Comments on the IMF WEO October 2009 “Lessons for Monetary Policy from Asset Price Fluctuations”
Since the mid-80’s the global consensus has been that central banks should orientated their policy only towards the control of retail price inflation. The development of the financial crisis has sparked a renewed debate on enlarging to role of central banks to include financial market price stability in their domain. Various central banks are currently supporting different forms of intervention to reduce the impact of the crisis on the financial system. The IMF study appears at a time when a more structured approach to intervention by central banks is being discussed in many countries.
Despite this recent intervention, the central banks have generally limited their intervention in capital markets to consultation on market regulation. The core role of central banks remains that of lender of last resort to the banking system and guardian of monetary stability. Monetary stability has generally been defined as stable retail prices via guided by short term interest rates (and indirectly by controlling the money supply). The place many central banks used to play in the regulation of financial markets has been moved to independent agencies.
Nevertheless, there have been frequent calls for central banks to intervene to dampen rising asset prices to avoid bubbles (with higher interest rates) and to support soft markets (with interest rate cuts). Mostly central banks have only responded to such calls in the context of the indirect implications for consumer price inflation, their primary goal.
Three key questions for central bank intervention on asset prices
We can identify 3 key questions on central bank intervention to control asset inflation :
• Why should asset inflation be targeted ?
• How can central banks recognize asset inflation ?
• Which tools does a central bank have to effectively influence asset price inflation ?
The IMF study looks at the first two questions. The study pays particular attention to the predictive power of various indicators for recognizing asset inflation. This is clearly an important point for central bank policy- asset price inflation must be recognized as it builds, if any kind of effective counter measures are be taken. Why should asset inflation be targeted ?
On the face of it, there does not seem to be much of a case against rising asset prices. The general perception is that rising stocks and rising housing prices are a reflection of economic growth and optimism for the future. Admittedly, housing prices directly impact directly prices as experienced by households. Therefore, housing prices are a component in CPI in most countries. Rising share prices on the other hand have no direct link to inflation. Attempts at verbal intervention by high profile central bankers warning of over-heating in asset markets are usually met by hails of dismay from the stock market, bankers and government officials alike.
There can be good reasons for rising share prices which are not inflationary. In sectors where technological, political or organizational changes promise to create greater profit opportunities in the future above trend share price increases are not inflationary. The future profit flow discounted to the present value may justify significant share price increases.
It is harder to make a case for the rising productivity of housing. Still, better quality housing (energy saving) etc. can justify price increases. Apart from this, we would expect price increases to be inflationary, increasing parallel to CPI, with a small factor for crowding out by commercial real-estate and for real economic growth (increasing marginal propensity to purchase property) and demographic factors. The IMF tried to neutralize these long-term trends in the study.
The problem with asset price inflation from the point of view of monetary stability lies in the cyclical nature of asset prices movements and the secondary effects in the economy. Periods of boom are succeeded by periods of bust. If prices just went up and down, without impacting overall economic activity there would be little reason for any attempt at regulation. However, the social costs in terms of lost output and economic dislocation can be severe. The IMF study looks at the correlation between changes in the output gap and asset price busts in housing and in shares.
Asset price bust = recession ?
The IMF study could not identify a clear positive correlation between a share price bust and an increasing output gap. This does not mean that there is no connection, just that there are also numerous share price busts without measurable impact on the economy. A bust in housing prices on the other hand, showed fairly good correlation with an increasing output gap*.
[*The social cost of changes in economic activity can be summarized by the notion of changes in the output gap. The output gap attempts to measure the difference between the full utilization of the economic potential present in the economy and actual use of this potential. The difference is the social cost of a level of sub-optimal economic activity.]
On the face of this analysis, the social implications of asset price inflation and a subsequent price down-turn in shares does not clearly warrant central bank intervention. On the other hand, the current recession (and the experience of the 1929 crash and subsequent depression) indicates that there may be major risks of serious social costs due to an economic downturn if share prices adjust downward too rapidly or if this movement coincides with other developments.
