RDP 9709: Asset-price Bubbles and Monetary Policy 4. Financial Regulation and the Inflation Environment

In the above model we assumed that the parameters linking asset price changes to inflation are fixed, and not influenced by other policy instruments or by the general macroeconomic environment. In this section we relax this assumption and discuss how financial regulation and the central bank's target rate of inflation might affect these parameters and thus affect the appropriate monetary policy response to an asset-price bubble.

4.1 The Role of Financial Regulation

So far we have assumed that the authorities have only one policy instrument. If other instruments – such as prudential regulation – are available, the need for monetary policy to respond to changes in asset prices is lessened. Sound bank regulation and supervision are likely to reduce the links between asset price growth and credit growth and reduce the financial system's exposure to a decline in asset prices. This should lessen the asymmetric inflation effects of asset price changes (measured by ?).

Financial regulation might also be able to reduce ?. By leading to tighter liquidity constraints on borrowers as the asset-price bubble grows, appropriate regulation and good credit assessment techniques reduce the scope for higher perceived wealth to be translated into higher consumption, and thus higher inflation.

The greater is the ability of regulatory policy to insulate the economy and the financial system from the effects of asset-price bubbles, the smaller will be the required monetary policy response to an asset-price bubble, and thus the closer expected inflation will be to target and economic growth to potential growth. In this way, sound bank regulation makes it easier for the central bank to achieve its inflation target. In the limit, if regulatory policy were able to reduce both ? and ? to zero, asset price changes need have no implications for monetary policy.

Note, however, that a role for monetary policy remains if regulatory policy is unable to eliminate the symmetric effect of asset price movement on inflation, even if it does eliminate the asymmetric effect (that is, α > 0 and β = 0). In this case, if the asset-price bubble is allowed to run unchecked, the expected variability of inflation and output are still unnecessarily high because of the direct effects of the bubble. Optimal monetary policy would still seek to bring forward the correction in prices and benefit from the fact that once the bubble collapses it will not reappear for some time.^[9] The important point, however, is that using regulatory policy to reduce the asymmetric effect of asset price changes on inflation, reduces the need for monetary policy to respond to asset-price bubbles and reduces the variability of inflation and output.

While sound bank regulation and supervision are likely to reduce the need for the monetary authorities to run tight policy to burst a bubble, the more difficult issue is what form that policy should take. In principle, cyclical changes in banking regulation might be appropriate; for example, increasing loan to valuation ratios or imposing higher capital ratios in the upswing of the cycle might slow a credit/asset-price boom and enhance the stability of the financial system. However, the very real danger with such an approach is that it is likely to induce a shift towards financial institutions that are more lightly regulated. In a sense this was the experience in Australia in the early 1970s (Kent and Lowe 1997).^[10] Tight regulation on banks simply saw the rapid growth of non-bank financial institutions (some of which were owned by banks) and the transfer of risk to these institutions. The lesson from this period is that imposing restrictive regulation on just one part of the financial system leads to a flow of resources to another part of the system. This reduces the effectiveness of the regulation and distorts the dynamic efficiency of the financial system. In fact, it was for these reasons that many central banks around the world embarked on a program of extensive deregulation of the financial system in the late 1970s and early 1980s.

While the case for using changes in prudential regulation to deal with asset price fluctuations is relatively weak, there is a much stronger consensus that well structured prudential policy, even if it does not respond dynamically to movements in asset markets, makes the financial system less prone to instability. Many of the principles of such a regulatory system are set out in the Basle Core Principles. While these measures do not rule out lending booms on the back of asset price rises, they help reduce the costs of asset price booms and busts. As such, they make the task of monetary policy easier and contribute to the stability of both output and inflation.

4.2 Implications of Low Inflation

A second issue is whether the target rate of inflation affects the critical parameters of our model.

We have argued above that asset-price declines can have large effects on output and inflation. However, we have largely ignored the issue of whether the size of these effects depends upon the rate of inflation. While this was useful from a modelling point of view, the asymmetric effects of asset price changes may well depend upon the background rate of inflation. When inflation is low, a given correction in real asset prices must occur through falling nominal prices, rather than asset-price inflation failing to match goods and services price inflation. The Australian experience is instructive here. After the property-price boom of the early 1970s, real property prices declined significantly; while this involved some fall in nominal prices, the high background rate of goods and services price inflation accounted for much of the adjustment in real prices. In contrast, in the low-inflation 1990s, relatively more of the adjustment in real property prices has taken place through the nominal price of property falling (Kent and Lowe 1997).

Since in a low-inflation environment falls in nominal asset prices are more likely, financial institutions run a greater risk of not being able to collect the full collateral value on bad loans. In response, some reduction in average loan-to-valuation ratios may be appropriate. If this does not occur, bubbles in asset markets may be more costly in a low-inflation environment because they increase the risk of difficulties in the financial system. In terms of our model, this means that the size of the asymmetric effect, ?, of a bubble collapsing may be higher in an environment of steady low inflation than in an environment of steady higher inflation (other things being equal). If this is the case, low inflation increases the importance of monetary policy responding relatively early in the life of the bubble.

While low inflation increases the probability of nominal asset price declines, it should also reduce the probability of an asset-price bubble emerging in the first place. In an environment in which inflation is high and variable, property acts as a hedge against inflation and there are also substantial tax advantages to property investments. In a low-inflation environment these advantages are reduced, making speculative increases in property prices less likely.^[11]

While low and stable inflation should reduce the likelihood of an asset-price bubble occurring, it does not guarantee that bubbles will not occur; the Japanese experience in the late 1980s is an obvious example. If a bubble does occur in a low-inflation environment, and financial institutions do not adjust their lending practices, there are perhaps stronger implications for the future health of the financial system than would be the case if inflation was higher. As a result, low inflation may raise the returns to early action to increase the probability of the bubble collapsing.

Footnotes

Using the same type of reasoning, the same is true if α = 0 and β > 0. [9]

Large increases in property prices saw rapid growth of non-bank financial institutions as well as other vehicles for investing in property. When the property price crash eventuated, losses were concentrated in these institutions rather than in the banks, but there were still considerable contractionary effects on the economy. [10]

There is also the question of the likelihood of a bubble emerging during a transition phase between high and low inflation environments. A lowering of the rate of inflation should lead to an increase in the value of real assets because average real interest rates are likely to be falling when moving from a period of disinflation into a period of sustained low inflation. However, determining the extent of this effect is difficult, and improving fundamentals can themselves be conducive to generating bubbles. In addition, a number of authors have argued that low inflation reduces the ‘equity risk premium’. As a result, real equity prices should increase as low inflation becomes entrenched (Modigliani and Cohn 1979; Blanchard 1993). Furthermore, lower nominal interest rates associated with lower inflation can reduce liquidity constraints which can place further upward pressure on asset prices. In Australia, it is not uncommon for financial institutions to set an individual's maximum borrowing capacity for a home loan such that the loan would generate initial repayments equal to 30 per cent of the individual's income. As a result, lower interest rates mean larger loans and more individuals qualifying for a housing loan (Stevens 1997). [11]