Product Specification#
Status: Draft, 2026-08-20 (all cited sources accessed 2026-08-20).
Scope note. This is a standardized composite specification assembled for reference
liability cash-flow modeling. It does not describe any single insurer’s product. Facts
carrying a source tag — [S#] (primary product documents: policy conditions (yakkan, 約款),
policy booklet (go-keiyaku no shiori, ご契約のしおり), contract summary (keiyaku gaiyō, 契約概要),
product summary (shōhin gaiyō, 商品概要) and product pages) and [R#] (product-specific
regulatory, actuarial and statistical references), both numbered per _research/cancer.md
and resolved in sources.md (same directory; numbering frozen, never renumbered), and
[REG-R#] (the cross-product reference library
references/regulatory-and-actuarial-references.md, whose own R-numbering is distinct) —
were extracted from the cited document. Values marked std are standardizations
introduced for the reference implementation; each std table row carries a numbered
footnote giving the rationale and the observed range across insurers. Facts the research
file could not verify are flagged unverified. The composite is drawn from seven
carriers: five life insurers writing a stand-alone cancer main contract on the
fixed-benefit chassis [S1] [S2] [S3] [S5] [S6] [S7] [S8] [S11] [S12], one life insurer
writing a treatment-benefit-only main contract [S10], and one non-life insurer writing
the non-reimbursed medicine (jiyū shinryō, 自由診療) expense-reimbursement variant [S13].
Two of the cited documents downloaded but could not be read — an image-only 契約概要 [S9] and a
subset-font brochure whose numerals all drop out [S4] — and the facts they would have carried
are named as gaps where they arise.
Deltas against the medical insurance (iryō hoken, 医療保険) chassis. The
medical product specification (医療保険) is the jplib third-sector
(dai-san-bun’ya, 第三分野) chassis: daily amount (nichigaku, 日額) times paid days,
capped per hospitalization and again in aggregate, with event benefits layered on. This
product inherits the chassis and changes five things, each of which is a first-order modelling difference and none of which is
cosmetic.
A 90-day waiting period (menseki kikan, 免責期間) from attachment of cover (sekinin kaishi, 責任開始) before any cancer benefit is payable — and a diagnosis inside it makes the whole contract void, not merely unpayable. No other product in this library has a general waiting period on its main cover.
Cancer diagnosis lump sum (gan shindan ichijikin, がん診断一時金) as a repeating lump sum on a stated cycle, not a once-only benefit. The cycle clock is a state variable with memory of its own, separate from the 180-day one-hospitalization memory in
medical.がん入院給付金 with no day limit at all — no per-hospitalization day limit (shiharai gendo nissū, 支払限度日数) and no lifetime aggregate (tsūsan, 通算) limit. This is the sharpest contrast with
medical, whose 60/120-day cap and 1,095-day aggregate are the defining features of that product’s benefit formula, and it removesmedical’s benefit-driven termination decrement entirely.Carcinoma in situ (jōhinai shinseibutsu, 上皮内新生物) paid at a reduced rate, on a separate and separately capped benefit, rather than covered or not covered.
Treatment-based monthly benefits — chemotherapy benefit (kō-gan-zai chiryō kyūfukin, 抗がん剤治療給付金) and radiotherapy benefit (hōshasen chiryō kyūfukin, 放射線治療給付金) — which pay per calendar month in which a qualifying treatment occurred, not per event and not per day. They are increasingly the centre of the product, and on one retrieved contract they are the only main-contract benefits [S10].
Product overview and market role#
Cancer insurance (gan hoken, がん保険) is 第三分野 business. The regulator names it directly: 第三分野 covers 医療保険, がん保険 and 介護保険, and its benefits are paid on disease (疾病) or injury (傷害) as 保険金 or 給付金 for treatment R3. Because 保険業法 第3条 makes that class writable under either a life or a non-life licence REG-R1, the same risk is sold in two structurally different ways — a point the composite has to take a position on, since one of the seven carriers here holds a 損害保険業免許 [S13].
It is the fourth-largest individual line in Japan by policy count in force, and the third largest by new business count. At 31 March 2025 ガン保険 stood at 2,522万件 in force, 12.9% of all 個人保険 policies — behind 医療保険 (23.3%), 終身保険 (19.7%) and 定期保険 (13.9%) — and took 159万件, 12.8%, of new business by count, behind only 医療保険 (296万, 23.8%) and 終身保険 (231万, 18.6%) REG-R31. Third-sector 年換算保険料 in force was 7兆3,062億円, a series that has risen every year published REG-R31. The line has been open to the whole industry only since the January 2001 liberalisation; by 平成16年度 third-sector business already exceeded ¥3.5 trillion of annualised premium and more than 20% of life insurers’ in-force R3.
Penetration depends sharply on how it is measured, and the two published figures are not comparable. On a 民保加入世帯 base — households already holding private life cover, n = 3,085 — 68.2% hold がん保険 or a がん特約 (tokuyaku, rider) (60.7% of 世帯主, 46.5% of 配偶者) REG-R32. On the household base of a different survey covering all providers including JA and 共済, N = 4,837, cancer cover stands at 39.9%, and 35.4% through private life insurers alone, a series with no consistent trend over time R10. Both are quoted here with their bases attached; a model point should not be justified by silently picking the larger.
Why it is a mass-market product rather than a niche rider is arithmetic. Lifetime incidence probability in Japan is 61.1% for men and 50.1% for women REG-R28 — roughly one person in two will be diagnosed. Registered incidence in 2023 was 993,469 cases (male 556,059, female 437,406) at a crude rate of 798.92 per 100,000 R5 REG-R28. And the insurable event is survivable: five-year relative survival for 2018 diagnoses was 63.17% for men and 66.84% for women, all sites R6. A benefit menu built around a diagnosis lump sum and a treatment-linked payment stream, rather than a death benefit, follows directly from that survival rate.
