Page 1 of 1

How to use Automatic ARIMA to forecast stock prices

Posted: Thu Jan 24, 2019 11:33 am
by ronctl
This is my first time ever to use EVIEWS to forecast stock price. As I am still a beginner and I hit a lot of problems.

First I download a set of daily data of IBM stock price from 24/01/2009 to 24/01/2019, then using automatic ARIMA forecast with the following setting:
ARIMA specification: Max diff. is 2, Max AR and Max MA are both 4, SAR and SMA are both zero and periodicity is 10
Transformation is auto
Regressors is c

After I launched the application, I have the following results:

------------------------------------------------------------------------
Automatic ARIMA Forecasting
Selected dependent variable: DLOG(CLOSE, 2)
Date: 01/24/19 Time: 18:30
Sample: 1/20/2009 1/17/2019
Included observations: 2515
Forecast length: 0

Number of estimated ARMA models: 25
Number of non-converged estimations: 0
Selected ARMA model: (2,2)(0,0)
AIC value: -5.85479996399

------------------------------------------------------------------------
First of all, what does Selected ARMA model: (2,2)(0,0) mean here? How I check the coefficients of the regression? Second, should I set periodicity equal to zero?
Third, what is the right procedure to forecast stock prices?

Re: How to use Automatic ARIMA to forecast stock prices

Posted: Thu Jan 24, 2019 4:24 pm
by startz
Third, what is the right procedure to forecast stock prices?
The right way to forecast the stock price P_(t+1) is almost certainly P_t, or perhaps P_t times a constant very, very slightly above 1.0.