cummulative change

[Lu Asegu]

Dear listers,
I would like to apologize if my question is too naive. I am using Stata 15.

I would like to replicate the following graph from Blanchard and Katz (1992) paper. I tried to graph a similar graph using compound average growth rate (CAGR) in the following way :

* Example generated by -dataex-. To install: ssc install dataex


webuse grunfeld, clear

gen linvest=ln( invest)
bys company: gen cagr= (linvest[20]/linvest[1])^1/20 -1
line cagr year if company==6

I tried the

cumul

command as well

ssc install cumul

* Example generated by -dataex-. To install: ssc install dataex


sysuse citytemp, clear
       cumul tempjan, gen(cjan)
        cumul tempjuly, gen(cjuly)
        stack cjan tempjan cjuly tempjuly, into(c temp) wide clear
        line cjan cjuly temp, sort

But the line is far from looking like the one presented in the graph.
May anyone help me, how I could get a graph similar to theirs please?
Thank you.
cummulative change.pdf

[Andrew Musau]

Cumulative change is just summing up change over time. So cumulative change in time \(t=3\) is \(change_{t=2} \;+\; change_{t=3}\), cumulative change in time \(t=4\) is \(change_{t=2} \;+\; change_{t=3}\;+\; change_{t=4}\), and so on. In the attached figure, Blanchard and Katz are summing up change in log employment across a selection of US states, starting in 1945. Here is how you would graph cumulative change in log investment for Company 1 in the Grunfeld dataset.

webuse grunfeld, clear
gen loginvest= ln(invest)
xtset company year
*CUMULATIVE CHANGE IN LOG INVESTMENT
bys company (year): gen wanted= sum(D.loginvest)
set scheme s1mono
xtline wanted if co==1, xtitle("") ytitle("Cumulative Investment Growth") byopts(note("")) subtitle("Company 1: Grunfeld dataset")

[Lu Asegu]

Dear Andrew,

That's very kind of you to reply to my question. Your explanation is quite simplifying as well, appreciated.