The public incidence data is what makes this product modellable, and its shape constrains the model. 全国がん登録 publishes counts and rates by 5-year age band, site, sex and diagnosis year, freely downloadable R5 REG-R29 — a citable basis of a kind no UK protection model has for its morbidity. Age-specific crude rates per 100,000 (2023, both sexes) run 24.38 at 20–24, 220.28 at 40–44, 959.00 at 60–64, 1,948.71 at 70–74 and 2,497.39 at 85–89 R5. Two features of that data are load-bearing. First, the male and female curves cross: female incidence exceeds male from roughly age 25 to roughly age 55 (35–39: female 193.89 against male 72.92) and is far below it afterwards (70–74: male 2,684.60 against female 1,291.07) R5. A unisex cancer basis is materially wrong at every age. Second, the file publishes paired rows with and without 上皮内がん: 全部位 C00–C96 at 993,469 against 全部位（上皮内がん含む） C00–C96 D00–D09 at 1,114,642 — an in-situ increment of 121,173, 12.2% of the invasive count R5. The in-situ grading a contract applies can therefore be sourced rather than assumed, though the increment quoted is an all-ages figure and should not be assumed age-invariant.
What the inpatient data explains is why the product looks the way it does. Mean length of stay for 悪性新生物 discharges was 14.4 days in September 2023, against 28.4 days for all conditions, with only a mild age gradient (35–64: 10.7; 65+: 15.5; 75+: 17.6) R7 REG-R27. A daily benefit with no day limit is affordable precisely because cancer stays are short — and, symmetrically, the economic weight of the product has migrated to the diagnosis and treatment benefits, which do not depend on inpatient days at all. Two of the seven carriers no longer carry an inpatient benefit in the main contract [S7] [S10]. What is not available is a length-of-stay distribution: the finer 患者調査 grids sit on e-Stat under 統計表 Z124-x and Z134 and were not fetched for this product R7 REG-R33, so every statement below about stay length rests on a mean.
Two structural facts frame everything below. First, the market runs two incompatible chassis: the fixed-benefit (定額給付) chassis, where every benefit is a fixed amount or a fixed multiple of a base amount, at six of the seven carriers [S1] [S5] [S6] [S7] [S10] [S11]; and the expense-reimbursement (実損てん補) chassis, where the actual treatment cost is indemnified including 先進医療 and 自由診療, at the non-life carrier [S13]. Second, within the 定額給付 chassis the design has been moving from diagnosis-and-hospitalisation centred [S1] [S5] [S6] to treatment centred [S10] [S11] — on one contract the main contract pays only the two monthly treatment benefits and everything else is a 特約 [S10]. The composite specified below sits deliberately between the two, because a reference model that implements only one cannot express the other.
Representative specification#
Product identity and issue rules#
Parameter |
Representative value |
Basis |
|---|---|---|
Design type |
がん保険, non-participating, no surrender value (無配当・無解約払戻金型); main contract (shu-keiyaku, 主契約) paying diagnosis, inpatient, surgery, treatment and outpatient benefits, with 特約 attached |
[S7] [S10] [S11]; menu std (1) |
Regulatory class |
第三分野 (保険業法 第3条第4項第2号 ／ 第5項第2号) |
|
Chassis |
Whole-of-life cover (終身); 10年更新 定期型 as a model-point flag |
[S1] [S6] [S10] [S11] vs [S5] [S7] [S13]; std (1) |
Issue age (契約年齢) range |
20–75 |
observed 0–75 [S15], 18–80 [S12]; std (2) |
Age basis |
契約年齢 = age last birthday at the 契約日, incremented at each 年単位の契約応当日 (「24歳7か月の被保険者の契約年齢は24歳」) |
[S8] |
Policy term (保険期間) |
終身 (to the terminal age of the mortality table); no maturity benefit |
[S1] [S6] [S10] [S11] |
Premium-paying period (保険料払込期間) |
Pay for life (終身払) default; 65歳払済 短期払 as a variant |
[S10] [S12] [S15]; std (3) |
Base benefit amount (基本給付金額) |
¥5,000 / ¥10,000 menu; composite default ¥10,000 |
[S2] [S3] [S15]; default std (4) |
Underwriting |
Written declaration (kokuchi-atsukai, 告知扱い): the application and the 告知 together are the acts that complete underwriting and start cover [S8], and 復活 needs a fresh 告知 [S1] [S8]. Whether a medical examination is ever required is not stated in any retrieved document |
[S1] [S8]; unverified on the examination |
Lives basis |
Single life only — no retrieved contract writes a joint life, but none states the restriction either |
[S1] [S5] [S6] [S7] [S10] [S11]; observed absence, std (1) |
死亡保険金 in the main contract |
None — 「この保険に死亡保険金はありません」 |
[S1]; against [S6]; std (1) |
Cancer cover start (がん責任開始日) |
The 91st day counting the 責任開始日 as day 1 |
[S1] [S5] [S6] [S10] [S13]; std (9) |
Anchor model cell |
Male, 契約年齢 40, 終身 chassis, 終身払, 基本給付金額 ¥10,000 — giving がん診断一時金 ¥1,000,000 on a 2-year repeat cycle, がん入院給付金日額 ¥10,000 with no day limit, がん手術給付金 ¥200,000, がん治療給付金 ¥100,000 per treatment month capped at 60 months, がん通院給付金日額 ¥10,000; 上皮内新生物 at 50% of the diagnosis lump sum; がん責任開始日 at day 91; premium waiver on first invasive diagnosis; 先進医療特約 attached; level premium ¥3,000 per month |
std (5) |
Footnotes to std rows:
The retrieved contracts fall into three menu shapes and they are not interchangeable: diagnosis + inpatient + surgery (+ discharge or outpatient) at three carriers [S1] [S5] [S6]; diagnosis + outpatient in the main contract with every treatment benefit in riders at a fourth [S7]; and a main contract paying only 放射線治療給付金 and 抗がん剤・ホルモン剤治療給付金, with hospitalisation, surgery, diagnosis and outpatient all in nine named 特約, at a fifth [S10]. A sixth folds surgery, radiation and chemotherapy into a single monthly 治療サポート給付金 [S11]; the seventh indemnifies actual cost [S13]. The composite main contract carries all five benefit types on the 定額給付 chassis, each independently switchable, so that a model point can be configured into any of the first six shapes. 終身 is the chassis because four of the seven write it [S1] [S6] [S10] [S11], against a 10-year renewable term at two [S5] [S7] and a 5-year renewable term at the expense carrier [S13]; the 定期 flag carries automatic renewal with benefit history, premium-waiver history and 責任開始期 treated as continuous across the renewal [S5]. A 死亡保険金 exists on exactly one retrieved contract — がん入院給付金日額 times a chosen 倍率, floored at the policy reserve [S6] — and is excluded: it is the minority design and it would import a death-benefit reserve into a morbidity product.
