2008 from below: what apparatus did the crisis demand?

The story did not start at BNP Paribas. It started two months earlier, inside Bear Stearns. On June 22, 2007, Bear committed a \$1.6 billion credit line to its High-Grade Structured Credit Strategies Fund as investors rushed to pull their money out. On July 17 the bank told clients the fund and its more leveraged sibling had lost essentially all of their value. On July 31 both funds — the two run by Ralph Cioffi and Matthew Tannin — filed for bankruptcy. They had been earning their returns by holding bonds built out of subprime mortgages and borrowing short-term to hold more of them. When subprime borrowers began missing payments in early 2007, the market for those bonds thinned out, and "thinned out" became "no buyer at any price" faster than the funds' own models said it could. The Bear hedge funds were not the crisis. They were the first event in it.

Then came August. The BNP Paribas statement on the 9th was the first time a large, ordinary bank said in public that it simply could not price its own holdings. Within hours the European Central Bank injected €95 billion into the banking system to keep overnight lending from seizing; the Federal Reserve followed on August 17 with a discount-rate cut and longer-term loans to banks. The market for asset-backed commercial paper — the short-term IOUs that funded vast pools of mortgage securities — shrank from roughly \$1.18 trillion to about \$0.95 trillion over the following weeks as lenders refused to roll the paper over. The TED spread, the gap between what banks charge each other and what the U.S. government pays, blew out from about 40 basis points in early July to 240 by late August. A reader following the wires through that month watched losses in one specific corner of the mortgage market spread into one specific funding market, watched that funding market stop working, and watched central banks step in to supply the cash the market would normally have supplied itself. The BNP Paribas press release was the moment liquidity stopped being an abstraction.

For the deeper chronology — the build-up of subprime origination, the securitization chain, and the summer-2007 funding stress as a connected sequence — the canonical account is Economic History Ch.19 §19.1 (The Crisis: September 2008 and the Global Financial Collapse), with the asset-backed-commercial-paper funding-market detail in the same section’s institutional-run passage.

The spring ran on rescues. On March 16, 2008, the Fed engineered the sale of Bear Stearns to JPMorgan — at \$2 a share initially, later raised to \$10 — and backstopped \$29 billion of Bear's worst assets through a special vehicle named Maiden Lane. On July 11, regulators seized IndyMac, a \$32 billion thrift, in the largest such failure to that point. On September 7, the Treasury took Fannie Mae and Freddie Mac into conservatorship, eventually drawing \$187.5 billion on the lines of credit it extended them. Running through every one of these decisions was a phrase the policymakers used out loud in testimony and press conferences: too big to fail. They did not mean it as a model. They meant that some institutions were so wired into the rest of the system that letting them collapse would take the system down with them — and that this concern, not the firm's own deserts, justified the public money.

Then came the weekend of September 12–15. At the New York Fed, a Barclays bid for Lehman foundered on a Fed guarantee Paulson and Bernanke would not provide; a Bank of America bid pivoted to Merrill Lynch instead; Lehman filed at 1:45 AM Monday. The next morning the Reserve Primary Money Fund disclosed \$785 million of now-worthless Lehman commercial paper, and its share price fell to \$0.97 — "breaking the buck," the first time in decades that a money-market fund, the vehicle ordinary Americans treated as indistinguishable from cash, had lost principal. The run that had started in one funding market in 2007 had now reached the safest non-Treasury short-term investment in the country. The policymakers reached for two more phrases. Counterparty exposure: Paulson's later account in On the Brink turns the Lehman decision on whether the contagion would run worse with a Fed-funded rescue, which would set a moral-hazard precedent, or without it, which was what actually happened. And the run on the shadow-banking system: Timothy Geithner's framing for the emergency money-market guarantee announced September 19, in which the Treasury stood its \$50 billion Exchange Stabilization Fund behind every dollar of every money-market fund to stop the run that had started that Tuesday.

This walkthrough keeps the Lehman weekend at chronology-paragraph depth; for the hour-by-hour account of the negotiations at the New York Fed and the systemic-risk apparatus they demanded, the deeper treatment is the sibling walkthrough The Lehman weekend.

The response stage came fast. On September 16 the Fed extended AIG an \$85 billion credit line, eventually \$182 billion — not because AIG's own books were the problem but because AIG had written credit-default-swap protection on hundreds of billions of dollars of securities held by major banks across the world, so letting AIG fail would have pushed losses straight into that web of counterparties. On October 3 the Troubled Asset Relief Program was signed: \$700 billion in authority, of which \$245 billion went to banks through the Capital Purchase Program by year-end and \$79.7 billion went to the automakers. On December 19 the Bush administration extended \$13.4 billion to General Motors and Chrysler to keep them out of immediate bankruptcy. The phrase behind the AIG rescue was systemic interconnection — Bernanke's account in The Courage to Act turns the decision on AIG's exposure to the whole investment-banking complex rather than on AIG's own solvency. The information-economics literature these concepts lean on — adverse selection when buyers cannot tell good assets from bad, and the fire-sale externality where one firm's forced selling lowers the price every other holder marks against — runs through Stiglitz-Weiss and Shleifer-Vishny; the textbook home is Economics Ch.4 §4.6 (Information Asymmetry). But in September 2008 these were concepts the policymakers invoked, not models they could run. The formal apparatus arrives in the next stage.

