Asymmetry in the EMS: New evidence based on non-linear forecasts

Abstract

In this paper we provide new evidence on the hypothesis of German leadership and asymmetric performance in the EMS, in the framework of causality tests, using daily data. Given the evidence about non-linearity in financial series, we propose applying non-linear forecasting methods based on the literature on complex dynamic systems. Our analysis covers nine countries, and the sample period runs until 30 April 1998, so including the more recent events in the EMS history. A comparison of our results with those obtained from standard linear econometric techniques leads us to conclude that inference on causality based on our non-linear predictors would be preferable to that based on the standard linear approach.

Portfolio diversification benefits within Europe: Implications for a US investor

“…Nonetheless, in this paper's case, the cointegrability property of these returns has decreased in the European Monetary System, perhaps due to changes within the exchange rate mechanism (like the widening of the fluctuation bands from F2.25% in the 1980s to F15% in the early 1990s), or to the gradual process of convergence of the interest rates of individual EMS countries (with that of Germany's), a process that makes the cointegrating relationship variant over time (see Caporale & Pittis, 1993;Hassapis, Pittis, & Prodromides, 1999). Finally, other authors (Bajo-Rubio, Sosvilla-Rivero, & Fernandez-Rodriguez, 2001) found some evidence of cointegration among some EMS members but not with others.…”

Monetary policy implications of comovements among long-term interest rates

“…Therefore, the absence of cointegration may merely reflect the gradual process of convergence of the interest rates of individual EMS countries (with that of Germany's), a process that makes the cointegrating relationship variant over time (see Caporale and Pittis, 1993;Hassapis et al, 1999). Finally, other authors (Bajo-Rubio et al, 2001) found some evidence of cointegration among some EMS members but not with others. 1992; Hamilton, 1994).…”

Forecasting interest rate swap spreads using domestic and international risk factors: evidence from linear and non‐linear models

“…Gençay (1999) found that NN models outperform other linear and non-linear models for a number of exchange rates, whereas Bajo-Rubio et al (2001) found evidence that NN models outperform linear models in predicting European interest rates. On the other hand, Teräsvirta et al (2005) used a number of G7 macroeconomic time series to show that STAR models forecast better when compared to linear models.…”

“…In contrast to STVAR models, MRS models assume that an unobserved Markov state variable drives the switching between regimes. 4 Recent applications of NN models in finance include, for example, Diebold and Nason (1990), Meese and Rose (1990), Gençay (1999), Jaditz and Riddick (2000) and Bajo-Rubio et al (2001). 5 Bootstrapping is a particular methodology for extracting zero-coupon rates from interest rate swaps that trade at par.…”