Observed: 0 (from 15 days old at 告知) to 75 [S15], on a distributor listing because the carrier states it only in the unreadable [S4]; 18–80 [S12]; a specimen contract at age 30 [S5]; not published in the retrieved documents at the other four carriers. The two published whole-of-life ranges overlap on 18–75. The composite takes 20–75: the lower bound is raised to 20 because juvenile cancer is a different risk written by one carrier only, and the age-specific incidence rate below 20 is an order of magnitude below the adult rates (16.51 per 100,000 at 15–19 against 220.28 at 40–44) R5; the upper bound is the tighter of the two published values.
Observed: 60歳／65歳/終身 [S15]; 55/60/65/70歳払済 or 終身払 [S10]; 終身払 only [S12]. 終身払 is the default because it is offered everywhere and is the design under which no surrender value ever arises on any retrieved contract (footnote 22). The 短期払 variant is retained because it is the only route by which this chassis acquires a surrender value at all.
The benefit schedule is expressed as multiples of a base amount at three carriers — 基本給付金額 [S2] [S3], がん入院給付金日額 [S6], 特約給付金額 [S7] — and as free-standing yen amounts at the rest. The only published course menu is ¥5,000 and ¥10,000, with the ¥5,000 course restricted to ages 50–75 [S15]; the ¥10,000 course is the one the carrier’s own 契約例 uses and the one for which the 100× diagnosis multiple gives the ¥1,000,000 headline [S2] [S3]. ¥10,000 is therefore the default: it is the amount at which the published multiples and the published headline figure are mutually consistent. It also sits at the bottom of the independent ¥1,000,000–¥3,000,000 selectable ladder at another carrier [S12] and equals the flat ¥1,000,000 of the expense product [S13].
Age 40 male is chosen because it is the age at which the only retrieved premium example renews [S5] and because it sits on the steep part of the incidence curve without being in the tail (220.28 per 100,000 at 40–44, against 959.00 at 60–64) R5. No carrier publishes a rate table for this product (footnote 7), so the ¥3,000 monthly premium is a pure modelling value, not a computed or quoted one. It is anchored as follows: the single published price point is a 10-year term at 診断給付金額 ¥2,000,000 and 入院給付金日額 ¥20,000 — twice the composite’s benefit amounts — costing ¥1,456 per month at age 30 and ¥2,082 on renewal at age 40, on a 2013-10-22 calculation basis [S5]. Halving the benefit amounts and repricing a 10-year term as a whole-life-pay 終身 contract with a repeating diagnosis benefit, a monthly treatment benefit and an unlimited outpatient benefit pushes it back up; ¥3,000 is the round figure in that neighbourhood. It is a model-point input in every sense and no result in this library depends on its being a market rate.
Benefit provisions#
Parameter |
Representative value |
Basis |
|---|---|---|
Waiting period |
90 days: cover attaches on the 91st day counting the 責任開始日 as day 1 |
|
Diagnosis inside the waiting period |
The contract is void (無効), not merely the claim unpayable; premiums refunded where neither policyholder nor insured knew of the diagnosis before 告知, retained where either knew |
[S1] [S5] [S10]; std (9) |
Reach of the voidness rule |
Not applied where no benefit event occurs within 5 years of the がん責任開始日 |
[S1]; std (9) |
Definition of がん |
By 約款別表 keyed to ICD codes; 診断確定 by a Japanese-qualified physician or dentist on biopsy (病理組織学的所見), other findings admitted where biopsy is not obtainable |
[S1] [S6] [S10] [S11] |
Date of diagnosis |
The date the diagnostic test was performed, not the date the result was communicated |
[S1] |
がん診断一時金 |
100 × 基本給付金額 = ¥1,000,000 at the default course |
[S2] [S3] [S13]; level std (11) |
Repeat cycle |
Payable again on 再発 / 転移 / 新たに生じた cancer, at most once in any 2 years, with no lifetime cap |
[S5] [S7] [S10]; std (12) |
2-year clock measured from |
The date of the previous payment trigger |
[S5]; std (13) |
Continuing hospitalisation at cycle expiry |
Still an inpatient on the day after the 2-year period expires ⇒ deemed a fresh trigger |
[S1]; std (13) |
上皮内新生物 diagnosis benefit |
50% of the diagnosis lump sum, payable once over the policy term, on a separate cap; not payable after a full-rate cancer benefit has been paid |
[S6] [S11]; std (10) |
上皮内新生物 on other benefits |
Paid in full on the inpatient, surgery, treatment and outpatient benefits; does not trigger the premium waiver |
[S7] [S10]; std (10) |
がん入院給付金 |
入院給付金日額 × 入院日数 for a stay whose direct purpose is cancer treatment — no per-hospitalization day limit and no 通算 limit |
|
Adjacent non-cancer days |
Days after a non-cancer illness begins during a covered cancer stay, and days before the diagnosis date in a stay begun for another reason, count into the cancer benefit where the insurer accepts the stay was for cancer treatment |
[S6] |
がん手術給付金 |
20 × 基本給付金額 = ¥200,000; unlimited number of payments; two or more simultaneous procedures count as one |
[S2] [S6]; std (15) |
がん治療給付金 (monthly) |
¥100,000 (= 10 × 基本給付金額) per calendar month in which a qualifying chemotherapy, hormone-therapy or radiotherapy treatment occurred; several treatments in one month pay once; lifetime cap 60 months |
[S5] [S10] [S11]; std (16) |
Qualifying treatment definition |
Anchored to public classifications, not a clinical list: 総務大臣が定める日本標準商品分類 「8742 腫瘍用薬」 (oral administration excluded) [S1], or drugs scored under 薬剤料 / 処方せん料 in the 医科・歯科診療報酬点数表 [S5] [S10] |
[S1] [S5] [S10] |
Trigger-change clause |
The insurer may change the payment triggers prospectively, with 主務官庁 approval, if the public medical insurance scheme changes; two months’ notice |
[S1] [S5] |
がん通院給付金 |
通院給付金日額 ¥10,000 per day of outpatient attendance for surgery, radiation, thermal therapy or non-oral chemotherapy — treatment-linked, no day limit and no 通算 limit |
[S1] [S7]; std (17) |
Outpatient during a paid stay |
Not payable |
[S7] [S10] |
Exclusions (免責事由) |
Not extracted for this product line — the exclusion lists sit in the non-extractable 約款 halves |
unverified; std (18) |
Suicide |
No death benefit exists in the composite, so the 免責 clause has nothing to bite on |
