Analysis of the evolution of sovereign bond yields by wavelet techniques

Abstract

The term “wavelets” covers a set of resources from the mathematical analysis that has proven their efficiency in system identification on areas such as hydrology, geology, glaciology, climatology and energy resources optimization. The methodology undergone on systems engineering could be extrapolated to everything conceptualized as “complex system” whatever its nature be. The wavelet techniques provide the description of non-stationary components and the evolution of macroeconomic variables in the frequency domain. The identification of predominant frequential scales and transient effects in time series highlights the multiresolutional analysis that would be more difficult to treat with traditional methods of econometrics. A review of the literature will show the potential problems that can be solved with these techniques, including prediction of benefits calculated on the evolution of the risk premium of a country, the extraction of symmetric macroeconomic shocks in country clusters, or detection of transient effects on the mutual influence of sovereign bonds between pairs of countries, among others. The dissertation will culminate in specific applications that show the power of wavelet techniques in identifying possible determinants and correlation of the evolution of sovereign bond yields in the euro area countries.