RDP 2008-03: Monetary Transmission and the Yield Curve in a Small Open Economy 1. Introduction

Long-term nominal interest rates in a number of inflation-targeting small open economies, such as Australia, Canada, New Zealand, Norway, Sweden and the United Kingdom, have moved very closely with those of the United States over the past 15 years or so. Figure 1 shows the pattern of interest rate correlations at different points on the yield curve for each country with the US. The pattern is clear: long-term nominal rates are highly correlated with their US counterparts, generally more so than rates at shorter maturities, and more so than with their respective short-term rates (not shown in Figure 1). This pattern has led to the view that long-term nominal interest rates are somehow determined abroad.^[1]

Traditionally, the transmission mechanism of monetary policy is understood (among other things) as linking a short-term nominal interest rate to a long-term real interest rate which, in turn, influences aggregate demand. Of course, for a short-term nominal rate to influence a long-term real rate, it must be that the short-term nominal rate influences long-term nominal rates.

In this paper we assess whether these reduced-form patterns of correlations are consistent with theory and examine the forces at work in the determination of a small open economy's yield curve. We find that the expectations hypothesis together with uncovered interest rate parity can account for much of the observed co-movement in interest rates of different currencies. In fact, the main contribution of our paper is to uncover a mechanism that can give rise to these observed reduced-form correlations in optimising general equilibrium. As we discuss below, differences in the persistence of domestic and foreign disturbances can bring about the pattern of correlations in Figure 1.

Other papers have tackled the related question of how much domestic and foreign factors influence domestic interest rates. For example, Campbell and Lewis (1998) use an event study to examine how Australian bond yields respond to new information and find that US economic news has a larger effect than domestic news on Australian yields. Tarditi (1996) estimates a reduced-form model of the Australian 10-year bond yield and finds that a one percentage point increase in the US 10-year bond yield is associated with around half a percentage point increase in Australian long-term yields.

There is also a large literature that analyses the yield curve with affine term structure models. These studies typically assume that bond yields are affine functions of unobservable factors and incorporate cross-equation restrictions that eliminate arbitrage opportunities (Knez, Litterman and Scheinkman 1994; Duffie and Kan 1996; Dai and Singleton 2000; Backus, Foresi and Telmer 2001; Backus and Wright 2007). But while factor models have been relatively successful in matching key statistical properties of the yield curve, factor models are not structural. Recent work addresses this issue by fitting the term structure to macroeconomic factors, either by combining them within unobserved factors, as in Ang and Piazzesi (2003) and Bernanke, Reinhart and Sack (2004), or by incorporating a no-arbitrage model of the term structure within a macroeconomic model as in Rudebusch and Wu (2004), Bekaert, Cho and Moreno (2005) and Hördahl, Tristani and Vestin (2006).

Our focus here is different. We set up a micro-founded two-block model consisting of a small open economy and a large (closed) economy and extend the set of equilibrium conditions in both the large and small economies to allow for an explicit consideration of the co-movement of foreign and domestic interest rates. In our model, the expectations hypothesis links interest rates of different maturities and uncovered interest rate parity links interest rates of different currencies. Short-term nominal rates are set by the monetary authorities on the basis of the fundamentals of their economies. In this respect our analysis resembles that of Evans and Marshall (1998), but unlike them, we study the behaviour of a small open economy's yield curve and pay particular attention to its relation to the large economy's yield curve. We then estimate the model's parameters and examine its ability to match the co-movement of interest rates of different currencies.

The rest of the paper is organised as follows. Section 2 describes the model and the role of the yield curve in the transmission mechanism of a small open economy. Section 3 discusses the estimation of the model. Section 4 examines the dynamics of the yield curve. Section 5 contrasts the model's moments with their empirical counterparts and Section 6 concludes.

Footnote

For the unfiltered data, the pattern of correlation is also increasing along the yield curve and quite similar. In the case of Australia, for example, these correlations are 0.56, 0.76 and 0.85 for 3-month, 5-year and 10-year maturities, respectively. As we discuss below, our conclusions are robust to the filtering of the data. [1]