While the IMF study could not show good correlation between a share price slump and changes in the output gap, central banks may reasonably view indications of a bubble in share prices as a warning. Clearly, if shares and housing prices boom and slump at the same time, the potentially serious impact on economic activity may be compounded in the subsequent period. A slump may affect financial markets simultaneously in various ways : bank capital will fall ; bank lending policy will become risk adverse and collateral values for mortgages will fall. It has been a favorite topic of market discussion in the last 10 years that central banks somehow had to engineer a soft landing from a share price boom. The IMF study shows that successfully intervening to smooth cycles in housing prices would make much more difference to the economy than actions to smooth share price slumps.
Some of the main points noted in the study
Above trend asset price increases preceded busts both in housing and in shares in the period since 1985. The period before 1985 was not seen as very relevant, as the most significant asset price busts in that period were related to oil price shocks. Also central bank policy and market regulation was much less consistent across the sample countries in the earlier period.
Several factors were identified which gave a fair indication of an approaching asset price bust, which did not also give too many false signals. These were credit growth and the investment/GDP ratio for both housing and share price busts. Growth in the current account deficit was positively correlated with a subsequent housing price bust. Output growth was positively correlated with a subsequent share price bust.
Can asset inflation be recognized ?
The IMF study concludes that central banks can access indicators with some predictive power for cycles in asset prices. However, central banks cannot use these indicators to guide monetary policy automatically. The indicators visible in the run up to a bust are not sufficiently distinct to distinguish between asset price increases due to strong output growth resulting from positive productivity shock and asset price inflation before a bust.
The effects of an asset price shock in the housing market on output growth are significant. A bust in share prices, on the other hand, does not show a statistically relevant impact on economic output.
The new paradigm theory- positive productivity shock
The matter of central bank management of potential inflationary impacts during a productivity shock was discussed as the new paradigm, notably by Alan Greenspan in the 1990’s. The reason for the interest in this subject lies in the assumption that asset price inflation may indicate approaching general inflation pressure, thus conflicting with the central bank’s goal of monetary stability.
To quickly summarize the “new paradigm” theory of the 1990’s : a positive productivity shock could permit above trend growth without consumer inflation. Large share price increases may be justified in this case, as they discount a change in the power to generate future profits. While Greenspan remained cautious about this theory, he and the Fed largely appeared to accept it in the 1990’s faced with the apparent paradox of the US economy, which had grown 27% in three years to 1997, while inflation remained modest. The Fed continued to leave interest rates low, while CPI continued to remain fairly low too.
Various crises encouraged the Fed to continue a policy of low rates (the Asian Crisis, Long Term Capital Management, Russian bonds and finally the internet bubble bursting). Critics pointed out that cheap central bank money may have fuelled the booms which led to some of these crises. However, other circumstances, such as shrinking government debt markets and low Japanese rates were arguably equally important. The Fed remained watchful. Greenspan : … “The argument for the so-called new paradigm has slowly shifted from the not unreasonable notion that productivity is in the process of accelerating, to a less than credible view…that we need no longer be concerned about the risk that inflation can rise again.” But rates were generally kept low.
Fed funds fell from 6% in early 1995 to 1% by 2004, 19 decreases and 7 increases. CPI in this period largely remained below 3%. It is possible that the lack of inflationary pressure had at least as much to do with Chinese exchange rate policy and Chinese productivity as with American productivity. Still, a more aggressive Fed policy could not really be justified in terms of CPI. The DJIA went from around 4500 to around 10 000 in this period.
Does a central bank have the tools to intervene in asset markets ?
Apart from their role as lender of last resort, central banks possess the power to set the short term lending rate to the banking system via their open market operations. The central bank can use this rate to encourage or discourage borrowing by the banking system, with a subsequent indirect impact on the money supply and on inflation. Other instances also play a role in this transmission mechanism.
