RDP 8906: A Random Walk Around the $A: Expectations, Risk, Interest Rates and Consequences for External Imbalance I. Introduction

Australia is a small open economy operating with essentially no impediments to the movement of capital into or out of the country. As a consequence, Australian interest rates should satisfy international arbitrage conditions. The arbitrage condition for a representative US investor can be expressed either in terms of nominal or real interest rates. Thus,

where i and r denote the nominal and real interest rates for some asset, S is the nominal $US/$A exchange rate, ΔS is the change in S over the life of the asset, and the superscript R denotes real. The risk premia, rp and rpR, are the excess returns demanded by a US investor to hold the Australian denominated asset.[1]

This paper presents a detailed examination of these two equations. Our almost exclusive focus is on short-term nominal assets with the horizon of our analysis ranging from one week to three months. The paper is laid out as follows. Section II presents evidence on the forward rate and on survey market expectations as predictors of the future spot exchange rate. In section III, after drawing implications from the survey on the size of the risk premium, two theoretical models are presented of the risk premium necessary to induce a representative US consumer-investor to hold a small proportion of assets in short-term nominal Australian assets. Section IV discusses the behaviour over the past fifteen years of consumer price inflation and short-term nominal and real interest rates for several OECD economies. The fifth section demonstrates that since Dec 83, the $A, unlike all the other currencies we examine, has exhibited significant skewness because of many large rapid unpredictable depreciations. The final section discusses the results of the paper in terms of either (i) a time-varying risk premium, or (ii) an inefficient foreign exchange market, or (iii) a peso problem (see definition on page 39) for the $A. Since late 85, all our evidence is that the risk premium has been much too small to explain the short-term real interest differential between Australia and the US, and so we focus on the latter two explanations. Evidence on the inefficiency of the foreign exchange market can provide a rationalization of the results – but a puzzle remains. As an alternative, we provide evidence that the $A suffers from a peso problem because of a market perception that, in the longer run, the real economy must adjust to put Australia on a sustainable net external debt/GDP path – with a lower real and nominal exchange rate during the adjustment process.


The term ‘risk premium’ is often used loosely to mean the excess return demanded by investors to compensate them for the ‘risk’ of an exchange rate fall. In this paper, we use the term in its technical sense. For a given expectation of the return on an asset, the risk premium is the excess return required because of the expected distribution of that return which may be summarised by the expected higher moments of the return on the asset. Equations (1) and (2) are approximations because, for example, (1 + iA)/(1 + ius) is only approximately equal to 1 + iA − ius. In the paper, the exact expressions are used when required. [1]