|
Termination on benefit payment |
None — payment of the diagnosis lump sum neither terminates the contract nor exhausts it; cover runs to the terminal age |
[S1]; against [S11]; std (12) |
The defining feature of the product, and the one modelling delta that has no analogue anywhere else in this library. Every retrieved cancer contract has a waiting period, and it is expressed two ways. Five carriers write 90 days with cover from day 91: 「責任開始日を含めて91日目」 [S1], 「保険期間の始期からその日を含めて90日を経過した日の翌日」 [S5], 「責任開始期の属する日からその日を含めて90日目の日の翌日」 [S6], 「保険期間の始期からその日を含めて91日目」 [S10], and 「90日の待機期間」 with cover from day 91 [S13] — four different wordings denoting the same day. Two write three calendar months: 「申込および告知がともに完了した日…から３か月を経過した日の翌日」 [S8] and 「その日を含めて3ヶ月を経過した日の翌日（応当日がない場合はその月の末日）」 [S11]. 生命保険文化センター states 90 days as the market norm R11. The composite takes 90 days. On a monthly projection grid the two formulations collapse into the same thing — three months of the grid — and the model must not pretend to a precision it does not have. What does not collapse is the consequence: the waiting period is not an exclusion but an invalidity rule, 「保険契約・特約は無効とします」 [S1], 「知、不知にかかわらず、ご契約は無効とします」 [S5], and the same at two more carriers [S10] [S6]. One carrier caps its reach at 5 years from the がん責任開始日 [S1] and the composite adopts that cap, because an uncapped voidness rule is a rescission right of indefinite duration and no other retrieved contract states one. Premium treatment during the window also differs: five carriers charge from inception, since the premium waiver and any non-cancer cover are already running [S1] [S5] [S6] [S7] [S10]; one charges nothing at all during the three months and says explicitly that this is not a discount [S11]; the seventh does not state it [S13]. The composite charges, the majority position. On reinstatement (fukkatsu, 復活) the waiting period restarts from the 復活日, except that a 復活日 earlier than the original がん責任開始日 leaves the original date standing [S1] [S6].
The single largest source of benefit-level variation in the product, and it splits three ways. Full rate — in-situ inside the definition of がん — at four carriers: 「悪性新生物（上皮内新生物を含みます。）」 [S1], 「別表5に定める悪性新生物および上皮内新生物」 [S5], 「ガン（悪性新生物・上皮内新生物）」 [S10], and the expense product [S13]. Half rate, as a separate benefit, at two: 上皮内がん診断給付金 at 指定倍率の1/2 [S6] and 上皮内新生物診断一時金 at 50% of the cancer lump sum [S11]. 10% at one: 「上皮内新生物の場合：診断給付金額の10％」, applied to both the once-only and the repeating diagnosis benefits [S7]. The composite takes 50% — the median of the three treatments, and the design under which the in-situ benefit is a distinct, separately capped benefit whose incidence can be sourced from the with/without-上皮内がん rows of the registry R5 rather than folded invisibly into the cancer rate. Two consequential rules come with the half-rate design and the composite adopts both: the in-situ benefit is payable once in its own right, and it is not payable once the full-rate cancer benefit has already been triggered [S6] [S11]. Grading is not applied uniformly across the menu: the 10%-grading carrier pays 手術治療給付金, 放射線治療給付金 and 通院給付金 in full for in-situ while excluding it entirely from the chemotherapy, non-scheduled-treatment, appearance-care, palliative-care and women’s-cancer benefits [S7], and in-situ is excluded from the premium waiver at another [S10] and from the advanced-medicine benefit at a third [S11]. The composite grades the diagnosis benefit and the waiver only, and pays every other benefit in full.
The 100× multiple is sourced [S2], as is the resulting ¥1,000,000 at the ¥10,000 course [S2] [S3]; what is standardized is the level, because the market spread is wide. The observed points are ¥1,000,000 at the default course [S2] [S3]; a selectable ladder of ¥1,000,000–¥3,000,000 in ¥500,000 steps, with ¥500,000 additionally available from age 66 [S12]; ¥1,000,000 flat on the expense product [S13]; がん入院給付金日額 × a 給付倍率 chosen by the policyholder from a range the insurer sets, with the in-situ version at half that 倍率 [S6]; and ¥2,000,000 in the one published 契約例 [S5]. ¥1,000,000 is taken because it is the bottom of the selectable ladder, the published headline at two carriers and the flat value at a third — three of the five points that state one — and because at a 2-year repeat cycle with no lifetime cap the level is the least interesting parameter in the product and should be the one a model point varies first.
Observed repeat cycles: once ever at three carriers [S1] [S6] [S11]; 2 years at three [S5] [S7] [S10]; 3 years, unlimited, at the expense carrier [S13]. No retrieved document contains a 1年に1回 cancer diagnosis benefit — the only annual cycles found are on a heart/stroke rider [S7] and on a survival income benefit [S11], so a one-year cycle is unverified and outside the sourced range. The composite takes 2 years with no lifetime cap: it is the modal repeating design, it is the shape that makes the diagnosis benefit a recurring liability rather than a single event, and a model that can express a 2-year cycle can express a once-only one by setting the cycle beyond the projection horizon. The once-only designs also differ in what payment does: at one carrier the A-type contract terminates on payment while C/D-type contracts instead waive all future premiums [S11]; at another the policy simply continues with the other benefits live [S1]. The composite continues and waives (footnote 8) — termination on the first diagnosis would delete the treatment, inpatient and outpatient benefits precisely when they are incurred. The composite additionally does not condition the second and later payments on the insured being under treatment: two of the three 2-year designs do so, requiring a hospitalisation [S10] or inpatient-or-prescribed-outpatient treatment [S7], while the third requires only a fresh 再発／転移／新生 [S5]. The unconditioned form is taken as the base and the treatment condition is carried as a switch, because the two differ by the whole conditional probability of being in treatment at the cycle date.
