When a financial system breaks

The last chapter of the subject. Crises differ in their details and repeat in their structure, and the structure is built from everything in the preceding eleven chapters.

The anatomy

1. Credit expands. Something makes lending look safe — a new segment, rising collateral values, a period without defaults. Chapter 2's pro-cyclicality: good conditions make borrowers look creditworthy, which expands credit, which improves conditions.

2. Standards slip, invisibly. Competition for growth means lending to weaker borrowers at thinner spreads. Nothing shows in the numbers, because chapter 4's loans go bad slowly. Reported quality is at its best when the underlying decisions are at their worst.

3. Leverage rises. Borrowers and lenders both. Chapter 11 of the Corporate finance subject: the good years are a multiple.

4. A shock arrives. The particular shock is the part nobody predicts and the part least worth predicting, because the fragility was built beforehand.

5. Collateral values fall. Loans that were safe at the old price are not at the new one, and selling collateral pushes the price down further for everyone holding it.

6. Funding stops. Not for the weakest institution — for the category, because lenders cannot tell quickly which is which. Chapter 11's contagion.

7. Liquidity and solvency blur. Below.

8. Credit stops reaching the real economy, and a financial problem becomes an economic one.

Why liquidity and solvency cannot be told apart in the moment

In theory the distinction is clean. Illiquid means good assets, no cash. Insolvent means the assets are not worth the liabilities.

In a crisis the two collapse into each other, for a specific reason: an illiquid institution must sell assets, and forced selling into a falling market makes it insolvent. The act of addressing illiquidity creates insolvency.

Worse, asset values during a panic are not observable. The market price of something nobody will buy is not a valuation; it is the absence of one.

So the authorities must decide whether to rescue an institution without being able to know which problem it has — and they must decide in days. That, rather than any failure of analysis, is why crisis decisions look poor afterwards.

Working the problem

The case for rescuing: if it is illiquid but solvent, letting it fail destroys value for no reason and triggers the contagion of chapter 11 — other institutions lose funding, credit stops, and the damage spreads to firms and households who did nothing. The cost of an unnecessary failure is borne by people far outside it.

The case for letting it fail: rescuing an insolvent institution transfers losses from the people who took the risk and earned the return to the public. It also teaches every other lender that reckless growth is underwritten, which makes the next crisis larger. Moral hazard is not a theoretical objection; it is the mechanism by which rescues cause the thing they are rescuing from.

The information that would most change the answer: whether the institution is solvent on a realistic valuation of its assets — not a panic price, not its own mark. In practice that means supervisory knowledge built up before the crisis, which is why the quality of ongoing supervision determines the quality of crisis decisions. A supervisor who does not already know the book cannot learn it in a weekend.

The classical answer — lend freely, against good collateral, at a penalty rate — is an attempt to split the difference: rescue illiquidity, not insolvency, and make it expensive enough that nobody relies on it. It works to the extent that "good collateral" can be assessed, which is precisely what a crisis makes hard.

Why India's structure matters here

Three features, from earlier chapters:

One institution holds several mandates. The RBI regulates banks, sets monetary policy, manages government debt and acts as lender of last resort (chapter 5). That concentrates information usefully and concentrates conflicts too.

Deposit insurance covers small depositors up to a limit, which removes most of the retail run risk — the stabiliser chapter 1 described.

NBFCs sit outside that protection (chapter 11), which is where the fragility has concentrated.

What the subject was arguing

Twelve chapters, one thread: a financial system is a confidence mechanism with a solvency requirement attached, and almost every rule in it exists because someone's incentive pointed the wrong way.

Capital is regulated because banks do not bear the full cost of their own failure. The ninety-day rule exists because a bank left to judge when a loan went bad would always say "not yet". Lending rates were benchmarked externally because a rate a bank computes is a rate a bank manages. Each rule is a response to a specific, documented failure of self-interest.

Which is the useful way to read any financial regulation: ask what incentive it was written to counteract. The answer is usually in the rule's own history, and it explains the rule better than the rule's text does.

The point

Crises follow a structure: credit expands against rising collateral, standards slip while reported numbers improve, leverage builds, a shock arrives, funding stops for a whole category, and liquidity becomes indistinguishable from insolvency because forced selling turns one into the other. Rescuing transfers losses to the public and teaches recklessness; not rescuing destroys solvent institutions and spreads damage to people outside the bargain. The decision turns on supervisory knowledge built long before the crisis.