Event Study Analysis for Panel Data

[Arbnor Gashi]

I want to run a regression of the variable on a constant and a set of 7 dummies, where the dummies are t_3, t_2, t_1 representing the time before the crisis; t representing the crisis; and t1, t2, t3 representing the time after the crisis. I have two problems:

1) I have run a regression using mvreg t_1 t_2 t_3 t t1 t2 t3 = REINTR, but I am not sure if this the right way to do it.
2) I also want to plot using the twoway plot of the coefficients on the constant and the beta of the dummies and standard deviation, and using the dummies t_3, t_2, t_1, t, t1, t2, t3 as the time variable.

[Arbnor Gashi]

I made mistake in this question so I am gonna ask again. I have panel data of 69 countries starting from 1980 to 2010. As mentioned in my last post I have already created 7 dummy variables corresponding to the crisis year (t), pre-crisis years ( t_3, t_2, t_1 ), and post-crisis years (t1, t2, t3).

1) In order to run a regression, I also need to create an additional variable which I call constant. This variable is the average of TotReserves when there is no crisis and no pre- or post-crisis dummy (tranquil time). For these variable I have used the following command: bys country: egen constant= mean(TotReserves) if tranquil==1, however, I am not sure if this is the right way to do.

2) After that, I need to run a simple regression of variable TotReserves on constant and seven dummies.

3) Then I have to plot this as follow:
a) I have to plot beta of constant (tranquil time) and,
b) Betas of the dummies including 95% confidence interval.
P.S. X-axis has to be t_3, t_2, t_1, t, t1, t2, t3.

For more information the paper below explains it in details:
Ciarlone, A., & Trebeschi, G. (2005). Designing an early warning system for debt crises. p.378-379:

[Phil Bromiley]

You didn't get a quick answer. You'll increase your chances of a useful answer by following the FAQ on asking questions - provide Stata code in code delimiters, Stata output, and sample data.

We normally do not help with homework. I'd never name a variable constant - constant is the intercept in the model.