Minimum reserves banks are required to hold and the capital adequacy criteria influence the way central bank monetary policy is transmitted to the economy. Institutional friction in lending practice in banks and the range of products available in the market will also play a role in transmitting central bank policy to the economy. Central banks do not generally possess a tool to directly impact bank lending.
If central banks were to try to achieve not only the goal of monetary stability (measured by consumer prices), but also the goal of monetary stability (measured by asset market inflation), then only one tool of short term lending would not be sufficient. Central banks would require what is being called a macro prudential tool to target bank lending directly.
As the IMF study points out : there is no significant correlation between changes in leading interest rates and asset price movements. This indicates that the current tool of managing short term lending rates to the banking system is not well adapted to managing asset markets. Anyway, in view of the weak correlation between share price busts and the output gap, there appears to be little reason for central banks to focus on share price fluctuations alone. An effective macro-prudential tool should focus on bank lending which impacts the asset prices in housing.
We can imagine the central bank intervening in the market not only to manage the short term lending rate, but also to directly target bank lending. This could, for instance, be managed by intervening to increase or reduce the spread between the current short-term rate and commercial lending rates.
Recent worries have not been about asset price inflation- It is deflation which has been the problem. The measures put in place by central banks have been crisis measures, many of them directly targeting bank lending. We can assume that these crisis measures will not work in the same way in a boom as in a bust. The current crisis measures range from penalties (Sweden) to support (USA). Looking forward to the end of the crisis, we can be certain that it will be easier for the central bank to withdraw a penalty than a support, without upsetting the market. Mervyn King (Governor of the Bank of England) has presented the notion of applying a spread to central bank money to boost bank lending. At the moment he is talking about a negative rate to encourage commercial lending.
The Riksbank in Sweden has already introduced such a penalty spread. Presently, the aim of these measures is to encourage bank lending (via a negative deposit rate on bank deposits at the central bank), but a similar tool could be applied to influence lending in boom times too. This would presumably work by managing the spread between central bank short term lending rates and commercial lending rates. The ECB does not currently have a tool to directly target bank lending and there does not seem to be a discussion on this matter at present. The Fed introduced a program of buying mortgage backed securities to encourage lending in that sector as a crisis measure. Ben Bernanke is now faced with removing this emergency measure without upsetting the market or replacing it with something more permanent. We know that Bernanke has been opposed to measures to manage asset price inflation in the past.
Other relevant factors
Central bank policy is obviously not the only policy variable with a strong impact on the market environment and on the economy. Fiscal and regulatory policies play an important role in influencing how banks and markets react.
High levels of government debt, especially when combine with a large current account deficit can influence both the way an asset boom runs and the way the economy reacts to a bust. The IMF study indicates good correlation between the development of a large current account deficit and the above trend inflation in asset prices. While this is a good predictor for a more restrictive policy via a macro prudential instrument, one could argue that direct government measures to control the government deficit could be a more promising.
The regulatory environment is also an important factor in financial stability. The development of large unregulated sectors in the financial markets had proved to be risky for the stability of the banking system. Permitting financial products to be originated and traded with no obligation to maintain a ratio to the underlying cash product has proved to be a source of systemic risk. Developments indicate that in liquidity and capital requirements for some business areas in banking were inadequate.
These fiscal and regulatory issues are not in the hands of the central banks. Effective use of a macro prudential instrument directly targeting bank lending would not change this framework. Nevertheless, in the end a financial crisis is always a credit crunch. Banks lend money to bad guys who do not pay back. A direct central bank tool to manage bank lending rates combined with adequate forecasting instruments to effectively smooth asset price cycles could be a useful additional tool to help central bank monetary management of financial stability.
Lessons for Monetary Policy from Asset Price Fluctuations World Economic Outlook Published by the IMF in October 2009.
The main authors of this chapter are Antonio Fatás, Prakash Kannan, Pau Rabanal, and Alasdair Scott, with support from Gavin Asdorian, Andy Salazar, and Jessie Yang. With support from Jordi Galí.