MIDAS

[EV_an]

Hello,

I am using the new tool of EViews 9.5 to estimate MIDAS models.

I am wondering if exists a similar tool to estimate Unrestricted MIDAS model (U-MIDAS).
Further, I would like to understand better how I can forecast using the MIDAS function. For instance, I want to forecast (static forecast) a quarterly variable using monthly data. When forecasting, is it possible to update the estimated function for each month of the quarter (i.e. I use the same model every time I am in the first month of the quarter). Finally, can I get dynamic forecasts re-estimating the model for each h horizon? Do I have to do it with a loop?

Thanks,

A.

[EViews Gareth]

U-MIDAS is not built in. It wasn't added since it is fairly simple to estimate outside of the MIDAS engine. We've been contemplating adding it nevertheless though.

To do it manually, simply perform point frequency conversion to bring over the observations, and lags of the observations to the lower frequency page.


I'm not sure I understand what you mean by a different model for each month of the quarter, but, yes, using the manual frequency conversion method, it should be possible.

To perform h-step forecasting, you will have to write a for loop, yes.



Since it might be useful, here's a program I wrote to verify that the PDL/Almon MIDAS approach was estimating correctly - it brings over the lags and then creates a polynomial from them. It should give you some pointers on writing a manual U-MIDAS.

Code: Select all

!xlags = 4 !ylags=1 !horizon=2 !polyorder = 3 wfopen(wf=midas, page=quarterly) .\mydata.xlsx range="Sheet1" pageload(page=monthly) .\mydata.xlsx range="Sheet2" series x1 = dlog(value(-!horizon))*100 for !i=2 to !xlags series x!i = x1(-!i+1) next for !i=1 to !xlags series z!i=x!i(-2) next pageselect quarterly series y = dlog(value)*100 smpl 1985 2009/1/1 for !i=1 to !xlags copy(c=l) monthly\z!i x!i next for !i=1 to !polyorder series z!i=0 scalar w!i=0 next for !i=1 to !xlags for !j=1 to !polyorder z!j = z!j + !i^(!j-1)*x!i w!j = w!j + !i^(!j-1) next next group xlags z* group ylags for !i=1 to !ylags ylags.add y(-!i) next equation eq2.ls y c xlags ylags show eq2

[EV_an]

Thanks! That was useful, I estimated the u-midas model and got the forecast for both models.