Parity conditions, currency forward market efficiency and risk premium : Australian evidence

The objective of this thesis is to undertake an empirical investigation of three wellknown exchange rate relationships, namely, covered interest parity (CIP), uncovered interest parity (UIP) and forward market efficiency (FME), using daily data as it applied to the Australian and US economies over the period 2 December 1985-29 December 2000. Due to the issue of non-stationary of the regression variables, ordinary least squares (OLS) results are corrected according to West (1988)'s procedures and two new estimation techniques ‚ÄövÑvÆ fully modified ordinary least squares (FM-OLS) developed by Phillips and Hansen (1990) and fully modified least absolute deviation (FM-LAD) by Phillips (1995) ‚ÄövÑvÆ are introduced and utilised. Modified versions of the UIP and FME models are also developed to incorporate a time-varying risk premium to rationalise the rejections of the UIP and FME in their original forms. Estimation of these modified UIP and FME models is conducted in the context of three types of generalised-auto-regressive-conditional-heteroskedasticityin-mean (GARCH-M) processes. An attempt is also made to investigate the effects of monetary policy actions not necessarily targeted at the exchange rate on the conditional volatility of the spot exchange rate, that is, the time-varying risk premium by including the monetary policy target (the overnight cash rate) in these models. While both the full sample (2 December 1985 ‚ÄövÑvÆ 29 December 2000) and the subsample (2 December 1985 ‚ÄövÑvÆ 31 December 1991 and 2 January 1992 ‚ÄövÑvÆ 29 December 2000) analysis reveal evidence supporting OP over the period December 1985 ‚ÄövÑvÆ December 2000, the UIP and FME conditions are only found to hold in the 90-day market over the period 2 January 1992 ‚ÄövÑvÆ 29 December 2000. The results of the subsample analysis reveal no conclusive evidence to support the notion of a time-varying risk premium over the period 2 December 1985 ‚ÄövÑvÆ 31 December 1991 leading to rejections of both the HIP and FME models in the 90-day market. However, when the cash rate is included, the presence of a time-varying risk premium is supported in that market, suggesting that the Reserve Bank of Australia (RBA)'s monetary policy plays an important role in influencing the volatility of exchange rates.