Veiga, HelenaUniversidad Carlos III de Madrid. Departamento de Estadística2007-02-212007-02-212007-02https://hdl.handle.net/10016/625According to the Taylor-Effect the autocorrelations of absolute financial returns are higher than the ones of squared returns. In this work, we analyze this empirical property for three different asymmetric stochastic volatility models, with short and/or long memory. Specially, we investigate how the Taylor-Effect relates to the most important model characteristics: its asymmetry and its capacity to generate volatility persistence and kurtosis. Finally, we realize Monte Carlo experiments to infer about possible biases of the sample Taylor-Effect and fit the models to the return series of the Dow Jones.162123 bytesapplication/pdfspaAtribución-NoComercial-SinDerivadas 3.0 EspañaAsymmetryKurtosisLong and short memoryTaylor-Effectworking paperEstadísticaopen accessws070702