The 2-year clock is measured from a different event at every carrier that runs one, and a model that treats them as identical is wrong. From the date the previous payment trigger occurred [S5]; from the first day of the calendar month in which the previous payment was made [S7]; from the start date of the last hospitalisation for which the benefit was paid [S10]. The composite takes the previous trigger date — the only one of the three that does not require a second state variable (a payment date, or an admission date) to be carried alongside the trigger. The deeming rule that converts a continuing hospitalisation at cycle expiry into a fresh trigger [S1] is adopted because without it the benefit is unpayable exactly for the insured whose treatment never stopped.
No day limit is the invariant of this product, and it is stated as such by every source that speaks to it: 「がん入院給付金は給付日数無制限です」 [S1]; 「支払日数に制限はありません」 [S3]; a limits column that states a 支払日数 cap for every other benefit and none for this one [S5]; no cap in the benefit article [S6] [S10]; unlimited on the expense product [S13]; and, from the independent consumer-education source, 「入院給付金の支払日数は無制限です」 stated as the product-type norm R11. Against
medical’s 60/120-day per-hospitalization cap and 1,095-day aggregate, this removes both limits and the benefit-driven termination decrement they create — a cancer contract cannot exhaust itself. Two caveats are recorded rather than smoothed: at one carrier the absence is read from the 契約概要 limits column and not from the benefit article, because the 約款 body did not extract [S5]; and at the two treatment-centred carriers there is no inpatient benefit in the main contract to limit, it being a rider [S10] or absent [S7]. The composite therefore has noL1and noLAon the inpatient limb, and the one-hospitalization test that dominatesmedicaldoes not arise at all — a simplification, and the reason a cancer model is not a re-parameterization of a medical model.Observed: 20 × 基本給付金額, unlimited count, simultaneous procedures counted as one [S1] [S2]; がん入院給付金日額 × 20 on the same counting rule [S6]; unlimited count but certain procedures (fibrescopic malignant-neoplasm surgery among them) limited to one payment per 60 days, with some procedures outside cover altogether [S5]; 手術治療給付金 at the rider amount, limited to once per 14 days for a 一連の手術 [S7]; and per-surgery payment with fee-schedule de-duplication, where a repeated identical procedure scored once in the schedule pays once [S10]. The composite takes 20× with no count limit, the value two carriers publish at the composite’s own base amount [S2] [S6]; the 14-day and 60-day re-payment bars are carried as switches, since they bind only on repeated procedures and the retrieved documents do not agree on which procedures they apply to.
This is the benefit the market has moved to, and the one whose unit of payment is easiest to get wrong: it is a month, not an event and not a day. Observed: 治療給付金額 per month in which the trigger occurs, lifetime cap 60 months [S5]; 特約給付金額 × a 倍率 of 1 for hormone therapy in breast or prostate cancer and 2 otherwise, with a combined lifetime cap of 120× across chemotherapy and hormone therapy, counted across renewals [S7]; 給付月額 per qualifying month, where a prescription covering two months still pays one month and two triggers on one day pay once [S10]; and ¥100,000 per treatment month, or ¥50,000 for a month of hormone therapy only, the hormone-only months capped at 60 months [S11]. The composite takes ¥100,000 per qualifying month with a 60-month lifetime cap. ¥100,000 is the only published monthly amount [S11] and equals 10 × the composite’s base amount. 60 months is chosen because it is the only absolute cap two carriers state in months [S5] [S11] and because the third carrier’s 120× cap is exhausted by 60 qualifying months at 倍率 2 — the non-hormone case [S7] — so the same number is the effective ceiling under two of the three designs. Radiotherapy is folded into the same monthly benefit rather than paid separately: two carriers pay it that way [S10] [S11], against a per-treatment payment limited to once per 60 days at a third [S7]. Folding it in is what makes the benefit a single monthly indicator variable, which is the whole computational advantage of the design; the separate per-treatment form is a switch. The 倍率-1 hormone-only reduction is not adopted, because only one carrier grades that way in months [S11] and one in 倍率 [S7], and averaging a 50% amount reduction with a 50% multiplier reduction would produce a figure neither carrier writes.
Two structurally different outpatient designs coexist, and one carrier runs both at once. Treatment-linked, no day limit — attendance for surgery, radiation, thermal therapy or non-oral chemotherapy, with 「給付の日数の限度はありません」 [S1] and 支払日数は無制限 with no 通算 limit at all [S7]. Hospitalisation-linked and day-limited — attendance in a window around an inpatient stay: 60 days before admission and 180 days after discharge, capped at 45 days per hospitalisation and 730 days lifetime [S5]; 60 days before admission and 365 days after discharge, capped at 120 days per post-discharge period [S10]; a 通院治療期間 of one year from discharge capped at 60 days [S1] [S2]; a 通院期間 of 365 days from a defined start date with unlimited days inside it [S7]. The composite takes the treatment-linked, unlimited-day form: it is the design that does not require the inpatient-stay state the composite has otherwise discarded (footnote 14), and it is the form two carriers write [S1] [S7]. Where a carrier runs both, the tie-break is contractual and specified — a day qualifying under both is paid as the hospitalisation-linked one until that limit is exhausted, then as the treatment-linked one [S1]; the composite has one stream and no tie-break.
An honest gap. The 免責事由 lists sit in the 約款 halves that did not extract at both carriers whose booklets were retrieved in full form [S1] [S5], and no other retrieved document reproduces them. They are therefore unverified for this product line, and the composite carries no exclusion set beyond the waiting-period invalidity rule and the non-disclosure remedy. The one exclusion whose absence is not a gap is suicide: 保険法 第51条 carries no statutory time limit and a 免責期間 is purely contractual REG-R34, but the composite has no death benefit, so the clause has nothing to attach to. It becomes live only if the 死亡保険金 design of the minority carrier [S6] is switched on.
