The credit crisis that hit the world economy in August teaches us many lessons about the workings of integrated financial markets. It also teaches us a lesson about the responsibilities of central banks.1
Until the eruption of the credit crisis, the consensus view was that central banks should target inflation, and that’s pretty much all they should do2. In this view, central banks should not target (or try to influence) asset prices, either - as was stressed by Alan Greenspan - because central banks cannot recognise bubbles ex ante, or - if they can - the macroeconomic consequences of bubbles and crashes are limited as long as central banks keep inflation on track. Inflation targeting, we were told, is the new 'best practice' central banking that makes it unnecessary for central bankers to try to influence asset prices3.
The credit crisis has unveiled the fallacy of this hands-off view. If the banking system were insulated from the asset markets, the view that monetary policies should not be influenced by what happens in asset markets would make sense. In that case, asset bubbles and crashes would only affect the non-banking sector, and a central bank is not in the business of insuring private portfolios.
The problem that we have seen in the recent crisis is that the banking sectors were not insulated at all from movements in the asset markets. Banks were heavily implicated both in the development of the bubble in the housing markets, and its subsequent crash. And since the banking system was heavily implicated, the central banks were also heavily involved by the very fact that they provide insurance to the banks in the form of the lender of last resort. Some may wish that central banks would abstain from supplying this insurance. Economic theory, however, tells us that central banks should intervene to provide liquidity if the liquidity crisis risks disrupting the payments system, thereby hurting many “innocent bystanders”. In addition, reality ensures that central banks are forced to provide liquidity when a crisis erupts, as they are the only institutions capable of doing so.
Thus, when asset prices experience a bubble, it should be a matter of concern for the central bank because the bubble will be followed by a crash, and that’s when the balance sheet of the central bank will inevitably be affected. It is not reasonable for a central bank to argue that asset bubbles and crashes should not be a source of concern and therefore that it should not try to intervene when a bubble arises, when it knows that the bubble will have large implications for its future balance sheet, and its profits and losses.
There is a second reason why the hands-off approach has been shown to be wanting. During the last few years, a significant part of liquidity and credit creation has occurred outside the banking system. Hedge funds and special conduits have been borrowing short and lending long, and as a result, have created credit and liquidity on a massive scale, thereby circumventing the supervisory and regulatory framework. As long as this liquidity creation was not affecting banks, it was not a source of concern for the central bank. However, banks were heavily implicated. Thus, the central bank was implicitly extending its liquidity insurance to institutions outside the regulatory framework. It is unreasonable for a central bank to insure activities of agents over which it has no oversight, very much as it would be unreasonable for an insurance company selling fire insurance not to check whether the insured persons take sufficient precautions against the outbreak of fire.
So, what can be done about this? There are two possible solutions. The first one is for the central bank to recognise that asset bubbles are a source of concern and that it should act upon the emergence of such a bubble. The argument that a bubble can never be recognised ex ante is a very weak one. One had to be blind not to see the bubble in the US housing market, or the internet bubble. And this is the case for most asset bubbles in history. When asset prices increase at a rate of 20% or more per year, and when credit aggregates increase by similar percentages in a sustained way during several years, one can be pretty sure that a bubble is on the move, and that a crash is imminent.
It has been argued that even if central banks can detect bubbles, they are pretty much powerless to stop them. This argument is not very convincing. It is not inherently more difficult to stop asset bubbles than it is to stop inflation. And central banks have been very successful at stopping inflation.
This is not an argument to target asset prices. Few economists today would make that argument. What is possible, however, is a “leaning against the wind” approach, whereby the emergence of a bubble leads the central bank to tighten policy more than it would do otherwise. This was in fact proposed by the ECB in its Monthly Bulletin of April 2005.
Second, central banks should be involved in the supervision and regulation of all institutions that create credit and liquidity. The UK approach of dissociating monetary policy from banking supervision has not worked. Central banks are the only insurers against liquidity risks. Therefore they are the ones who should control those who create credit and liquidity. Failure to do so will continue to induce agents to create excessive amounts of liquidity, endangering the financial system.
The fashionable inflation targeting view is a minimalist view of the responsibilities of a central bank. The central bank cannot avoid taking more responsibilities beyond inflation targeting. These responsibilities include the prevention of bubbles and the supervision of all institutions that are in the business of creating credit and liquidity.
1 This is an expanded version of an FT-column published on November 2, 2007
2 An influential paper making the case that central banks should not try to influence asset prices is Bernanke, B. and Gertler, M. (2001), Should Central Banks Respond to Movements in Asset Prices?, American Economic Review, May, pp. 253-57. Although this has become the consensus view, there are prominent dissenting views also. An example is Cecchetti, S., H. Genberg, J. Lipsky and S. Wadhwani (2000): Asset Prices and Central Bank Policy, Geneva Report on the World Economy 2, CEPR and ICMB. Among the major central banks it is remarkable that the ECB has defended the view that central banks should lean against the wind when asset bubbles arise, (see Monthly Bulletin, April 2005).
3 Proponents of this view have argued that flexible inflation targeting that takes a sufficiently long term perspective is sufficient to deal with asset bubbles, i.e. flexible inflation-targeting can be tailored in such a way that the longer-run consequences of asset prices are taken into account when setting interest rates. (see Charles Bean, Asset Prices and Monetary Policy, Federal Reserve Bank of Australia, Nov. 2003)