Browsing Statistics by series "Economics and Society – 262"

Sort by: Order: Results:

Now showing items 1-1 of 1
  • Catani, Paul (Svenska handelshögskolan, 2013)
    Economics and Society – 262
    Conditional heteroskedasticity is often encountered in economic and financial time series. Since the introduction of autoregressive conditional heteroskedasticity (ARCH) by Engle in 1982, modelling volatility has received much attention in financial econometrics. Conditional heteroskedasticity also causes many asymptotic tests in time series models not to be valid. For example, tests for autocorrelation typically assume independent and identically distributed errors. The wild bootstrap provides a solution to the problem with inference under conditional heteroskedasticity. This thesis consists of an introduction and four papers dealing with conditional heteroskedasticity in multivariate time series models. The first paper studies wild bootstrap tests for autocorrelation in vector autoregressive (VAR) models with conditional heteroskedasticity. The second paper is an empirical study of tests for cointegration in Chinese stock price data in the presence of conditional heteroskedasticity. The third paper proposes and studies a new Lagrange multiplier test for testing the adequacy of an estimated constant conditional correlation generalized ARCH model. The fourth paper studies tests for ARCH in VAR models.