The Data Space

Input data shared by every by-contract projection of VA_KR_S.

The eight input CSVs are read here, once per model, and referenced from Projection as data. Projection is parameterized by point_id, so each Projection[N] is a separate ItemSpace with its own cells cache; if the readers lived there, every model point would re-read every file. Holding them in an unparameterized Space reads each file once no matter how many contracts are projected.

Inputs are external files: plain CSVs in the model folder’s parent directory, products/variable_annuity/, rather than data stored inside the model. The model folder therefore holds nothing but formulas — no _data/, no IOSpec, no embedded values — so a diff of the model shows logic changes only, and a table can be swapped without touching a formula. This follows annuallife.TradLife_A; contrast basiclife.BasicTerm_S, which keeps its inputs inside the model.

The consequence worth knowing: the model is not portable on its own. Copying the VA_KR_S folder without its parent’s CSVs produces a model that reads and then fails on first evaluation.

input_dir() resolves the directory from _model.path.parent at run time, so the model works wherever the repository is checked out. Each table has a filename Reference and a reader Cells:

Reference	Cells	File
model_point_file	model_point_table()	model_point_table.csv
mort_table_file	mort_table()	mort_table.csv
lapse_table_file	lapse_table()	lapse_table.csv
fund_file	fund_table()	fund_table.csv
charge_file	charge_table()	charge_table.csv
risk_prem_file	risk_prem_table()	risk_prem_table.csv
return_scenario_file	return_scenario()	return_scenario.csv
crediting_file	crediting_table()	crediting_table.csv

Three of them carry a compound key — mort_table by (sex, age), fund_table by (fund_set, fund_id) and return_scenario by (scenario_id, fund_id) — and three are read without an index because the lookup is a band test on a half-open interval rather than a key lookup: lapse_table on completed policy years, risk_prem_table on attained 보험나이, crediting_table on completed years since annuitisation.

charge_table is the 상품요약서’s own 수수료 안내표, one row per charge line, and it is a table rather than a block of scalar References for a reason: the five lines this product turns on — 계약체결비용, 계약관리비용, 위험보험료, 최저보증비용 and 특별계정 운용보수 — are deducted from different bases at different times and land in different accounts, and the base, timing and account columns carry that beside each rate so a reader can see it without reading a formula.

Every file but model_point_table.csv carries a provenance column and every cell in it begins with a citation tag. That is the library’s rule, and here it is load-bearing rather than decorative: no Korean mortality table in this library is a transcription, the 제10회 경험생명표 not being published in full [REG-R33] [REG-R34], and every return assumption on this product is a standardization as well. The rows say which authority each number stands on.

Cells Descriptions

input_dir()

The directory holding the input CSVs: the model folder’s parent.

Inputs are external files, not data stored inside the model, so the model folder is pure formulas. The path is resolved at run time from where the model was read, following annuallife.TradLife_A.

model_point_table()

The model point table, read from model_point_table.csv.

Ten contracts indexed by point_id. Point 1 is the technical notes’ anchor cell: 남자 보험나이 40, 기본보험료 ₩300,000 월납, 10년납, 연금개시나이 60, 보증형, 채권형 50% / 주식형 50%. This is the only file exempt from the provenance rule, a model point being a configuration rather than an assumption.

mort_table()

Annual mortality by sex and attained 보험나이, from mort_table.csv.

Two bases on one file. mort_rate is the 보험사망률 used for the death decrement of the 연금개시 전 보험기간, against which the GMDB is written; ann_mort_rate is the 연금사망률 (연금생명표) proxy used for the annuity factor and for the payout phase. Both are [std] Makeham constructions: the annuity basis is fitted so its complete expectation of life at 65 rounds to the 제10회 경험생명표 65세 기대여명 of 23.7 years male and 27.1 female [REG-R33] — the fitted values are 23.663 and 27.060 — and the insurance basis is that curve at mu / 0.80. The 경험생명표 qx table is not published [REG-R34]; this file must never be presented as it.

lapse_table()

The annual 해지율 scale by completed policy year, from lapse_table.csv.

Read without an index: the lookup is a band test, each row giving the rate that holds from dur_from completed policy years until the next row. The scale is calibrated so the seven-year persistency is 28.9%, against the only published Korean figure for this product — 「변액보험의 7년 평균 유지율은 30% 미만」, itself second-hand inside [R1].

fund_table()

The 특별계정 fund menu and its allocations, from fund_table.csv.

Indexed by (fund_set, fund_id). Two funds — 채권형 and 주식형 — are the minimum that exercises a pro-rata allocation, a per-fund 운용보수 and the mandatory bond floor at once; real Korean menus run from 5 to 51 funds [S4] [S10]. The three sets shipped sit exactly on the three rungs of the mandatory 채권형 ladder [S1] [R1], asset_class naming which fund the floor is measured on.

charge_table()

The 상품요약서 fee stack, one row per charge line, from charge_table.csv.

Indexed by line. value is a rate or an amount and the base column says which; timing says when it is taken and account which side of the 특별계정 / 일반계정 boundary it lands on. The commission scale and the insurer’s own unit expenses live here too, so that every parameter with a provenance is in a file that carries one.

risk_prem_table()

The 위험보험료 scale by attained 보험나이 band, from risk_prem_table.csv.

Read without an index, the lookup being a band test on age_from. The rate is a fraction of the 기본보험료 and the applicable row is the highest age_from at or below the attained age. The published band is 0.004%–0.011% (₩12–₩32 a month on the anchor cell) [S2] [S4]; the grading across it by age is [std].

return_scenario()

The deterministic gross asset returns, from return_scenario.csv.

Indexed by (scenario_id, fund_id): one annual gross separate-account asset return per fund, held constant for the whole projection. Gross, not net — the 특별계정 운용보수 of fund_table() is deducted inside the 기준가격 afterwards, so the management fee is a modelled cash flow rather than an assumption. Three scenarios are shipped, the base one and the two other returns a Korean variable illustration must show [R2] [REG-R48].

crediting_table()

The payout-phase 공시이율 and 최저보증이율 ladder, from crediting_table.csv.

Read without an index, the lookup being a band test on completed years since annuitisation. Two bases are shipped: decl_2026, on which the declared rate is the 2026 평균공시이율 of 2.50% [REG-R48] and the 최저보증이율 floor of 1.00% / 0.75% / 0.50% [S1] never binds, and min_guar, on which the declared rate is nil so that Max[공시이율, 최저보증이율] resolves to the floor and the ladder is exercised by a shipped model point.