Bagehot, central banking, and the financial crisis

Xavier Vives 31 March 2008



The present financial crisis poses two main questions: whether it is similar to past crises and how central banks should intervene to preserve the stability of the system.

The current financial turmoil seems extraordinary because it has unexpectedly affected the heart of the functioning of our sophisticated money markets. Despite the Northern Rock episode, the main contours of the current crisis seem very distant from scenes of crises past where newspapers were covered with photos of depositors queuing to withdraw their money during a panic. Yet this crisis is just a modern-market form of a traditional banking crisis.

An old-fashioned bank run happened if enough people tried to withdraw their funds from a bank; even if the bank was solvent, it might not be able to meet all the withdrawals and thus the fear of bank failure could become a self-fulfilling prophecy. In the current crisis, participants in the interbank market take the place of long queues of withdrawers. They have stopped extending credit to other banks that they suspect to have been contaminated by the subprime loans and which therefore may face solvency problems. The commercial bond market and structured investment vehicles are facing similar trouble.

Both the old and new forms of crisis have at their heart a coordination problem. In the current one, participants in the interbank market and in the commercial bond market do not renew their credit because of fear others will not either. Witness the demise of the investment bank Bear Stearns at the heart of the dealing on structured vehicles.

In reaction, central banks have intervened massively, injecting liquidity and allowing banks to access fresh cash at the discount window in exchange for collateral that includes the illiquid packages of mortgage obligations. Have central banks done the right thing or are they provoking the next wave of excessive risk taking by bailing out banks and markets? Is monetary policy the only tool available for the central bank to address the market crisis?

Bagehot’s wisdom

Bagehot advocated in 1873 that a Lender of Last Resort in a crisis should lend at a penalty rate to solvent but illiquid banks that have adequate collateral. The doctrine has been criticised as having no place in our modern interbank market, but this is wrong. Bagehot’s prescription aims to eliminate the coordination problem of investors at the base of the crisis. It is still a useful guide for action when the interbank market stalls.1 It makes clear that discount-window lending to entities in need may be necessary in a crisis.

Bagehot's doctrine, however, is easy to state and hard to apply. It requires the central bank to distinguish between institutions that are insolvent and those that are merely illiquid. It also requires them to assess the collateral offered. Central banks, because of information limitations, are bound to make mistakes, losing face and money in the process. This doesn’t mean they should not try.

Poor collateral versus massive liquidity

The collateral should be valued under “normal circumstances”, that is, in a situation where the coordination failure of investors does not occur. This involves a judgment call in which the central bank values the illiquid assets. A central bank that only takes high quality collateral will be safe, but will have to inject much more liquidity and/or set lower interest rates to stabilise the market. This may fuel future speculative behavior. Some of this may have happened in the Greenspan era, in the aftermath of the crisis in Russia and LTCM, and after the crash of the technological bubble. The ECB and the Federal Reserve have accepted now partially illiquid collateral that the market would not. This seems appropriate and releases pressure to lower interest rates to solve the problem, something that should be done only if there are signs of deterioration in the real economy. The problem is that central banks are extending the lender of last resort facility outside the realm of traditional banks to entities, like Bear Stearns, that they do not supervise and, therefore, over which they do not have first hand information. How does the Fed know whether Bear Stearns or other similar institutions are solvent? It seems that the Fed is not following Bagehot’s doctrine here.

Finally, if banks and investors are bailed out now, why should they be careful next time? This is the moral hazard problem: help to the market that is optimal once the crisis starts has perverse effects in the incentives of market players at the investment stage. The issue is that only when the moral hazard problem is moderate does it pay to eliminate completely the coordination failure of investors with central bank help. When the moral hazard problem is severe, a certain degree of coordination failure of investors - that is, allowing some crises - is optimal to maintain discipline when investing and, amending Bagehot, some barely solvent institutions should not be helped.

Therefore, a key question to assess the future consequences of current central bank policy is whether the subprime mortgage crisis arises in the context of a moderate or a severe underlying moral hazard problem. The important extent of asymmetric information in this market points to a severe problem. Be as it may, this issue will determine whether current help will plug the hole for good,or only temporarily to make a larger one in the future. The challenge for central banks is to find the right balance between preserving current stability and imposing discipline for the future. Bagehot’s doctrine is still a reference today.


1 See X. Vives and J.C. Rochet (2004)”Coordination Failures and the Lender of Last Resort: Was Bagehot Right after all?” Journal of the European Economic Association (



Topics:  Financial markets

Tags:  Central Banks, subprime crisis, liquidity, financial stability, asymmetric information

Professor of Economics and Finance and Academic Director of the Public-Private Sector Research Center at IESE Business School; CEPR Research Fellow


CEPR Policy Research