Return to PSG Models documentation page
## GEMINI output file .ARC

--- .ARC file contains revenue and cost statistics for social-
security-account immediate-annuity provider.
The GEMINI runNNNNN.arc file has no header or footer lines.
There is one line for each cohort age for each scenario as well as one
summary line for each scenario.
The statistics on a line are separated by the tab character.
The arc statistics on each scenario AGE LINE are as follows:
(1) scenar: scenario number
(2) age: age of individuals in simulated birth cohort
(3) ap_rev: annuity provider revenue from selling annuities at this age
(4) ap_cst: annuity provider cost from making annuity payments at this age
(5) ap_d_r: annuity provider discount rate at this age (in percent)
The arc statistics on each scenario SUMMARY LINE are as follows:
(1) scenar: scenario number
(2) pv@65
(3) pv_rev: present value of annuity provider revenue from selling annuities
(4) pv_cst: present value of annuity provider cost from making annuity payments
(5) rc_rat: ratio of pv_rev to pv_cst for this scenario (pv_rev/pv_cst)
On all lines, rev and cst statistics, which have been inflated up to
population totals using the PENSIM:RUN.sample_pct, are expressed in
billions of SSASIM:COHORT.cpi_year dollars.
The present value statistics are computed using nominal cashflows and the
nominal discount rate (ap_d_r, which equals the nominal yield on Treasury
bonds), and then the nominal pv amount at age 65 is converted to real terms
so that it is also expressed in billions SSASIM:COHORT.cpi_year dollars.
NOTE: in a stochastic run that has more than one scenario, be sure
that the all-scenario pv@65 revenue-to-cost ratio is at least
one. To compute the all-scenario ratio at the Windows command
prompt in the directory containing the runNNNNN.arc file use
the following two commands:
> gawk "$2~/pv/" runNNNNN.arc | mean - 3
> gawk "$2~/pv/" runNNNNN.arc | mean - 4
The first command calculates the mean value across all scenarios
of the revenue statistic (3), while the second command calculates
the mean value across all scenarios of the cost statistic (4).
The all-scenario ratio is simply the first mean divided by the
second mean.