German leadership: evidence from retail mortgage rates

Abstract

The article finds evidence that the converging trend of euro-zone countries' retail mortgage rates, observed before and shortly after the introduction of the euro, is dictated by the German mortgage rate, hence supporting the German leadership role.

Notes

¹ The beginning of the sample period coincides with the implementation of the Second Banking Directive for a single banking market across the EU. In order to find out whether or not the inclusion of some post-1999:1 period makes any differences in the article's analysis, additional 18 observations (1999:1 to 2000:6) are added to the sample. This article investigates 11 founder members of EMU: however, Luxembourg, Austria, Italy and France are omitted because, respectively, no retail mortgage rate data, only partial data (from 1995:4 on), only partial data (from 1995:1 on), and only quarterly data, are available.

Fig. 1. The retail mortgage rates (in %). The vertical line indicates 1999:1Source: European Central Bank. Series: lending rates, N2, mortgage loans to households.

² The results of the investigation are mixed. See Hafer and Kutan (1994) for a survey.

³ The lag length was selected among lags from 1 to 12 based on the Akaike information criterion (AIC) and, separately, the Schwarz information criterion (SC). For the Ireland case, the t-statistic is −5.93. The equation includes a constant and one lag (chosen based on AIC and SC).

⁴ AIC chooses the lag length two, while SC and HQ choose the lag length one among lags from 1 to 12. Reimers (1993) suggests that HQ is most reliable when VARs include cointegrating relations. Lütkepohl (1993) explains these three criteria.

⁵ The interpretation of complete convergence follows Hafer and Kutan (1994) and Haug et al. (2000): complete convergence is obtained if there are p − 1 cointegrating vectors (or equivalently there is a single shared common stochastic trend) among p mortgage rates.