Code for GARCH-ARIMA

I have time series data,

My input series is TEMP., HUMIDITY, PRESSURE.

and MY TARGET SERIES IS AVERAGE WIND SPEED.

I am arranging my data for 1 day ahead prediction like

d= ones(1,365);

X_new=[Temp.;Humidity;Pressure];

X=mat2cell(X_new,3,24*d);

input series = X

T_new=[wind speed];

T=mat2cell(T_NEW,1,24*d);

Target series = T

I WANT TO GO NEXT STEP FOR garch-arima, I WANT CODE TO CALCULATE P,D,Q VALUE

ALSO PREDICTION CODING, for GARCH-ARIMA

Answer 1

[Subhajyoti]

Hi Seemant,

You can perform the GARCH-ARIMA modelling in MATLAB, using the Econometric Toolbox. The ‘arima’ function in MATLAB helps to estimate the parameters of the ARIMA model. The residuals from the ARIMA model can then be used to fit a GARCH(1,1) model with ‘garch’ function. This allows for forecasting future values and volatility of the series.

Note: Since MATLAB does not support an automatic function, you may need to try different combinations of ‘p’, ‘d’, ‘q’, using criterias like AIC (Akaike Information Criterion), BIC (Bayesian Information Criterion), Autocorrelation functions like ‘autocorr’ and ‘parcorr’, etc.

You may go through the following MathWorks documentation link to learn more about GARCH-ARIMA modelling in MATLAB.

• ‘arima’: https://www.mathworks.com/help/econ/arima.html
• Estimate ARIMA models: https://www.mathworks.com/help/ident/ug/estimating-arima-models.html
• ‘garch’: https://www.mathworks.com/help/econ/garch.html
• Specify GARCH models: https://www.mathworks.com/help/econ/specify-garch-models-using-garch.html

I hope this helps.