Quant - more autoregressive stuff

I know this is alot to ask, but there is no way I can get the formatting for this question to work. Schweser q-bank #12798. Can you guys take a look at this and explain the last question to me? I understand if it is too much effort.

To forecast Q2 2004, you need the t-1 and t-4 lagged terms. You are given the t-4 value (Q2 2003), and you need to forecast the t-1 value (Q1 2004). You forecast the Q1 2004 value using the appropriate lagged terms (Q4 2003 and Q1 2003). Q1 2004 = 0.155 + 0.24(0.24) + 0.168(0.26) = 0.256 Q2 2004 = 0.155 + 0.24(0.256) + 0.168(0.22) = 0.253 === Answer

Regression Equation is GM = 0.155 + 0.240*X1 + 0.168*X4 GM(of 2004.2) = TO CALCULATE GM(2004.1) = 0.155 + 0.240*0.240 + 0.168*0.260 =0.155 + 0.0576 + 0.04368 = 0.25628 Use chain forcasting… GM(2004.2) = 0.155 + 0.240*0.25628 + 0.168*0.220 0.155 + 0.0615072 + 0.03696 = 0.2534672 Hence GM(of 2004.2) = 0.2534672

You mean part 6??? I just took this question a few hours ago (reviewing every section in order right now). First, just write out the equation so you have it. Take the info from the first table. I’ll do it in increments for you: Y = 0.155… (this is the “constant” coefficient or the intercept). Y= 0.155 + .24 (Y lag 1) … (add .24 of whatever Y was last period) Y= 0.155 + .24 (Y lag 1) + .168 (Y lag 4) … (add .168 times Y four periods ago). So we are given the value of Y up to the 4th quarter of 2003 and we are asked for Y (forecast) for the second quarter of 2004. The first thing we need to do is calculate Y for the FIRST quarter of 2004. Plug and chug into the above equation. Once you have that result you can plug and chug the equation for the second quarter of 2004. Is it still not clear what they’re doing here?

Thanks guys.

plyon Wrote: ------------------------------------------------------ > > Is it still not clear what they’re doing here? Not much is clear right now. :frowning: There is no way I can learn this stuff in 27 days. I don’t even know what to do next.

just using the time series regression to forecast the future

mwvt9 same here… i went through quant understood a big chunk but i still need to go over it but I cant waste too much time now so maybe later… go to keep on going…

Okay, I got it. There explanation confused me a little. Just chain forecasting.

I just realized I was looking at the table of years backwards. I was using the oldest data as the latest. Stupid mistakes.