The blog has been searching the websites of the major central banks, such as the IMF, World Bank, Federal Reserve and Bank of England, for research on the history of credit crises. Several readers, including Paul Noble of Parsons Brinckerhoff, have also kindly forwarded helpful studies.
The most comprehensive study that it has found analysed 33 banking crises between 1977-2002 and concluded:
• The average length of each crisis was 4.3 years
• The median loss of GDP was 7.1%
• Major crises (such as today's) caused GDP losses of at least 10%.
• GDP losses can double if the banking crisis leads to a currency crisis
The studies also suggest that lack of effective government action (eg depositor guarantees and liquidity support) causes even greater GDP losses. The US Depression led to 30% of GDP being lost.
Another key message from the research is that even "successful" government intervention comes at a high price. This is because it causes banks to lower their risk profile in two key ways:
• They prefer to hold government debt rather than make corporate loans
• They only lend to the very safest borrowers
This change in risk profile means that government intervention has the side-effect of breaking the process by which banks provide credit for the real economy. Inevitably, therefore, credit crunches are deflationary.
History's lessons on the likely course of today's crisis are thus not encouraging. Governments will initially find it easy to borrow, but face the risk of a currency crisis if foreign lenders begin to suspect they will never be able to repay the money borrowed. Companies however, will find it more difficult to borrow, as banks "de-risk" their balance sheets.
Consumers therefore face an increased risk of unemployment, and so will tend to save more, rather than spend money. In turn, this will reduce demand - further pressuring companies, and government's ability to provide fiscal stimulus.