An algorithm using GARCH process , Monte-Carlo simulation and wavelets analysis for stock prediction

Electronic copy available at: http://ssrn. com/abstract=1271248 1 An algorithm using GARCH process , Monte-Carlo simulation and wavelets analysis for stock prediction Eleftherios Giovanis Abstract This paper examines and presents a simple algorithm... More

Electronic copy available at: http://ssrn.
com/abstract=1271248
1
An algorithm using GARCH process , Monte-Carlo simulation and wavelets
analysis for stock prediction
Eleftherios Giovanis
Abstract
This paper examines and presents a simple algorithm for prediction stock written in
MATLAB code.
We apply it to thirty stocks of the Athens exchange stock market .

We obtain the stock returns and we would like to predict, not the actual price , but the
sign of stock returns.
The results are very satisfying while we predict the right sign
for 25 out of 30 cases or else we have a success of 83.
33%.
The problem with the
algorithm is that we don’t have the ability to predict zero returns.

Keywords GARCH, wavelets˙ forecasting˙ Monte-Carlo˙ wavelet discrete
transformation
Introduction
We use GARCH model because financial time series are characterized by
leptokurtosis , clustering volatility and leverage effects.
GARCH (1,1) as a symmetry
model can capture with succe Less