Statistical analysis of wavelets pre-processed financial time series
by
Enrico Capobianco
Technical University of Denmark, Institute of Mathematical Modelling

We study high frequency Nikkei stock index series, sampled at daily and intradaily time intervals. We investigate what certain wavelet transforms suggest in terms of volatility features underlying the observed returns process. One of our scopes is to use wavelets as a pre-processing tool so to de-noise the data; we believe that this procedure may help in identifying, estimating and predicting volatility. We also aim to investigate some well-known temporal aggregation results, already described in other work on financial time series. We thus conduct our study with an interest toward both a statistical inference perspective and an empirical finance viewpoint. We definitely put emphasis on the importance of working with high frequency data for exploring many of those features which are usually claimed to characterize volatility models, but we also report evidence on how new nonparametric statistical procedures such as wavelets become useful for uncovering features in the data generating process which are not clearly seen from the noisy observed returns.

Date received: February 23, 1999

Copyright © 1999 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cacq-16.