Of Volatility Spillovers In European Government Bond Markets

In this paper we engineer an information mapping of transmission linkages across various European government bond markets. The research introduces a calibration methodology for the application of an optimizing radial basis function (RBF) artificial neural network (ANN). Utilizing a closed-form derivation of the regularization parameter, the Kajiji-4 RBF ANN is known to efficiently minimize the ill-effects of multicollinearity. This innovation benefits the ANN-based financial economeaics in a comparative re-examination of volatility spillover effects from the US and aggregate European government bond markets into the bond markets of individual European countries. The research offers a two-fold extension to computational analytics in this field. First, the analysis presents a philosophy for the artfbl engineering of the time-series directed RBF ANN. Second, a two-step procedure is introduced that relies upon both the ARCH-framework and a classification directed artificial network to test overall modeling efficiency. We find that all post-modeling efficiency tests certify the ex-ante expectation for the Kajiji-4 RBF ANN to produce residuals that are devoid of latent economic covariance and conditional volatility effects. Moreover, we find the efficiently estimated parameters yield evidence to confirm a substantial body of extant literature on bond volatility-spillover effects. But, these same parameter estimates also provide strong evidence to contradict the contemporary view of a weak US volatility-spillover for EMU countries with a correspondingly strong spillover effect for non-EMU countries. 4" Printing. Not for public quotation