Why Are There So Many Banking Crises?
By Jean-Charles Rochet
Princeton University Press, $50

In the midst of the worst financial hurricane for decades, a book that asks why there are so many banking crises looks more than timely. That said, the current decade has been marked by fewer such upheavals than the period between the mid-1970s and the millennium, which had to cope with shocks stemming from the breakdown of the Bretton Woods exchange rate system and a global trend towards financial deregulation.

The consensus among central bankers before the interbank market seized up last August was that while financial crises had become less frequent in the 21st century, their impact would probably be exceptionally severe when they happened. For once, the consensus was right.

Crises are endemic because commercial banking is fundamentally unstable. This reflects the fact that deposits, which can be withdrawn on demand, are used to finance illiquid loans. The snag is that if everyone asks for their money at the same time in a bank run, even solvent banks will fail.

Until the Northern Rock saga, it was assumed that deposit insurance had eliminated the risk of a 19th-century-style run in which panicky depositors queued to retrieve money. Most recent runs have been “silent”, with the damage done by financial institutions that fail to renew short-term IOUs such as certificates of deposit.

The other means to stabilise unstable banking systems involves the central bank acting as lender of last resort where the collapse of a bank would pose a threat to the whole financial system through contagion. Since the mid-19th century, central bankers have increasingly followed Walter Bagehot’s doctrine that they should lend in a crisis on good collateral to solvent banks, while making clear their willingness to lend without limits to pre-empt contagion.

Yet in most bank bail-outs, illiquid banks turn out to be insolvent. Last-resort lending, like deposit insurance, introduces moral hazard to the system. The existence of a safety net reduces incentives to monitor and discipline risk-taking managers. This in turn has given rise to banking regulation and supervision. Yet in spite of a huge regulatory infrastructure that now operates at the international level via the Basel Accord, crises keep coming.

Among economists’ explanations are moral hazard, ill-judged capital adequacy rules and the incompetence of supervisors. Jean-Charles Rochet, a leading authority on banking, argues the real problem lies with politicians who too often insist on rescuing insolvent banks for short-term reasons of their own. He also highlights the dangers of failing to impose timely corrective measures at ailing banks and the perverse incentives that arise if shareholders and managers are not expropriated in the event of failure – a point highly relevant to the nationalisation of Northern Rock.

The author questions the efficacy of market solutions here. Investors and markets behave erratically in crises and, if people assume that the government will intervene, the market cannot work. He has no time for central bankers who resort to “constructive ambiguity” over whether they will bail out banks. This remedy for moral hazard does not lend itself to accountability.

His preferred solution to the core political issue is to ensure the independence and accountability of bank supervisors along the lines that now apply to monetary policymakers. Rochet would like to see more weight given to the Basel provisions for supervision and market discipline, including explicit rules about how and when supervisors should intervene and the debt instruments to be used to promote market discipline.

This proposal would operate in a wider framework whereby banks that have large exposure to macroeconomic shocks would be denied emergency liquidity assistance from the central bank. Banks with low exposure would have access to the lender of last resort, but with capital ratios and deposit insurance premiums that increase proportionately with exposure. This is elegant, but devising rules for intervention would be hard. Even if politicians could be persuaded to grant independence to supervisors, it is hard to believe they would keep away where liquidity assistance was denied by the central bank in politically sensitive circumstances, even if retail depositors were absent from the scene.

Potential readers should be warned that much of the book is highly technical. There are odd lapses too. The author gets the name of the Bank for International Settlements wrong. Overend, Gurney, which crashed in 1866, is mis-spelt. While there are ample references in the text, there is no index. Yet whatever the verdict on the policy proposals, the book makes interesting reading in the current circumstances.

The writer is an FT columnist