Given atom values differ across Monte Carlo paths
When persisting a Tier 1 run
Then each stored atom is persisted as a distribution summary per year using the 7 percentiles (p5/p10/p25/p50/p75/p90/p95)
Stored distributions (per year)
capital (end-of-year snapshot distribution)
deposit_year (yearly total distribution)
return_year (yearly credited P&L total distribution)
withdraw_year (yearly total distribution)
tax_year (yearly total distribution)
fee_year (yearly total distribution)
inflation (end-of-year multiplier; store as scalar if deterministic, distribution if stochastic)
Acceptance criteria
For each year and each stored atom metric, DB contains values for p5/p10/p25/p50/p75/p90/p95
Deterministic metrics are stored consistently (either as a scalar + documented, or as a distribution with identical percentile values)
Given atom values differ across Monte Carlo paths
When persisting a Tier 1 run
Then each stored atom is persisted as a distribution summary per year using the 7 percentiles (p5/p10/p25/p50/p75/p90/p95)
Stored distributions (per year)
capital (end-of-year snapshot distribution)
deposit_year (yearly total distribution)
return_year (yearly credited P&L total distribution)
withdraw_year (yearly total distribution)
tax_year (yearly total distribution)
fee_year (yearly total distribution)
inflation (end-of-year multiplier; store as scalar if deterministic, distribution if stochastic)
Acceptance criteria
For each year and each stored atom metric, DB contains values for p5/p10/p25/p50/p75/p90/p95
Deterministic metrics are stored consistently (either as a scalar + documented, or as a distribution with identical percentile values)