Options — 特約 and 特則#
Parameter |
Representative value |
Basis |
|---|---|---|
がん先進医療特約 |
The 技術料 of a 厚生労働大臣-designated 先進医療 reimbursed in full, lifetime aggregate ¥20,000,000; the rider terminates on reaching the cap |
[S1] [S7] [S11]; std (19) |
先進医療 cash top-up |
10% of the 先進医療給付金, capped at ¥500,000 per 療養 |
[S1]; std (19) |
患者申出療養 |
Excluded from the composite’s advanced-medicine benefit |
[S11] vs [S7]; std (19) |
がん退院一時金 |
10 × 基本給付金額 = ¥100,000 on discharge from a covered stay of 10 or more consecutive days; unlimited count, but not payable for a stay beginning within 30 days of a discharge already paid; on discharge by death the policyholder becomes the payee |
[S1] [S2]; std (20) |
がん収入サポート給付金 |
50% of the diagnosis lump sum on each policy anniversary after the diagnosis year while the insured is alive, up to 5 payments; stops on death |
[S11]; scope std (21) |
特定保険外診療給付金 |
Out of scope |
[S7]; scope std (21) |
緩和療養給付金, 外見ケア給付金, 女性がん特約, がんゲノムプロファイリング検査給付金, がん要精検後精密検査給付金 |
Out of scope |
[S7] [S11]; scope std (21) |
Expense-reimbursement (実損てん補) chassis |
Out of scope |
[S13]; scope std (21) |
The ¥20,000,000 lifetime cap is uniform at every carrier that publishes one for a current product [S1] [S7] [S11] — as it is on the
medicalchassis — against ¥10,000,000 on the 2013-edition contract [S5], which is the same parameter at an earlier vintage rather than a genuine alternative. Reaching the cap extinguishes the rider [S1]. The cash top-up is not uniform: 10% of the benefit capped at ¥500,000 per 療養 [S1] against ¥150,000 per 療養, once per policy year [S7]. The composite takes the former, which is the form that scales with the underlying claim. 患者申出療養 is expressly excluded at one carrier [S11] and expressly included at another [S7]; the composite excludes it, matching themedicalchassis so that the two products’ advanced-medicine streams are configured identically.The discharge lump sum is specified at exactly one carrier in the retrieved set [S1] [S2] and is retained because its shape — a lump sum conditioned on stay length, with a 30-day re-payment bar — is the one benefit in this product that reads back onto the length-of-stay distribution, and 平均在院日数 for 悪性新生物 is 14.4 days R7 REG-R27, close enough to the 10-day threshold that the benefit’s value is sensitive to the distribution rather than to its mean. It is off in the base run.
Out-of-scope items are named rather than dropped, because each is a real benefit some insured actually holds. The income-style benefit [S11] is the closest thing in the market to a cancer income stream and would be modelled as an annuity-certain contingent on survival, which is the
income_guaranteemachinery, not this product’s; 特定保険外診療給付金 — per qualifying month for treatment at a がん診療連携拠点病院等 by a procedure not in the public fee schedule, once per month, 12 payments lifetime — is the 定額給付 chassis’s answer to 自由診療 [S7] and is the point at which the two market chassis meet, but it needs a non-scheduled-treatment incidence basis that no retrieved source supplies. The expense chassis [S13] is out of scope entirely: indemnifying actual cost, with ガン入院保険金 unlimited, ガン外来保険金 capped at ¥20,000,000 resetting at each 5-year renewal, and a ¥1,000,000 diagnosis benefit on a 3-year cycle, it is a different model — severity distributions of cost rather than fixed amounts — and it is written under a non-life licence REG-R1.
Termination and values#
Parameter |
Representative value |
Basis |
|---|---|---|
Surrender value (kaiyaku-henreikin, 解約返戻金) |
None, at any duration, under 終身払 |
[S7] [S10] [S11]; std (22) |
Policyholder dividend (配当金) |
None — 無配当; no 満期保険金 |
[S5] [S6] [S11]; REG-R9 |
契約者貸付 / 自動振替貸付 |
Neither offered — there is no surrender value to lend against, so a missed premium lapses the policy |
[S1] [S7] [S10] [S11]; std (22) |
Rider values |
Riders carry no surrender value at any point |
[S1] [S5] |
Grace (払込猶予期間), 月払 |
From the first day of the month following the 払込期月 to the last day of that month |
[S1] [S6] [S8]; std (23) |
払込猶予期間, 半年払・年払 |
From the first day of the following month to the 月単位の契約応当日 in the month after that, with the stated month-end and February/June/November extensions |
[S1] [S6] |
Claims during grace |
Unpaid premium deducted from the benefit; for a monthly contract where the event falls on or after the contract anniversary inside the window, two months’ premium is deducted; if the benefit cannot cover the arrears the contract lapses at the end of grace |
[S1] [S6] |
Lapse (失効) |
From the day after the grace period expires |
[S1] [S6] [S11] |
Reinstatement (復活) |
Within 1 year of 失効, on payment of arrears and fresh 告知; may be refused on health grounds; the waiting period restarts from the 復活日 |
[S1] [S8]; std (24) |
First-premium failure |
The contract is void (無効), not lapsed |
[S1] [S8] |
Non-disclosure (告知義務違反) |
Rescission within 2 years of the 責任開始日, 復活日 or 特約中途付加日; 詐欺 and unlawful-purpose voidness are not time-barred |
[S1] [S6]; ceiling REG-R35 |
Benefit reduction (減額) |
Available where premiums become unaffordable, subject to an insurer-set minimum; increases require a new contract with fresh underwriting; where the daily amount is reduced mid-stay the benefit is computed at the amount in force on each day |
[S1] [S6] [S12] |
クーリング・オフ |
15 days [S1] / 8 days [S11] from the application date; out of scope |
|
Death of the insured |
Cover terminates; no death benefit in the composite |
[S1] |
Four positions on surrender value, and the spread is the widest structural variation in the product. None at all — 無解約払戻金型 — at three carriers [S7] [S10] [S11] and on the riders of every contract examined [S1] [S5]. Suppressed then a fixed step: no value ever under 終身払, and under a 短期払 no value during the premium-paying period and 10 × 基本給付金額 afterwards, which — since the product pays no death benefit — is also what is returned on death after the premium term [S1]; or no value during the premium-paying period and afterwards the lesser of 30% of the unsuppressed value and the greater of 入院給付金日額 × 10 and 診断給付金額 ÷ 10 [S5]. Conventional value throughout, computed from months paid and months elapsed, at exactly one carrier [S6] — and it is the only retrieved cancer contract carrying the full savings machinery: 契約者貸付, 自動振替貸付 at ≤ 8% p.a. compounded into principal at each anniversary and continuing until principal plus interest would exceed the surrender value, with one month’s notice before the 失効予定日 and retrospective cancellation if surrender or 払済 is requested within 3 months of grace expiry, and 払済保険への変更 setting a paid-up がん入院給付金日額 from the surrender value capped at the pre-conversion 日額 [S6]. The composite takes no surrender value: three carriers write the contract that way outright and a fourth gives zero under the 終身払 default, so it is the majority position at the composite’s own specification. The consequence is the one to carry into the model — the lapse decrement has no cash-value offset and there is no 自動振替貸付 to delay it. A lapse is a pure release of future premium and future benefit, and a missed premium lapses the policy at the end of grace. On the one contract with a value, a model that lapses immediately is wrong, because the APL carries the policy; that is a switch, not the base.
