Too Big to Believe

The behavioural logic behind Lehman’s collapse

The film Too Big to Fail is useful less as a historical record than as a way of seeing how decisions were made inside a financial system that had already lost the ability to locate its own risk clearly. The aim here is not to reconstruct every event in sequence. It is to examine the logic the crisis exposed: how funding dependence compresses decision time, how institutional belief functions as a live financial resource, and how a system behaves once that belief breaks.

By September 2008, the central question was no longer who had been wrong in the years before. The pressing question was narrower and more dangerous: what would fail next if nothing was done. The mortgage market had already been producing losses for months. Risk had been built, sliced, repackaged and sold through the system for years. It was still present, but it had been distributed through securitised instruments and funding chains so widely that clear ownership had become hard to trace. Gorton and Metrick (2012) later described the crisis as centred on securitised banking and a run on repo. That formulation captures something essential. The assets were still there, but confidence in how they could be financed was weakening.

That background helps explain the behaviour of the major investment banks before the collapse. They kept growing inside a market structure that had already taught a dangerous lesson: size and interconnectedness could create political reluctance around failure. The larger the balance sheet, the more transactions passed through it, and the more complicated disorderly failure became for everyone else. Institutions did not need a formal guarantee to act on that expectation. The system itself had already made the logic legible.

Lehman Brothers relied heavily on that environment. It expanded rapidly, carried substantial exposure, and depended on short-term funding, particularly through repo. The mechanism was technically ordinary and behaviourally fragile. Securities were pledged in exchange for cash and then repurchased shortly afterwards, often the next day. The arrangement functioned while lenders continued to believe two things at once: that the collateral would still be acceptable tomorrow, and that the institution itself would still be standing tomorrow. Copeland, Martin, and Walker (2014) showed how fragile tri-party repo funding could become under pressure, even if the market did not break in the same way for every borrower. In Lehman’s case, confidence deteriorated sharply in the final phase.

This is where the crisis becomes easier to read as a behavioural problem as well as a financial one. A funding structure built on daily renewal is also built on repeated acts of confidence. Those acts are not sentimental. They are market judgements made under uncertainty. When the judgement shifts, the institution’s time horizon collapses. Lehman did not need a long period of deterioration once funding confidence disappeared. It needed only a very short period without renewed lending. For an institution dependent on continuous short-term finance, that was enough.

The endgame exposed this with unusual clarity. Suspicion had surfaced earlier, but the decisive phase began on Friday, 12 September 2008, in New York, behind closed doors. The people in the room gradually understood that this was no longer a question of general market stress. Lehman could become insolvent by Monday morning. Executives, regulators and potential counterparties worked through options under severe time pressure. Barclays showed serious interest, but a rapid acquisition ran into regulatory and timing constraints. Other possibilities circulated and then weakened. The problem was not always that an option made no economic sense. In several cases there was no usable legal route, no speed, or no political willingness to carry it through in time.

By Sunday evening, the outcome was clear. Lehman would not be rescued. That decision had an immediate practical meaning and a wider systemic meaning. For years, markets had operated with a growing expectation that institutions of sufficient size and interconnection would not be allowed to fail disorderly. That expectation had become part of the system’s own risk structure. Letting Lehman go was therefore also a test of that belief. Sorkin’s account of the weekend captures the extent to which the decision-makers were operating under both financial and symbolic pressure: they were dealing with a failing institution and with the precedent that rescuing it would reinforce (Sorkin, 2009).

The market response showed quickly that the decision did not restore calm. It intensified uncertainty. The next question was not whether Lehman had deserved to fail. The next question was who might now be read as vulnerable through the same funding lens. Morgan Stanley and Goldman Sachs depended on similar short-term market confidence. Morgan Stanley in particular came under severe pressure as credit default swap spreads widened and counterparties became more cautious. The firm had not suddenly become a different institution over a weekend. What changed was the market’s reading of its survivability inside the same funding environment.

That is the point at which the rescue logic changed. The Federal Reserve moved more openly into the role of lender where markets would no longer lend. Morgan Stanley and Goldman Sachs were allowed to convert to bank holding companies, giving them access to a more stable funding and liquidity framework. On paper, this looked like a regulatory status change. In practical terms, it was part of the mechanism that kept them alive. The distinction matters because it shows what the crisis management process was actually tracking. The dividing line was not moral worth. It was funding fragility and systemic consequence.

The same logic appears when the story widens beyond the investment banks. AIG required extraordinary intervention because the scale and opacity of its obligations created fears of broader contagion. The Congressional Oversight Panel (2010) documented the extent to which concerns about interconnected loss transmission shaped the rescue decision. General Electric also encountered severe strain through short-term funding markets, despite being known primarily as an industrial company. The crisis did not respect sector labels. It moved along financing structures.

This is one reason the episode remains so important. The fault line did not run neatly between reckless firms and prudent firms, or between finance and the “real economy”. It ran through systems built on rollover funding, collateral confidence and assumptions about what markets and states would tolerate. Once confidence weakened, distinctions that had looked stable became much less protective.

In a narrow technical sense, crisis management worked. Markets stabilised. Funding channels were restored. Institutions that might have failed under continued panic remained standing. But the longer view is harder to settle comfortably. The interventions did not produce a complete redesign of the underlying system. Banks returned to profitability. Losses outside the protected core remained widely distributed across households and workers. The basic question therefore remained open.

How long can a system keep functioning when some of its central truths are too destabilising to state directly?

That question sits underneath Too Big to Fail more than any individual heroic or villainous role. A system had been built in which confidence was treated as permanently renewable, risk was treated as distributable beyond clear ownership, and collapse was treated as less likely for institutions that had become difficult to unwind. Lehman showed what happened when one of those beliefs was tested and not defended.

That is why the ending still matters. The crisis did not simply reveal losses. It revealed the behavioural architecture of a system that needed confidence, obscurity and political hesitation in order to keep functioning at scale. Once one of those supports gave way, decision-makers were no longer choosing between clean options. They were trying to slow a sequence in which belief, funding and survival had become tightly bound to one another.

The story ends historically in 2008. The mechanism did not end there.

References

Congressional Oversight Panel. (2010). The AIG rescue, its impact on markets, and the government’s exit strategy.

Copeland, A., Martin, A., & Walker, M. (2014). Repo runs: Evidence from the tri-party repo market. The Journal of Finance, 69(6), 2343–2380.

Gorton, G., & Metrick, A. (2012). Securitized banking and the run on repo. Journal of Financial Economics, 104(3), 425–451.

Sorkin, A. R. (2009). Too big to fail: The inside story of how Wall Street and Washington fought to save the financial system—and themselves. Viking.