RDP 9709: Asset-price Bubbles and Monetary Policy 1. Introduction

What role do asset prices have in a monetary-policy framework that involves targeting medium-term inflation of prices of goods and services? This question has attracted considerable interest in recent times, in part due to the coexistence in some countries of low inflation of prices of goods and services and rapid increases in asset prices, most notably stock prices. In this paper, we argue that the main reason that a monetary authority with an inflation target might respond to movements in asset prices is that these movements can create future difficulties in the financial system, which in turn can affect future output and inflation. Furthermore, we argue that the effect of asset price changes on output and inflation is asymmetric, with declines in asset prices having stronger effects on output and inflation than do increases in asset prices. The asymmetry arises from the adverse effects that a collapse in asset prices can have on the process of financial intermediation.

Of particular importance are the prices of assets used as collateral for loans, for increases in these prices can generate the backing for additional loans and, in turn, a boom in credit growth. If the price increases turn out to be unsustainable, and a correction in prices occurs, much of this collateral can disappear, causing large losses for financial institutions and other firms. The outcome can be a pronounced and protracted slowdown in economic activity, and inflation below target.

In principle, the adverse effects of asset-price bubbles on the stability of the financial system can be moderated through appropriate financial system regulation and supervision. Nevertheless, provided that the effects of asset-price bubbles on the economy are not entirely eliminated, a role for monetary policy may remain. In particular, if the central bank can affect the probability of an asset-price bubble bursting by changing interest rates, it may be optimal to use monetary policy to influence the path of the bubble, even if it means that expected inflation deviates from the central bank's target in the short term. By seeking to change the path of the bubble, monetary policy may be able to simultaneously contribute to maintaining financial system stability and to reducing the variance of inflation.

The reasoning behind this result is relatively straightforward. Suppose that the real asset price is above, and moving further away, from some fundamental level – that is, that there is an asset-price bubble. As the asset price rises it is likely to put some upward pressure on goods and services price inflation by increasing aggregate demand and expectations of inflation. When the bubble eventually bursts, these inflationary pressures will abate, but we argue that there will be an additional contractionary effect on inflation arising from the adverse effect that falling asset prices have on collateral values, financial system stability and the provision of intermediated finance.

There are two broad scenarios to consider. The asset-price bubble could burst in the near future or it could burst in the more distant future. If it bursts in the near future, there is likely to be a mild contraction in output, and inflation is likely to be below target. If on the other hand, the bubble survives and grows, there will be continuing expansionary effects on output and inflation. But eventually the bubble must burst, and when it does so, the collapse will be much larger with the potential to adversely affect the stability of the financial system. When this occurs we could expect a prolonged period of output below potential and inflation (substantially) below target.

Now suppose that the central bank, can increase the likelihood of a bubble bursting by raising interest rates. In this case it may make sense for the central bank to do so early on in the life of the bubble, even though this will increase the likelihood of inflation being below target in the near term. This is desirable, however, because it also decreases the chance of the bubble continuing, and hence, of much more extreme outcomes for inflation and output in the longer term. Of course the path of interest rates will depend on many factors, including the probability of the bubble bursting of its own accord, the expected growth rate of a surviving bubble, the magnitude of the bubble's direct effect on inflation, as well as the magnitude of the asymmetric effect of the bubble's collapse on inflation through reductions in intermediated finance. The objective of our paper is to develop a formal framework to help analyse the interplay of these different factors.

The structure of the paper is as follows. Section 2 outlines various arguments why monetary authorities might be concerned about asset prices. These arguments are discussed in the context of our modelling strategy. Section 3 uses a simple model to analyse the links between monetary policy and asset-price bubbles. Section 4 presents a discussion of how financial regulation and the inflation environment might affect the parameters of the model and thus affect the appropriate monetary policy response to an asset-price bubble. Finally, Section 5 concludes by drawing some broad lessons for monetary policy.