Out of sample forecast
Posted: Fri Feb 21, 2014 2:31 am
Hello everyone,
first of all: Sorry for my bad english, but I am not a native speaker.
I am currently working on my master thesis, but I am quite new to Eviews, hence lacking lots of programming skills.
I want to conduct an out of sample analysis with expanding window, checking the predictability of risk. My idea is I want to run an AR(1) and then use the estimated coefficients and last obs´of realized variance to forecast the realized variance in the following month.(see here if you need more information (page
: http://papers.ssrn.com/sol3/papers.cfm? ... id=2041429)
I have attached my program (which I took from Esther, who posted it somewhere here and adjusted it to my needs). My point is: I´m not sure if the program does what it should. Further, I don´t understand, what the program really stores in 'fcast_wml'. I was thinking that it stores the predicted values for realized variance based on the estimated coefficients. But if I substitute the estimated coefficients for Alpha and Rho and the observed realized variance from the previous month into my equation, it does not give the forecast values. The problem is, I am not sure if it really had to give the forecasted values (if so, something in my code seems to be false) or if I am completely on the wrong track (and my code might be correct) and neglect something important.
I would really appreciate if someone here could help with this problem.
Greetings, Sasa
first of all: Sorry for my bad english, but I am not a native speaker.
I am currently working on my master thesis, but I am quite new to Eviews, hence lacking lots of programming skills.
I want to conduct an out of sample analysis with expanding window, checking the predictability of risk. My idea is I want to run an AR(1) and then use the estimated coefficients and last obs´of realized variance to forecast the realized variance in the following month.(see here if you need more information (page
: http://papers.ssrn.com/sol3/papers.cfm? ... id=2041429)I have attached my program (which I took from Esther, who posted it somewhere here and adjusted it to my needs). My point is: I´m not sure if the program does what it should. Further, I don´t understand, what the program really stores in 'fcast_wml'. I was thinking that it stores the predicted values for realized variance based on the estimated coefficients. But if I substitute the estimated coefficients for Alpha and Rho and the observed realized variance from the previous month into my equation, it does not give the forecast values. The problem is, I am not sure if it really had to give the forecasted values (if so, something in my code seems to be false) or if I am completely on the wrong track (and my code might be correct) and neglect something important.
I would really appreciate if someone here could help with this problem.
Greetings, Sasa