"MODELLING AND FORECASTING FINANCIAL LOG-RETURNS AS LOCALLY STATIONARY WAVELET PROCESSES"

This page accompanies the article Modelling and forecasting financial log-returns as locally stationary wavelet processes by Piotr Fryzlewicz. The reader can download here the routines written for the paper, and the data sets analysed in it. See below for the full list of objects available for download.

All the functions have been written in S and designed for S-Plus (version "2000 Windows"), which is a commercial package. However, they will also probably work in R, which can be downloaded free of charge here.

Additionally, some functions require WaveThresh3, a free add-on module for S-Plus written by Guy P. Nason.

Downloading and installing the functions and the data sets.

  1. Click here to download the archive "fints.zip".
  2. Extract the archive. You will obtain the file "fints.q".
  3. Run S-Plus and type (in Windows) "data.restore("D:\\PATH\\fints.q")", where D:\PATH is the directory containing "fints.q".
  4. The functions and the data sets are now available.

Alphabetic list of functions --- click on the name for help. Click here for the forecasting routines. We recommend that the "start.p" parameter in function adjust be set to 1 when forecasting financial log-returns.

Alphabetic list of data sets --- click on the name for description

Alphabetic list of time-varying local variance functions --- click on the name for description.