investment bank, unable to renew its short-term funding, is sold to JPMorgan Chase in a weekend deal backed by the Federal Reserve.
7 September 2008
Fannie Mae and Freddie Mac seized
The two government-sponsored giants standing behind roughly half of U.S. mortgages are placed into federal conservatorship.
15 September 2008
Lehman Brothers files for bankruptcy
With over $600 billion in assets, it is the largest bankruptcy in American history. No buyer and no bailout is found.
16 September 2008
AIG rescued; a money fund "breaks the buck"
The Federal Reserve extends an $85 billion lifeline to the insurer AIG. The Reserve Primary money market fund posts losses on Lehman debt, triggering a run on funds treated by savers as cash.
3 October 2008
The $700 billion rescue law
Congress passes the Emergency Economic Stabilization Act, creating the Troubled Assets Relief Program (TARP) — days after first voting it down.
8 October 2008
Central banks act together
The Federal Reserve, European Central Bank, Bank of England and others cut interest rates in a coordinated move; governments across Europe begin guaranteeing and recapitalising their banks.
25 November 2008
Quantitative easing begins
The Federal Reserve announces large-scale purchases of mortgage-related securities — the start of a new era of central banking.
16 December 2008
Interest rates hit zero
The Fed cuts its main rate to a range of 0–0.25%, the lowest in its history, and signals it will stay there for some time.
2009
June 2009
The recession ends — on paper
The U.S. economy stops shrinking after eighteen months, the longest downturn since the 1930s. Unemployment keeps rising for months afterwards.
2010
21 July 2010
The rules are rewritten
The Dodd-Frank Act becomes U.S. law; internationally, the Basel III accords force banks to fund themselves with more capital and less short-term debt.

Bear Stearns went first, in March — saved from bankruptcy only by a Fed-assisted fire sale. The lesson markets drew was double-edged: these firms were fragile, but the government would apparently catch them.

Then, on the weekend of 13–14 September, that assumption met Lehman Brothers. Officials could find no buyer; they insisted they lacked legal authority to lend into a hole of unknown depth. On Monday 15 September 2008, Lehman filed for bankruptcy.

What followed in the next seventy-two hours is why “2008” means what it means. The insurer AIG, facing payment on its vast mortgage guarantees, was rescued with an $85 billion public lifeline one day after Lehman was not. A major money market fund — a vehicle millions of savers treated as indistinguishable from cash — announced its dollars were now worth 97 cents, and institutional money stampeded out of the entire sector. The commercial paper market, which funds the everyday operations of even industrial companies, began closing. Banks would barely lend to each other overnight, at any price.

This is what “the financial system nearly collapsed” concretely means: the plumbing that carries payrolls, trade credit and corporate cash came measurably close to stopping — not for lack of money in the world, but because trust, the thing the system actually runs on, was gone.

The rescue

What stopped the spiral was state power, applied in overlapping layers over about six months. Treasuries put public capital directly into banks — the U.S. through TARP, Britain through the part-nationalisation of RBS and Lloyds-HBOS, with rescues across Ireland, Iceland, the Benelux and beyond. Governments guaranteed bank debts and money funds to halt the runs. Central banks lent freely against assets no one else would touch, cut rates towards zero, opened dollar pipelines to each other so foreign banks could pay dollar debts, and then began buying securities outright — quantitative easing. In spring 2009, American regulators published stress tests showing which big banks could survive a worse recession and forced them to raise capital, which finally began restoring the missing ingredient: verifiable trust.

It worked, in the narrow sense. No major economy’s payment system failed, and by mid-2009 the acute panic was over.

The bill

Rescuing the system did not rescue the economy. The United States lost output on a scale unseen since the 1930s — GDP fell over four per cent from peak to trough, and unemployment doubled from five to ten per cent, with nearly nine million jobs lost. Millions of families lost homes to foreclosure. And the damage was global: in late 2008, world trade and industrial production fell more steeply than at any time since the Second World War, dragging down economies that had never touched a subprime mortgage. In 2009, the world economy contracted — the first global fall in output of the post-war era.

