RDP 1977-06: Interest Rates and Exchange Rate Expectations in the RBA76 Model 7. Conclusions
October 1977
By any standards this paper raises more questions than it answers. Although it proved possible to include a measure of the company debenture rate in the RBA76 model, and to therefore obtain further evidence about the reactions of asset choices to changes in interest rates, the evidence suggests that short-run movements in the debenture rate have been dominated by the bond rate, rather than the other way around. This is not to deny that market forces have influenced nominal interest rates; indeed, the bond rate reaction function allows for adjustment towards a target rate determined by the monetary growth rate, although the lags in this adjustment are estimated to be relatively long and the traditional assignment variables explain a high proportion of short-run changes in the bond rate. Given that the models which fit the data best have the debenture rate adjusting in the short run to the bond rate, this means that these models react in a broadly similar way as the standard RBA76 model in which the bond rate is a proxy for all domestic nominal interest rates.
The behaviour of the version of the model – Model D – in which the bond rate is constrained to adjust to the debenture rate, which adjusts in turn to the world rate, is sufficiently different to that of the standard versions of the RBA76 model to illustrate the practical importance of the main questions discussed in this paper. Model D does not fit the data as well as the other models considered in this paper although there is some casual evidence from the data in figure 1 which suggests that the direction of causation may be altering in a way which would make Model D more relevant for current analysis. This illustrates the importance of further work on interaction among interest rates in Australia.
Perhaps the most positive finding is the way in which the model-building techniques of the RBA76 project can be used to construct a measure of expected changes in the exchange rate. This measure is conditional on the maintained hypothesis about exchange rate expectations, based largely on relative prices in domestic and foreign goods markets, but since the measure produces sensible results when it is used in converting world interest rates measured in $US to a rate of return in $A, this provides an independent test of the hypothesis. There is, however, urgent need to compare the measure of expected exchange rate changes derived in this paper with other measures which may be available.