Three carriers state the same one-month monthly grace [S1] [S6] [S8]; one direct writer runs two months — 「払込期月の翌月初日から翌々月末日まで」 [S11]. The composite takes one month, the majority and the same value as the
medicalchassis. The first-premium rule is separately defined and materially different in kind: the payment period runs from the 責任開始日 to the end of the following month, then grace to the end of the month after that, and failure makes the contract void, not lapsed [S1] [S8] — so a first-premium failure produces no in-force policy at all rather than a lapse decrement.Three regimes, and they decide whether lapse is absorbing. 1 year from 失効 with fresh 告知 [S1] [S8], with one of those carriers additionally refusing reinstatement outright once any cancer benefit has been paid and after a surrender request [S1]; 3 years, with 延滞保険料 payable and no new policy document issued [S6]; and none at all — 「失効した保険契約を元の状態に戻すこと（復活）はできません」 [S11]. The composite takes 1 year, the modal published window and the same as
medical. What makes reinstatement more than a persistency detail on this product is that the waiting period re-runs: がん責任開始日 becomes the 復活日, unless the 復活日 precedes the original date [S1] [S6]. A reinstated cancer policy is therefore not the policy that lapsed — it is a policy with 90 days of no cover in front of it, which is a real anti-selection control and a real modelling state.保険業法 第309条 sets an eight-day window from the later of delivery of the disclosure document and the application date, effective on dispatch REG-R36; one carrier contracts for 15 days with a full refund and no interest [S1], another for the statutory 8 [S11].
jplibmodels from the point cover is in force and scopes the window out — it is a pre-inception decrement and modelling it would need a new-business funnel this library does not have.
Contractual mechanics#
The waiting period, and why it is not an exclusion#
Write d0 for the 責任開始日. Cancer cover attaches on
がん責任開始日 = d0 + 90 days (i.e. the 91st day counting d0 as day 1)
and on a monthly grid, with t in months from issue, that is the boundary t = 3
std. Nothing turns on the 90-day/3-month distinction at monthly resolution, and the
model does not claim otherwise; the two contractual wordings [S1] [S11] are recorded in
footnote 9 so that a daily implementation can separate them.
The rule that matters is what happens to a diagnosis inside the window. It does not simply go unpaid: the contract is void [S1] [S5] [S10], premiums are refunded where neither party knew of the diagnosis before 告知 and retained where either did [S1], and — under the composite — the treatment is not applied at all if no benefit event occurs within 5 years of the がん責任開始日 [S1]. For a projection this is a de-recognition, not a decrement: the policy was never in force, so it releases premium already collected as well as future benefit, and it belongs in a validity adjustment at outset rather than in the lapse column. The diagnosis date that tests it is the date the diagnostic test was performed, not the date the physician communicated the result [S1] — a distinction that also moves a diagnosis across a repeat-cycle boundary.
On 復活 the clock restarts from the 復活日 unless the 復活日 falls before the original がん責任開始日 [S1] [S6].
The diagnosis benefit and its cycle#
Let A be the 基本給付金額 (¥10,000 at the anchor cell), DB = 100 × A the diagnosis lump sum,
and T_last the date of the most recent payment trigger. The benefit is payable when
the insured is first diagnosed with an 悪性新生物 on or after がん責任開始日 [first payment]
or 再発 / 転移 / 新たに生じた cancer is diagnosed and now - T_last >= 2 years [repeat]
with no limit on the number of payments [S5]. T_last is set to the trigger date, not the
payment date [S5], and if the insured is still an inpatient on the day after the two-year
period expires, that day is deemed a fresh trigger [S1].
For 上皮内新生物 the benefit is 0.50 × DB, payable once over the policy term, on its own
cap, and not payable once a full-rate cancer benefit has been paid [S6] [S11]. In-situ pays
in full on every other benefit and does not trigger the premium waiver [S7] [S10].
The cycle is a state variable with its own memory, and it is not the 180-day
one-hospitalization memory of the medical chassis: it is keyed to a diagnosis event, it
runs for two years rather than 180 days, and it is unaffected by admissions and discharges
except through the deeming rule. A model that reuses one clock for both is wrong.
The inpatient benefit — no limits#
hosp_benefit(i) = D * stay_days(i)
for every stay i whose direct purpose is cancer treatment, with D the がん入院給付金日額 [S1]
[S5] [S6] [S10]. There is no L1 and no LA [S1] [S3] R11. Days adjacent to the cancer
treatment count in where the insurer accepts the stay was for that purpose — days after a
non-cancer illness begins during a covered stay, and days before the diagnosis date in a
stay begun for another reason [S6].
Two consequences follow and both are deltas against medical. The 180-day
one-hospitalization test does not arise, so successive admissions need no grouping logic.
And the contract cannot exhaust itself: medical’s benefit-driven termination
decrement, which fires when both aggregate limbs are used up, has no analogue here. Cover
runs to the terminal age of the mortality table, and the only decrements are death, lapse
and — on the waiting-period rule — invalidity.
Surgery, and the monthly treatment benefit#
surg_benefit = 20 * A per surgery, unlimited count,
simultaneous procedures counted as one [S2] [S6]
treat_benefit(m) = 10 * A for each calendar month m in which at least one
qualifying chemotherapy, hormone-therapy or
radiotherapy treatment occurred,
subject to sum over m of 1 <= 60 months
The monthly benefit’s unit of payment is the calendar month, and every retrieved contract that writes one says so in a different way that means the same thing: a prescription covering two months still pays one month, two triggers in one month pay once, two triggers on one day pay once [S10]; several treatments in a month pay once [S11]. The 60-month lifetime cap is std (footnote 16). On a monthly projection grid the benefit is therefore an indicator variable per period, not a count and not a duration — which is why the treatment-centred design is computationally simpler than the inpatient one, and why the market moved to it.
