External macroeconomic factors and the link between short- and long-run European interest rates: a note.

• Author: Camarero, Mariam

1. Introduction

   The most recent literature on the term structure of interest rates suggests that the seminal model of Campbell and Shiller (1987), consisting of the spread and the change in the short-term rate in a vector autoregressive (VAR) specification, could be too parsimonious to fit the data (Carriero, Favero, and Kaminska 2006). In fact, this bivariate approach includes an implicit reaction function, where the only determinants of policy rates are long-term rates. Therefore, there are potential misspecifications due to the omission of macroeconomic factors. To solve this problem, some authors have considered the inclusion of other macroeconomic variables, "a la Taylor," to which monetary policymakers react. The usual potential omitted variables involved in the augmented models are internal ones, namely, inflation, output, or employment movements. However, the external determinants have been only scarcely studied. (1)

   In this article, we propose the inclusion of a new set of variables ("a la McCallum") compatible with the endogenous policy reaction model developed by McCallum (2005), Assuming that policymakers adjust interest rates in order to keep exchange rates stable and that they are interested in smoothing interest rate movements, McCallum develops a reduced-form equation for the spot exchange rate under rational expectations. From McCallum (2005), if the European Central Bank (ECB) smoothes the path for short-run interest rates, and if it weighs in the exchange rates when formulating policy, then simple tests of the expectations hypothesis (EH) could be severely biased. Therefore, by taking into account the external influences of foreign economy, the fit of the EH may be improved. In other words, expectations of the short rate may explain a greater share of long-run rate variability than simple tests may indicate. Kluger (2000) generalizes this result by taking into account the policy reaction to the term spread in addition to the exchange rate. According to this model, the central bank increases the short-run rate in response to a depreciating exchange rate or a widening spread, which signals higher expected future inflation.

   According to Neely (2001), there are at least three reasons central banks might tend to change their interest rate targets in a similar fashion. First, countries react similarly to common shocks. Central banks take into account the state of the economy, including international conditions, such as commodity prices (i.e., oil), to implement monetary policy. Therefore, if changes in such prices tend to affect countries in the same way, they lead to similar monetary policy measures. Second, in an integrated economy, countries may desire to maintain stable exchange rates, and, consequently, central banks might minimize swings in the external value of their currencies, leading to coordinated policy measures. Third, economic conditions in one country affect those in other countries through trade and capital flows. A U.S. recession that leads to lower U.S. interest rates might also slow down its trading partners' growth, prompting their central banks to lower rates as well.

   It has been claimed that with the formation of the European Monetary Union (EMU), monetary policy in Europe may be more concerned with European trends and less concerned with external factors than the central banks of the constituent states used to be. Although this statement can be true and monetary policy is not, certainly, a game of "follow the leader," all in all we think that the ECB considers the effect of external factors, including foreign interest rates, when making monetary policy for its own countries (Belke and Gross 2005; Bruggemann and Lfitkepohl 2005).

   Therefore, in this article, we contribute to previous empirical literature by augmenting the original VAR framework proposed by Campbell and Shiller (1987) with additional variables. Specifically, to control for worldwide effects on Euro-area interest rates, the variables we consider are the U.S. short-run interest rates, U.S. inflation, and the euro/dollar exchange rate. Once foreign interest rates and the exchange rate are included in the VAR, not only the EH of the term structure but also the EH in the foreign exchange market plays a role.

   According to our results, the inflation rate decreases not only because of the evolution of the short-run interest rates controlled by the ECB but also as a joint consequence of a higher degree of international competitiveness (and economic integration). Thus, long-run interest rates react to exchange rates, and this reaction might be precisely because of the anticipated future reactions of the ECB to the exchange rates. In turn, this interpretation supports the EH.

2. Empirical Results

   We analyze the following quarterly data:

   [x'.sub.t] = [[l.sub.t], [i.sub.t], [q.sub.t], [DELTA][p.sub.t], [[i.sub.t].sup.*], [DELTA][p.sub.t.sup.*] t = 1986 : 1 to 2005 : 2,

   where [l.sub.t], [i.sub.t], and [p.sub.t] are, respectively, the aggregated long- and short-run...