Is Long Rate Decoupling From Short Rate? Revisiting Expectation Hypothesis Of Australian Term Structure Of Interest Rate

Background and statement of the problem: Previous studies on Australian data provide inconclusive results on expectation hypothesis. Moreover, it is argued that Australian long-term rate is influenced by US long term rate, which signals decoupling of short rate from long rate. This decoupling, if true, makes monetary policy transmission ineffective. Accordingly, the paper examines expectation hypothesis with Australian data. Research methodology and data: Daily, weekly, monthly and quarterly data on 30-day, 90-day and 180-day dealers bill and 5-year and 10-year Australian government bond rate over the period from 1976 to 2016 are sourced from Datastream. Unit root test with structural break is applied to check the stationarity of data. Threshold co-integration approach is followed to test the expectation hypothesis of interest rate. Principal component analysis is used to identify common global factors affecting long rates in developed countries. Finally, Gewek’s instantaneous feedback method is applied to examine contemporaneous movement of long rate in developed countries. Research findings: Empirical analysis shows that all interest rate series are nonstationary at level; however, stationary at their first differences, which are suitable for co-integration analysis. Threshold co-integration results show that short rate and long rate in Australia are co-integrated, that is, expectation hypothesis holds. We also find evidence of threshold effect in some cases. Using principal component and instantaneous feedback method we find that long term bond yields in developed markets are driven by some unobserved global factors and the USA long rate has no causal influence on other developed markets’ long rate. We conclude that in Australia short rate is not de-coupling from long rate and hence effectiveness of monetary policy is well maintained. Policy implications: Expectation hypothesis is essential for the effective transmission of monetary policy. The hypothesis, if does not hold, causes interruption in monetary policy transmission. We find that expectation hypothesis holds for Australian data; hence RBA’s monetary policy transmission should not foresee any interruption. Moreover, bond contains premium for expected inflation, so it can be used as an indicator of RBA’s commitment for low inflation. Besides, financial intermediaries can effectively use the link between short and long rate in setting their interest on long-term loans based on short term interest rates.