The recession officially lasted from December 2007 to June 2009, but that dating flatters the recovery, which was slow and unequal almost everywhere.

In Europe, the crisis mutated: bank rescues and recession blew holes in public finances, helping trigger the sovereign debt crises that nearly broke the euro after 2010.

The crash and the crisis are not the same thing

It is tempting to picture 2008 as a stock-market event — traders with heads in hands as screens turn red. The falling stock market was real (American shares lost roughly half their value from their 2007 peak to early 2009), but it was a symptom, and not the dangerous part.

Stock prices falling means owners of shares are poorer. The economy can absorb that; it did after the dot-com crash of 2000–02, with only a mild recession. What it cannot absorb is the failure of the credit system — the machinery of loans, deposits and payments that every business, solvent or not, uses to operate. In 2008, that machinery itself was failing: companies with full order books could not roll over routine borrowing; banks could not fund themselves overnight. That is why 2008 was catastrophic when 2000 was merely painful, and why the policy response aimed first at banks rather than shareholders: the point was not to protect investors’ wealth but to keep the economy’s circulatory system running.

What changed afterward

The regulatory answer came in two main waves. In America, the Dodd-Frank Act of 2010 — the largest rewrite of financial rules since the 1930s — created a consumer financial protection agency, a council to watch for system-wide risks, resolution powers to wind down failing giants without either bankruptcy-by-surprise or open-ended bailout, and pushed derivatives towards transparent clearing. Internationally, the Basel III accords forced banks to hold several times more genuine loss-absorbing capital, capped their leverage, and — for the first time — regulated liquidity, limiting dependence on the overnight funding that had killed Bear and Lehman. Central banks kept the tools invented in the emergency; quantitative easing and near-zero rates shaped the entire following decade.

What did not change is equally telling. No senior executive of a major American financial firm went to prison for crisis-era conduct. The biggest banks emerged bigger. And credit creation migrated partly beyond the regulated banks — into funds and private credit vehicles — where some of the old questions about leverage and runnable funding are now asked again.

What remains disputed

We conclude this financial crisis was avoidable.

— Financial Crisis Inquiry Commission, final report, 2011

That was the verdict of the official inquiry’s majority. Yet serious people still disagree about 2008, and honest explanation means saying where.

Who deserves the blame? The official inquiry itself split three ways. The majority indicted reckless private risk-taking and failed regulation; dissenters emphasised, variously, a broader global credit bubble, or U.S. housing policy and the government-sponsored mortgage giants. Where the balance lies between Wall Street greed, Washington neglect and global capital flows remains genuinely contested.

Did Lehman have to fall? Officials have argued they lacked legal authority to save it; critics reply that the authority found for Bear Stearns and AIG could have been found for Lehman, and that letting it fail was a choice — the costliest miscalculation of the crisis.

Were the rescues just? Saving the system meant saving many of the people who broke it. Whether the bailouts were a regrettable necessity or a moral catastrophe that entrenched “too big to fail” is a live argument — one with obvious force the next time a rescue is needed.

Is the system safe now? Banks are unambiguously better capitalised. But 2008-style risk is adaptive: it moves to wherever the rules are thinnest. Whether reform fixed the system or merely moved the fault line is a question only the next crisis can settle.

Why 2008 still matters

The crisis is not really over, because its consequences became the world we live in. The long era of near-zero interest rates and central-bank asset buying reshaped everything from house prices to pension funds — and when inflation returned in the 2020s, the unwinding of that era caused tremors (including the 2023 American bank failures) that were 2008’s direct descendants. The public debts swollen by rescue and recession still constrain governments. And politically, the sight of banks rescued while households were foreclosed upon corroded trust in institutions across the West, feeding the populist revolts of the following decade.

2008 also left a lesson worth keeping, because it will be tested again: financial systems do not fail because assets lose value. They fail because trust — quiet, invisible, assumed — turns out to be the only thing holding them up.

Correction, 4 August 2026: an earlier version said national house prices “roughly doubled” between 2000 and 2006; the S&P Case-Shiller national index rose by more than 80 per cent. The comparison of the 2008–09 trade collapse to the early Great Depression has also been restated in more careful terms.