eduzhai > Applied Sciences > Computer Science >

A Higher-Order Correct Fast Moving-Average Bootstrap for Dependent Data

  • Peter
  • (0) Download
  • 20210216
  • Save

... pages left unread,continue reading

Document pages: 42 pages

Abstract: We develop and implement a novel fast bootstrap for dependent data. Ourscheme is based on the i.i.d. resampling of the smoothed moment indicators. Wecharacterize the class of parametric and semi-parametric estimation problemsfor which the method is valid. We show the asymptotic refinements of theproposed procedure, proving that it is higher-order correct under mildassumptions on the time series, the estimating functions, and the smoothingkernel. We illustrate the applicability and the advantages of our procedure forGeneralized Empirical Likelihood estimation. As a by-product, our fastbootstrap provides higher-order correct asymptotic confidence distributions.Monte Carlo simulations on an autoregressive conditional duration model providenumerical evidence that the novel bootstrap yields higher-order accurateconfidence intervals. A real-data application on dynamics of trading volume ofstocks illustrates the advantage of our method over the routinely-appliedfirst-order asymptotic theory, when the underlying distribution of the teststatistic is skewed or fat-tailed.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×
save success