Plotting event study by month

[Charlie Groscup]

Hi,

I am trying to run and plot an event study, but my regression is not reporting any statistics other than the coefficients and I am unsure why. I've posted the code below and have attached an image of the regression result:

format date %td
gen metal_type=0
replace metal_type=1 if steel_prices
label var metal_type "Type of metal"
reg metal_price metal_type metal_type#i.date i.date

Also in trying to plot the regression, the labels read " Steel#12815", but I want the labels to read dates such as "01feb1995" and I am unsure how to specify this in the coefplot command.

Thanks!

[Jared Greathouse]

Somethings wrong, deeply wrong, with your data. If you're only getting point estimates and no SEs, Something isn't right with your dataset.

[Charlie Groscup]

I would think so too, except that when I run the following regression it seems to work:

reg metal_price metal_type metal_type#tariff i.date

[Jared Greathouse]

Please, don't include screenshots of your output. You can simply copy and paste in in the code delimiters. Anyways, why do one way interactions? You can jut do

reg metal_price i.metal_type##i.date

Either way, and maybe I'm misunderstanding, this is not what I'd call an event study. In this instance, you'd simply place a set of lags and leads on your intervention of interest to check for anticipation effects.

[Charlie Groscup]

I went ahead and figured it out, thanks for the help though!