Hello Everyone, I am a last-year student at university, currently working on my Bachelor's (so still learning R), and I really hope that you would suggest a potential solution (even if to use Python).
I need to estimate this model by using the GMM (preferably Arellano and Bond (1991)). I have been searching for the solution for almost 2 months, yet still not succeeded.
The data set could be found here:
https://drive.google.com/file/d/1uMmZTBDCaq1YoKO8KXU1L8EiqcKVe5V3/view?usp=sharing
My code is the following:
PLVData_A <- pdata.frame(LVData_A, index = c("ID","Year"))
z1 <- pgmm(domegaACF_A ~ lag(domegaACF_A, 1) + ddebt + ff1 + ff1:ddebt + Age + ta + dsales | lag(domegaACF_A, 1:2),
data = PLVData_A, effect = "twoways", model = "onestep", transformation = "d")
summary(z1, robust = TRUE)
After I have created all the needed variables, I am trying to estimate coeffs by using pgmm function in plm package. And receive 3 types of errors:
In case I use transformation = "d"
- 1 type 1 picture
In case I use transformation = "ld"
, errors are as follows: 2 type of error - second picture
In case I delete lag from the model and put it after "|" sign I get: 3 type of error - third picture
I would really appreciate any comments and suggestions provided because I do not know how to get away from this dead end. Thank you in advance!)
Please ask for the data and any explanation, since I am really desperate to find the solution.
UPD: I have created lags as separate variables, so here is a new correlation matrix:
domegaACF_A ddebt ff1 Age ta dsales lag1_domegaACF_A lag2_domegaACF_A
domegaACF_A 1.000000000 -0.014777102 -0.002600866 -0.019160423 0.02456158 0.256801279 -0.379350157 -0.027687422
ddebt -0.014777102 1.000000000 0.128264730 0.004706522 0.03878795 0.057488971 0.018800962 -0.003222902
ff1 -0.002600866 0.128264730 1.000000000 -0.048072279 -0.02868682 0.008979745 -0.002808377 0.008062733
Age -0.019160423 0.004706522 -0.048072279 1.000000000 0.27884112 -0.021440539 -0.051829392 -0.048162046
ta 0.024561579 0.038787947 -0.028686815 0.278841116 1.00000000 0.078247618 0.015304990 0.015539599
dsales 0.256801279 0.057488971 0.008979745 -0.021440539 0.07824762 1.000000000 -0.049518811 -0.004134281
lag1_domegaACF_A -0.379350157 0.018800962 -0.002808377 -0.051829392 0.01530499 -0.049518811 1.000000000 -0.358671174
lag2_domegaACF_A -0.027687422 -0.003222902 0.008062733 -0.048162046 0.01553960 -0.004134281 -0.358671174 1.000000000
And I believe that correlation is not much trouble.