The Data Space

Input data shared by every by-policy projection.

The four 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 policies are projected.

Inputs are external files: plain CSVs in the model folder’s parent directory, products/ci_insurance/, 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. This follows annuallife.TradLife_A; contrast basiclife.BasicTerm_S, which keeps its inputs inside the model through modelx’s IOSpec machinery.

The consequence worth knowing: the model is not portable on its own. Copying the CI_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
ci_incidence_file	ci_incidence_table()	ci_incidence_table.csv
lapse_table_file	lapse_table()	lapse_table.csv

Every decrement in this model is a construction, and the reason is Korean

Korea publishes neither of the two tables this product needs.

경험생명표 (gyeongheom saengmyeongpyo, the industry experience life table, currently the 제10회 applied from 2024-04) is produced by 보험개발원 and is not published in full: only summary statistics — 평균수명 and 기대여명 — are released [REG-R33] [REG-R34]. The life 참조순보험요율 is defined by 감독규정 제1-2조제1호 as the 위험률 the bureau files with the supervisor, not as a published table [REG-R4] [REG-R34]. The 장기손해보험 참조순보험요율 display 보험개발원 does publish carries an insured-cancer 발생률 grid and a 질병입원율 grid [REG-R61] — the bases Cancer_KR_S and Medical_KR_S use — and no 중대한 질병 item at all, so nothing on it reaches this model. The 산출방법서 that would carry a carrier’s own 적용위험률 is a 기초서류, filed and never published [REG-R2].

What does exist, and it is the whole public evidence base for this product, is a single 2011 상품요약서 that prints its 예정위험률 by sex at ages 20, 40 and 60 [S3]. Both decrement files are built on it.

mort_table.csv — sex, attained age 15 to ω = 110, mort_rate, provenance. The male rates are a Makeham form fitted exactly to [S3]’s three disclosed male anchors in the rate itself, not in the force — q(y) = A + B c^y — used to age 60, and log-linear in q from there to q(110) = 1; extrapolated, that fit reaches a rate of 1 at about attained age 107 and stands a third above the shipped ramp by age 100, which is why the old-age shape is a separate [std] rule rather than a continuation. The female rates are 0.5294 times the male rates at every age below ω, that ratio being [S3]’s own disclosed female-to-male ratio at age 20 — the only usable female anchor, because [S3]’s female rates at 40 and 60 extract identical to the male ones and are a PDF column-merge artefact expressly marked [unverified]; at ω both sexes carry the terminal rate of 1. The construction is known to understate the female advantage: it gives a 15-to-80 death probability ratio of 0.560 against the 0.500 implied by 국가데이터처’s survival to age 80 of 남 64.4% / 여 82.2% [REG-R38]. The row-by-row tags say which rows are anchors and which are fill.

ci_incidence_table.csv — sex, attained age 15 to 100, cause, ci_rate, provenance, in long form so that each cause carries its own tag. Five causes:

cancer, ami, stroke

중대한 암, 중대한 급성심근경색증 and 중대한 뇌졸중, [S3] at ages 20, 40 and 60, log-linear in ln(rate) between and below those anchors, and above 60 on the 40-to-60 log-slope decaying geometrically at 0.90 a year [std] — undamped, the cancer rate passes 1 before age 100.

other

the five remaining 중대한 질병, the four 중대한 수술 and 중대한 화상 및 부식, as 10.5% of the three headline rates [std]: [S4]’s published 3대-to-17대 office premium step of 5.30% divided by the 50.6% share the CI benefit takes of the risk premium at male 40 on [S3]’s 보장위험별 연간보험료 disclosure.

ltc

장기요양상태 on 노인장기요양 1·2등급, nil below 65 and 0.12% at 65 growing 14% a year [std] — a placeholder for the construction LTC_KR_S owns, scaled so the 65+ mean is of the order implied by [REG-R42]’s 154,688 1·2등급 인정자 at an assumed three-year mean duration. This product’s own source set retrieved no inception rate; the library’s did — LTC_KR_S sources a disclosed 요양 1·2등급 발생률 grid at ages 40, 50 and 60 by sex, whose male 1·2등급 rate at 60 is 0.000530, about 2.5% of this model’s CI rate there. Holding this limb at nil below 65, and not modelling the 노인성 질병 route below 65 [REG-R55] at all, understates the CI decrement at insured ages by that order.

These are 예정위험률, not best estimate. [R1] records the margin regime around them — 안전할증 on the 기초발생률 capped at 30% in the early 2000s, raised to 50% in 2015 and uncapped from 2017 — so the base run is a valuation-basis run. mort_be_factor and ci_be_factor are the levers that turn it into a best estimate and they are 1.00 on every model point but one.

lapse_table.csv — policy year, lapse_rate, provenance. The 표준형 comparison curve, three rows short of a decade with a level tail. No CI lapse experience of any kind was retrieved [R1], so the curve is [std], bounded only by the 적용해지율 envelopes Korean 상품요약서 publish. The suppressed forms do not use it: they run the 로그-선형 원칙모형 the IFRS17 주요 계리가정 가이드라인 prescribes for 무·저해지 business, converging to 0.1% at 납입완료 with a 0.8% post-완납 ultimate [REG-R27], which is a formula rather than a table and lives in Projection. Carrying both is the comparison the guideline itself requires an insurer to disclose.

model_point_table.csv is the exception to all of this: it is a configuration rather than an assumption, and carries no provenance column.

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.

Indexed by point_id. Point 1 is the technical notes’ worked-example anchor cell; the other eight exercise the product’s variants, its optional modules and the two ends of its issue-age and sum-assured envelopes.

mort_table()

The mortality rates by sex and attained age, from mort_table.csv.

A [std] construction anchored on the 예정 경험 사망률 disclosed at three ages by one Korean CI 상품요약서 [S3]; not the 경험생명표, which 보험개발원 does not publish in full [REG-R33]. Indexed by (sex, age) with a single mort_rate column and a provenance column saying, row by row, whether the rate is an anchor, a fit, a ramp or the terminal age. ω is the first age at which the rate reaches 1.

ci_incidence_table()

The CI incidence rates by sex, attained age and cause, from ci_incidence_table.csv.

Long form — one row per (sex, age, cause) — so that each cause carries its own provenance tag, which matters because they do not rest on the same thing: the three headline causes are [S3] at three ages and [std] everywhere else, while other and ltc are standardizations throughout.

These are first-event rates across a competing-risk set, not marginal incidences. The benefit is payable once only across every trigger, and Korea’s supervisor required the overlap between CI causes to be reflected in the filed rate rather than ignored for rate stability as overseas practice does [R1]. Summing published site-specific incidences instead would be wrong in the direction the regulation specifically addresses.

lapse_table()

The 표준형 voluntary surrender rates by policy year, from lapse_table.csv.

The policy_year key is the contractual, 1-based label, not the model’s 0-based time index: the projection’s period t reads row t + 1, so the file is unaffected by the frame’s indexing. Six duration rows and a level tail read for every later year. Used only where the model point sets lapse_basis to table; the suppressed forms run the 로그-선형 원칙모형 of [REG-R27] instead, which is a formula and lives in Projection.