Asset Prices and Central Bank Policy
by Hans Genberg* Executive Director (Research) Hong Kong Monetary Authority
(Presentation at the Conference "The ECB and its Watchers", Frankfurt)
5 September, 2008
I am honoured to have been invited to participate in this prestigious conference and to speak on a topic which has been with us for some time already. As a matter of fact, eight years ago I was the co-author of a report entitled Asset Prices and Central Bank Policy in which we made the following statement:
"…a central bank concerned with stabilizing inflation about a specific target level is likely to achieve superior performance by adjusting its policy instruments not only in response to its forecasts of future inflation and the output gap, but also to asset prices." (Cecchetti, Genberg, Lipsky, and Wadhwani, 2000, p. xix)
The fact that the issue is still alive enough to justify a session in this conference suggests that we were not fully convincing in our arguments, but also that the counterargument, i.e. that a central bank need not pay attention to asset prices in setting its policy, is also not universally accepted. So what is the argument still about? One way to frame this question more specifically is to pose the issue as follows: When it formulates its policy, should a central bank consider the movements in asset prices over and above their influence on the inflation gap and the output gap? This is what I take to be the question at hand.
The objective of central bank policy and the case for paying attention to asset prices
In order to make some headway and to avoid misinterpretation, it is important to have a common understanding of what the assumed objectives of monetary policy are. In a first instance I will stick to the conventional case which we considered back in 2000 where the central bank attempts to minimize some combination of fluctuations in inflation around a target value and fluctuations of output around its natural level (the output gap for short). In view of the lags with which monetary policy operates on the economy, it is clear that the central bank must be forward looking and set its policy in response of deviations of future expected inflation and output levels from their respective targets. From this it follows trivially that if the central bank takes into account the expected levels of inflation and output at all future horizons, weighted appropriately by some discount factor, then all relevant information is already factored in and there is no reason to include asset prices, or for that matter monetary aggregates, in the decision process. So to make the issue interesting we must consider the case in which the central bank focuses on some particular horizon, two-years say, as this is commonly used by some central banks.
In our 2000 report, we addressed the issue by simulating a theoretical macro model which included a process that could generate asset price misalignments under different assumptions about the form of the monetary policy rule. Our conclusion based on these simulations was that giving a small weight to asset price developments improved the performance of the economy as judged by an ad hoc welfare function which depended on the present discounted value of inflation and output deviations. In an earlier paper I had used an estimated model of the Swiss economy to illustrate a similar point, namely that including the exchange rate (in a floating exchange rate context) in the policy rule together with the inflation rate and the output gap would be beneficial.
It is of course possible to construct counter examples where adjusting a policy interest rate in response to asset price misalignments would be counterproductive. The 2005 paper by Gruen et. al. is such an example (Gruen, , Plumb, and Stone, 2005). However it is also not robust to modification in the assumed stochastic process for asset prices (Haugh, 2008). So the conclusions we can draw appear to be model dependent which is a nuisance. But this does not mean that we should ignore potentially relevant information only that we can not react to it mechanically without reflection.
Objections and responses
A number of additional arguments have been brought up purporting to show that reaction to asset price developments would be a mistake. Some of these are based on a different notion from what I have in mind of what 'reacting to asset prices' means.
- One objection would argue that stabilizing asset prices would require such large adjustments in the policy interest rate that it would be destabilizing for inflation, output and employment. While this may be the case, it is not relevant for the issue as I see it. When we advocated taking asset price developments into account we made it clear that this was for the purpose of stabilizing inflation and output, NOT of stabilizing asset prices themselves. For this reason the reaction to asset prices is likely to be relatively muted, i.e. it would be to lean against the wind of asset price developments not attempting to target some particular value.
- A second objection claims that targeting an asset price is dangerous because we do not know what its equilibrium value is. As with the previous point, this argument is based on the idea that the central bank should try to achieve some particular value of the asset price, i.e. that it should target asset prices. This, I believe, is not what anyone has in mind. Of course, as with the output gap, the central bank does need to base its actions on an estimate of the difference between the actual and the equilibrium value of the variable it monitors. When there is uncertainty about the equilibrium value, it would be important to allow for some margin of error within which no action is taken as suggested by Haugh (2008). In other words, the policy reaction would be subject to some threshold effect. In addition, it may be useful to combine asset price information with information about credit or money growth in the economy to gauge whether there is a case for policy adjustment. (Borio and Lowe, 2002) The implication is that theoretical and empirical modelling efforts must incorporate non-linearities explicitly in order to be informative.
