Converting Annual to Monthly data

[zaki091]

Hi, I am researching the effects of Euroarea fiscal stimulus on inflation expectations during the crisis. The data for fiscal stimulus is given as an annual percentage of GDP, e.g. 1.5% of Euro Area GDP in 2009.

I am not sure if I could convert it to monthly data (by simply assuming 1.5% in every month for 2009) to show the effects of this stimulus on inflation expectations during this particular period.

P.S I need monthly data as my data for inflation expectations is monthly.

Can someone help with this please? Thanks

[trubador]

No, that would not be a wise approach. Since you are interested in "expectations", I do not think you need the actual size of the package or the exact date of implementation. Instead, you can simply take into account the before-and-after impact of the anouncement itself. If I remember correctly, the anouncement was made at the end-of November in 2008. So, you can add the following dummy variable into your model as an additional regressor:

Code: Select all

@after("2008m12")

[zaki091]

My model is as follows;

∆post π = α + β (Balsheetchg) + λ (FiscStimulus) + εi
∆post g = α + β (Balsheetchg) + λ(FiscStimulusi) + εi

where π = inflation expectations, g= output expectations, (Balsheetchg) = change in size of ECB balance sheet and (FiscStimulus)= is a measure of the Euroarea stimulus package (actual)

Are you suggesting that I drop the (Fiscstimulus) variable and add the dummy instead? Thanks

[trubador]

Yes, I think this is a better place to start to see if you can observe and measure a structural change. By the way, you may also want to add (Balsheetchg*Dummy) to check if slope changes.

[zaki091]

Great, thanks for your advice