The full event chronology — Bear and IndyMac and the GSEs, the Lehman–AIG–Reserve Primary week, and the TARP-and-automakers response — is the spine of Economic History Ch.19 §19.1 (The Crisis: September 2008 and the Global Financial Collapse), whose coordinated-response passage walks the emergency facilities and rescues in turn.

The pre-2008 workhorse had a financial sector that did nothing: saving flowed to investment through a frictionless conduit, and balance sheets had no effects of their own. The apparatus the events demanded had to put four specific dynamics — each one visible in Stages 1 and 2 — inside the model. Four research families did exactly that, and each maps onto a cascade-dynamic the crisis exhibited.

1. The financial accelerator (Bernanke-Gertler-Gilchrist 1999, extended by Christiano-Motto-Rostagno 2014). A contracting friction between borrowers and lenders generates an external-finance premium that rises as net worth falls, so an adverse shock weakens balance sheets, which raises the cost of borrowing, which weakens balance sheets further. This formalizes the Bear → Lehman → AIG counterparty cascade of Stage 2 — the loop in which losses propagate through interconnected balance sheets rather than dissipating.
2. Banking with leverage constraints (Gertler-Karadi 2011, Journal of Monetary Economics). Financial intermediaries face a balance-sheet constraint that binds when their net worth deteriorates, and central-bank asset purchases substitute for the lending impaired intermediaries can no longer do. This formalizes the September 2008 emergency-facility arc of Stage 2 — the alphabet soup of Fed lending programs and the AIG credit line, all of them stepping in where impaired intermediaries had stopped supplying credit and conventional rate cuts could not reach.
3. Liquidity-funding spirals (Brunnermeier-Pedersen 2009, Review of Financial Studies). Market liquidity (the cost of trading an asset) and funding liquidity (the cost of borrowing to hold it) feed back through margins, producing reinforcing loss spirals and margin spirals that can collapse a market with no change in fundamentals. This formalizes the funding-market freezes of both earlier stages — the August 2007 commercial-paper freeze, the March 2008 run on Bear's repo funding, and the September 2008 money-market-fund run — as the same mechanism running at three different scales.
4. Intermediary asset pricing (He-Krishnamurthy 2013, American Economic Review). When the intermediary sector's capital is impaired, risk premiums on the assets intermediaries hold jump, because the sector's capacity to bear risk — not the representative consumer's appetite for it — is what is constrained. This formalizes the late-2008 spread blowouts — credit, mortgage, and swap spreads all widening together — that the standard consumption-based asset-pricing model could not explain at all.

At the core of the whole program sits one idea that can be written in a line. The cost of borrowing is not a fixed market rate; it carries a premium that depends on the borrower's balance sheet.

The unmodified base these extensions retrofit onto is the canonical New Keynesian core; the toolkit for reaching the point where conventional policy runs out and the unconventional tools begin lives alongside it. Want the model walk rather than the identification? The three textbook homes are below.

Siehe Kap. 15 §15.6: Das 3-Gleichungen-NK-Modell Siehe Kap. 15 §15.8: NK vs. RBC Impulsantworten Siehe Kap. 4 §4.6: Informationsasymmetrie

For how this financial-friction program sits inside the New Keynesian tradition it patched — what the canonical synthesis model missed, and how the discipline bolted the missing structure on rather than abandoning the framework — the intellectual-history account is History of Economic Thought Ch.12 §12.7 (The reckoning: what 2008 exposed and what survived).

Start with what the apparatus now explains. The shadow-banking-system fragility of all three earlier funding-market freezes — the August 2007 commercial-paper freeze, the March 2008 run on Bear's repo funding, the September 2008 money-market-fund run — fits the liquidity-funding-spiral framework as a single mechanism at three scales. The counterparty-cascade transmission of Bear → Lehman → AIG fits the balance-sheet propagation of the accelerator and banking models. And the policy architecture that followed reads as the apparatus operationalized: macroprudential regulation as margin-and-leverage-constraint policy; lender-of-last-resort capacity as intermediary-balance-sheet support; the Fed's emergency-liquidity facilities — the TAF, the PDCF, eventually a standing repo facility — treated not as one-off improvisations but as a permanent part of the playbook. Dodd-Frank (July 2010) and Basel III (December 2010) built the regulatory frame the apparatus informed; the 2009-onward Fed stress-test program, institutionalized as the Comprehensive Capital Analysis and Review, made the lender-of-last-resort posture routine. That policy-response arc is the regulatory-aftermath stretch of Economic History Ch.19 §19.1 (the Dodd-Frank and stress-testing aftermath).

Now the gaps, which are real and load-bearing. The macro-finance integration is uneven: the financial-friction blocks are bolted onto standard New Keynesian skeletons rather than co-derived with the rest of the model, so the integration is technically genuine but conceptually layered. The endogenous-instability accounts in the Minsky tradition are not yet workhorse: the apparatus formalizes the back end of a financial cycle — the fire sales, the accelerator, the spread amplification once trouble starts — but not the front end, the way a long calm itself manufactures the leverage that later destabilizes the system. The interaction between household heterogeneity and financial frictions, where heterogeneous-agent New Keynesian models meet intermediary balance sheets, is a live research frontier rather than a settled policy-model standard. And the predictive-adequacy question is genuinely open: whether this apparatus would have flagged 2008 had it existed before the crisis is a counterfactual the literature cannot settle, and whether it can flag the next crisis is the live policy question.