- A third objection suggests that monetary policy should not worry about the build-up of asset price bubbles, but should react swiftly and forcefully to the bursting of the bubble. This argument for an asymmetric response is based on the idea that it is difficult to identify the emergence of a bubble, because the build-up is incremental in nature whereas it is obvious when a bubble bursts. In addition, the sharp decline is the asset price associated with the bursting of the bubble can be very costly. This argument seems to stand William McChesney Martin's dictum on its head, namely we should keep the party bowl well filled because we don't really know the capacity of each party goer to hold his or her liquor. Instead we should just be ready to clean up the mess if everyone gets too drunk. This asymmetry seems a recipe for moral hazard, because by promising to clean up after a bust in asset prices we may make investors less likely to exercise appropriate caution on the way up. Of course, I am not arguing that the central bank should stand idly on the sidelines if financial instability and recession are brought about by a sharp asset price decline. I am arguing, however, that tightening policy somewhat as the party gets rowdy would lead to a better outcome.
Financial stability as an additional objective
The build-up of unsustainable asset valuations followed as they usually are by large and rapid declines may create strains in the financial system with secondary negative effects on employment and output. It may therefore be argued that reacting to asset price misalignments would also be justified as a way to prevent financial stress and thereby adverse macroeconomic outcomes. Ensuring financial stability more generally may be legitimately added to the objectives of the central bank. But care must be taken not to overburden monetary policy in the process.
In principle, for a policy maker to achieve N policy objectives it is necessary to be able to control an equal number of independent policy instruments. A central bank that is trying to achieve an inflation target while at the same time minimizing variability of output should ideally have two instruments, and adding a third objective related to financial stability would require an additional instrument. We may need to look beyond the traditional tools of the central bank for this.
For the purpose of limiting output fluctuations when the economy is hit by shocks of various kinds it would be useful if other measures and policies than interest rate adjustments were in place. Automatic stabilizers in the fiscal system come to mind, but one might also think of institutional arrangements that render the economy more flexible to deal with shocks. Similarly, for the purpose of securing financial stability one should not rely principally on interest rate policy, but one should encourage the development of appropriate prudential regulations and practices. These might include rules that are capable of reacting to the economic cycle much as the automatic fiscal stabilizers do in that area. The system put in place by the Spanish authorities comes to mind. This system involves requiring financial institutions to make capital provisions which are based on the state of the business cycle in a way that make them counter-cyclical. (See Fernández de Lis, Martínez, and Saurina, 2001.)
Having additional policy instruments to deal with additional objectives does not imply that interest rate policy should be conducted completely independently of those other policy instruments. Coordinated actions involving all instruments are surely more efficient. Therefore, even if macro-prudential instruments were developed we should still take information in asset markets into account when we conduct monetary policy.
It is time to conclude and I will do so by summarizing my main arguments in three points.
- A central bank that practices flexible inflation targeting will achieve improved outcomes if it takes asset prices into account over and above their effect on the standard-horizon inflation forecast.
- Nonlinearities in the form of threshold effects or the joint presence of unusual developments in asset prices and credit/monetary aggregates may need to be incorporated in both modelling and implementation.
- If financial stability is added as an additional objective of policy, then an additional instrument should also be looked for. The use of this additional instrument should be coordinated with the use of the interest rate instrument.
Thank you very much for your attention.
Borio, C. and P. Lowe, 2002. "Asset prices financial and monetary stability: exploring the nexus". BIS Working Papers, no. 114.
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.
Cecchetti, S., H. Genberg, and S. Wadhwani, 2002. "Asset prices in a flexible inflation targeting framework". In Asset Price Bubbles: the Implications for Monetary, Regulatory, and International Policies (eds. W. C. Hunger, G. G. Kaufman, and M. Pomerleano). MIT Press, pp. 427-44.
Fernández de Lis, S., J. Martínez, and J. Saurina 2001. "Credit growth, problem loans and credit risk provisioning in Spain, BIS Papers, no. 1, March, pp. 331-53.
Gruen, D., M. Plumb, and A. Stone, 2005. "How should monetary policy respond to asset price bubbles?" International Journal of Central Banking, Vol. 1, no. 3, pp. 1-31.
Haugh, D., 2008. "Monetary policy under uncertainty about the nature of asset price shocks", momeo.
Wadhwani, S., 2008. "Should monetary policy respond to asset price bubbles? Revisiting the debate." Keynote Speech at the SIERF Colloquium, Munich, 12 June, 2008.