calculating lead values of an annual variable in a quarterly dataset

[Ama Perera]

Hi all,

I have a dataset with 40 countries. Some variables are available on a quarterly basis while some variables are available on an annual basis.

I have gdp forecasts annual data , I want to calculate GDP forecast at time t+1 , t+2, t+3 and t+4 (year t+1 , t+2 t+3 t+4)

Sample dataset included below:

* Example generated by -dataex-. To install: ssc install dataex
clear
input str97 location double(qdate year gdp_forecast_annual)
"Australia" 41 1970 6.95457121381335
"Australia" 42 1970 6.95457121381335
"Australia" 43 1970 6.95457121381335
"Australia" 44 1971 3.88811418259372
"Australia" 45 1971 3.88811418259372
"Australia" 46 1971 3.88811418259372
"Australia" 47 1971 3.88811418259372
"Australia" 48 1972 2.27675761035412
"Australia" 49 1972 2.27675761035412
"Australia" 50 1972 2.27675761035412
"Australia" 51 1972 2.27675761035412
"Australia" 52 1973 4.35304230982696
"Australia" 53 1973 4.35304230982696
"Australia" 54 1973 4.35304230982696
"Australia" 55 1973 4.35304230982696
"Australia" 56 1974 1.62215552938497
"Australia" 57 1974 1.62215552938497
"Australia" 58 1974 1.62215552938497
"Australia" 59 1974 1.62215552938497
"Australia" 60 1975 2.48283325135266
"Australia" 61 1975 2.48283325135266
"Australia" 62 1975 2.48283325135266
"Australia" 63 1975 2.48283325135266
"Australia" 64 1976 3.83800917313187
"Australia" 65 1976 3.83800917313187
"Australia" 66 1976 3.83800917313187
"Australia" 67 1976 3.83800917313187
"Australia" 68 1977 1.48596679362497
"Australia" 69 1977 1.48596679362497
"Australia" 70 1977 1.48596679362497
end
format %tq qdate

I've tried many codes but none of them works. Could some one please help me with this?

Thanks in advance.

[Clyde Schechter]

I'm not sure I understand your question. You say you have quarterly data, but in the example shown, the variable gdp_forecast_annual is the same for every quarter within the same year. So it looks more like annual data that has been simply reproduced over quarters. If I do understand your question, you have "quarterly" data, and you want to compute leads of the gdp_forecast_annual at 1, 2, 3, and 4 years.

encode location, gen(country)
xtset country qdate

gen lead_1_year = L4.gdp_forecast_annual
gen lead_2_year = L8.gdp_forecast_annual
gen lead_3_year = L12.gdp_forecast_annual
gen lead_4_year = L16.gdp_forecast_annual

If this isn't what you want, please explain more clearly. Better still, post back with an brief example where you show the actual results you want.