Hi all,
I've performed a time-series regression loop by using the following code (approximately 250 regressions):
gen sic_2 = real(substr(sic,1,3))
egen SIC_id = group(sic_2)
drop if SIC_id==.
bysort SIC_id: gen no_obs=[_N]
keep if n_obs>50
forval i = 1(1)258{
if inlist(`i', 65,66,67,170,195,232,236,249,250,251) continue
reg WD_WC WCFO_m1 WCFO_0 WCFO_p1 if `i' == SIC_id, r
estimates store reg`i'
}
where WD_WC means change in working capital, and WCFO_m1, WCFO_0, and WCFO_p1 are past, current, and future cash flows, respectively (model of Dechow and Dichev of 2002).
This gives me all the coefficients of the 250 regressions. However, I want to save their coefficients as 'one'. So, actually, I want to know the mean of these 250 coefficients. Specifically, I want the mean of alpha (the constant), B0, B1, and B2.
Does anyone know how I can do this? I would be really really happy to receive an answer.
Roy
Related Posts with Time-series regression loop (approximately 250 regressions) and saving their coefficients
convergent validityHow would I calculate the convergent validity of two diagnostic tests where both outcome measures ar…
how to replace missing value by using the information in previous or following year?I am dealing with panel data of 20000 firms for period 2010-2015. I found the identity of some firms…
MGARCH DVECH for panel dataDear all, I am trying to estimate Diagonal VECH GARCH(1,1) model for a panel of 600 firms (ID) and …
Small Variance estimating coefficientsI would just like to confirm my understanding is correct. I am running a panel regression. One of m…
More efficient way to moss p_name, match("([0-9])") regex in a very large dataset?Good morning, I need to find the first instance of [0-9] in a string. I tested the following and it …
Subscribe to:
Post Comments (Atom)
0 Response to Time-series regression loop (approximately 250 regressions) and saving their coefficients
Post a Comment