Topic forecast future share price levels

Topic forecast future share price levels
Name:
University:
Course:
Tutor:
Date of Submission:

The daily closing prices data for FTSE and DowJones values for American and British American is obtained from yahoo finance. The logarithm of the relative prices over the period from 2001 to 2010 is multiplied by 100 in the case of FTSE to calculate continuously compounded daily stock returns using GARCH mathematical model. Logarithm prices have been used to prevent nonstationarity on the level of stock prices as well as stock returns volatility
To effectively use FTSE and DJ respectively to predict the future share prices individually and together to predict future American and British share price levels. It will be important to examine the distributional properties of the daily stock price index change individually for FTSE and DJ. The descriptive statistics which will be calculated will include mean , variance, standard deviation, skewness, kurtosis and Jarque-Bera statistics for normality test.
Prediction of share prices using FTSE values (2001-2010
The average change in price index over the ten years period is 0.15
The daily share price fluctuations is assumed to vary within the range of 8.51 to -7.74
The null hypothesis of normality is rejected at 1% level through the Jarque-Bera statistics. Kurtosis value is assumed to be zero which actually corresponds to zero
GARCH MODEL
GARCH model developed by Boller is one of the most common nonlinear time series models used to predict future share prices. The model actually involves the joint estimation of conditional mean and a conditional variance equation as shown in the equations below (Andersen, 2001).
GARCH MODEL
σ = ω+αε +βσ t1
definition of the parameters in the equation ,
tr denotes the continuously compounded daily share prices on time t, and t 1
Ω denotes the information set of all observed returns up to time t 1
μ and a are constant parameters, t
ε is the innovation process,
α, β are nonnegative parameters
Using this model the following parameters can be tabulated as follows;
This table presents parameters estimates for the GARCH-in-Mean for the share prices basing on a 10 year period 2001- 2010
Parameters Coefficient p-value
β 0.033 0.045
δ 0.164 0.000
α 0.140 0.000 μ 0...