RDP 8811: Monetary Transmission in a Deregulated Financial System 6. Policy Implications

Before discussing the policy implications of the theoretical model, it is important to point out the limitations of the static fixed price Hicksian framework modeled above. First, it is not possible to treat the interest rate as an exogenous variable, even although it is closely influenced by the central bank. When put into a stochastic world, the model does not imply that fixing the nominal interest rate is an optimal policy. Fixing the nominal interest rate would, as discussed initially by Wicksell (1936) and more recently by Sargent and Wallace (1975), lead to price level indeterminacy and/or instability as soon as one relaxes the assumption of fixed prices. As an example, imagine a one-off positive shock to the economy (e.g., an improvement in the terms of trade), this would lead to a rise in income and, starting from initial equilibrium, a rise in inflation. With the nominal interest rate fixed, and inflation rising, the real interest rate will fall. This leads, in turn, to a second-round rise in income and inflation and a further fall in the real interest rate and so on. Any deviation from steady-state induces explosive behaviour.

To avoid such instability, the nominal interest rate needs to be tied to a nominal variable. Until recently, most countries, including Australia, have used nominal “money” growth as a nominal policy anchor. In a world without money is there an alternative nominal anchor? The most obvious alternative anchor is nominal GDP growth which is, in any case, the ultimate objective in a monetary targeting framework. A rise in GDP would then be followed by a policy-induced rise in interest rates, cutting off any inflationary spiral before it could get underway. The stance of monetary policy could be evaluated by comparing nominal GDP growth with the nominal interest rate. If the nominal interest rate was below current nominal GDP growth (and expected to remain there), this allows investors (including the government) to borrow at current interest rates, invest in the domestic economy and pay interest out of future growth. Such a monetary stance is clearly expansionary (and could not be maintained indefinitely without rising inflation). In contrast, if interest rates were allowed to move above the GDP growth rate, policy could be interpreted as contractionary.

Even more direct and simple policy regimes are available. One is fixing the nominal exchange rate, although this strategy is not available to every country (one country must tie down the world average price level). Another regime is to target directly a price level or the rate of inflation. This approach has been suggested by Hall (1983) and others in the New Monetary Economics school. A major problem with the pure price targets is that they largely remove the anti-cyclical or Keynesian role for monetary policy. Without going into the debate over discetionary monetary policy, it is clear that many central banks would be unwilling (or unable) to give up any role for real variables in setting interest and exchange rates. It is perhaps this factor that gives nominal GDP targets some advantage over pure price targets. It should also be stressed that none of the above alternative regimes are necessarily easy to implement. Because of lags in the collection of data and the impact of policy on targets, there will always be substantial discretion and uncertainty involved in the short-run operation of monetary policy.

Indeed, the crucial problem with many simple quantity or price level rules is that they require the authorities to forecast the future impact of current policies on private sector expectations. Such a task is certain to lead to frequent forecast errors and may aggravate rather than smooth economic cycles. Is there a monetary regime that can be neutral to the business cycle in a deregulated financial system, just as monetary targets were supposed to be neutral in a regulated system? This is clearly a neglected area requiring further research. A recent paper by Barro (1988) may provide a fertile starting point. He argues that monetary policy could be directed at stabilizing an interest rate. In the framework outlined above, such a policy is clearly not possible given that the interest rate is basically an operating instrument. However, if a long-term rate is introduced into the model, then targeting the long rate could be a feasible policy alternative. It has the major advantage that it removes the need for central banks to try and forecast their policy target and the effects of policy on the target: this would now be left entirely up to financial markets and would be instantly observable.

For some central banks this strategy would not provide adequate control over inflationary expectations because the long-term real interest rate is not observable and may vary through time. An extension of the Barro proposal would therefore be to target the difference between a nominal and an indexed long-term bond; i.e. the markets long-run inflationary expectations. Whenever the yield on nominal long bonds rose above the yield on indexed bonds policy would be tightened and the yield curve would invert; activity would slow and so expected inflation (and nominal long yields) would fall.

A number of technical and practical problems remain unresolved with many of the above policy proposals. Yet for many countries, including the US, the U.K., Canada, and Australia, where financial deregulation has undermined the simple monetary targeting option, regimes currently in operation are probably best described as some mixture (often ad hoc) of the above systems. Given this state of affairs, further research on alternative regimes is clearly required.