The covered set is defined by reference to public classifications rather than a private list
of named drugs — 日本標準商品分類 「8742 腫瘍用薬」 with oral administration excluded [S1], or whatever
attracts a 薬剤料 or 処方せん料 under the 医科・歯科診療報酬点数表 [S5] [S10]. As on the medical chassis, that
makes the trigger enumerable from a public document, and every such contract carries a
clause letting the insurer change the 支払事由 prospectively, with 主務官庁 approval and two months’
notice, if the public scheme changes [S1] [S5].
The outpatient benefit#
outp_benefit = D * days of attendance FOR surgery, radiation, thermal therapy
or non-oral chemotherapy, no day limit, no 通算 limit [S1] [S7]
The trigger is the treatment, not proximity to a hospital stay, so the benefit needs no admission or discharge date. A day of attendance during a stay for which the inpatient benefit is paid is not payable [S7] [S10].
先進医療#
adv_paid(k) = min( tech_fee(k), 20,000,000 - adv_paid_cumulative )
adv_top_up(k) = min( 0.10 * adv_paid(k), 500,000 )
and the rider terminates when the cumulative amount reaches ¥20,000,000 [S1] [S7] [S11]. 患者申出療養 is excluded [S11].
Grace, lapse and reinstatement#
For a 月払 contract, grace runs from the first day of the month following the 払込期月 to the last day of that month [S1] [S6] [S8]. A claim arising inside the window is paid net of the unpaid premium, and where the event falls on or after the contract anniversary inside the window two months’ premium is deducted [S1]; if the benefit cannot cover the arrears the contract lapses at the end of grace [S6]. 失効 runs from the day after grace expires.
There is nothing to break the fall. On a product with a 解約返戻金 the insurer would advance
the premium under 自動振替貸付, an election the 監督指針 requires be at the policyholder’s choice and
notified promptly REG-R14; the composite has no surrender value, so neither 契約者貸付 nor
自動振替貸付 can operate [S1] [S7] [S10] [S11]. The lapse is real and immediate — the same
position as the medical chassis, and the opposite of jplib’s savings chassis. On the one
retrieved cancer contract that does carry a value, the APL runs at ≤ 8% p.a. compounded
annually until principal plus interest would exceed the surrender value [S6], and a model
that lapses that contract on a missed premium is wrong.
復活 within one year restores cover from the 復活日 on payment of arrears and fresh 告知 [S1] [S8],
and the waiting period re-runs (above). One carrier refuses reinstatement outright once any
cancer benefit has been paid [S1] — a control with no analogue on the medical chassis, and
one that makes the reinstatement option conditional on the claim history the model is
already carrying.
Riders and options#
In scope (modeled or parameterized):
がん診断一時金 repeat cycle — 2 years, unconditioned, no lifetime cap; the under-treatment condition [S7] and the hospitalisation condition [S10] are switches, and a once-only design is the same cycle set beyond the horizon [S1] [S6] [S11].
上皮内新生物 grading — 50% of the diagnosis benefit, once, separately capped [S6] [S11]; the full-rate [S1] [S5] [S10] and 10% [S7] treatments are switches on the same parameter.
がん治療給付金 (monthly) — ¥100,000 per qualifying month, 60-month lifetime cap; the radiotherapy limb is folded in [S10] [S11], and paying it separately per treatment on a 60-day lockout [S7] is a switch.
がん通院給付金 — treatment-linked, unlimited days [S1] [S7]; the hospitalisation-linked, day-limited designs [S5] [S10] are switches.
がん先進医療特約 — a reimbursement stream with a ¥20,000,000 lifetime cap and a 10% capped top-up, the rider terminating on the cap [S1] [S7] [S11]. Switchable off.
がん退院一時金 — ¥100,000 on discharge from a stay of 10+ days, with the 30-day re-payment bar [S1] [S2]. Off in the base run.
保険料払込免除 — cancer-diagnosis trigger excluding in-situ [S10] [S11]; the disability-only trigger [S1] [S5] [S6] is a switch.
10年更新 定期 chassis flag — automatic renewal with benefit history, waiver history and 責任開始期 continuous across the renewal, and an option exercisable at least 2 months before maturity to convert to 終身 [S5] [S7].
Out of scope: the 実損てん補 expense chassis and its 自由診療 cover [S13]; 特定保険外診療給付金 [S7]; がん収入サポート給付金 [S11]; 緩和療養給付金 (24-month [S7] and 12-month [S10] caps); 外見ケア給付金 [S7]; 女性がん特約 and 女性手術給付金 [S7] [S11]; がんゲノムプロファイリング検査給付金 and がん要精検後精密検査給付金 [S7]; the 悪性新生物 death benefit floored at the policy reserve [S6]; 払済保険への変更, 契約者貸付 and 自動振替貸付 [S6]; the 特定障害不担保特約 underwriting carve-out [S1]; 重大疾病一時金特約 (heart and stroke, not cancer) [S7]; 減額 and 中途付加 alterations [S1] [S12]; and クーリング・オフ REG-R36.
Variations across insurers#
Chassis and benefit menu. Diagnosis + inpatient + surgery at three carriers [S1] [S5] [S6]; diagnosis + outpatient with treatment benefits in riders at a fourth [S7]; a treatment-only main contract with nine riders at a fifth [S10]; diagnosis + a single monthly treatment benefit at a sixth [S11]; expense reimbursement at the seventh [S13]. Composite: all five 定額給付 benefit types in one main contract, independently switchable, so that any of the first six shapes is a configuration (footnote 1).
Waiting period wording. 90 days with cover from day 91 at five carriers [S1] [S5] [S6] [S10] [S13]; three calendar months at two [S8] [S11]. Composite: 90 days, which on a monthly grid is three months either way (footnote 9).
Premium during the waiting period. Payable at five carriers [S1] [S5] [S6] [S7] [S10]; not charged at all at one, which says explicitly that this is not a discount [S11]; not stated at the seventh [S13]. Composite: payable.
上皮内新生物. Full rate at four [S1] [S5] [S10] [S13]; half rate on a separate benefit at two [S6] [S11]; 10% at one [S7]. Composite: 50% on the diagnosis benefit, full elsewhere, no waiver trigger — the median treatment, and the one whose incidence is separately sourceable from the registry’s with/without-上皮内がん rows R5 (footnote 10).
Diagnosis-benefit repeat cycle. Once ever at three [S1] [S6] [S11]; 2 years at three [S5] [S7] [S10]; 3 years at one [S13]. No retrieved contract writes a 1-year cycle unverified. Composite: 2 years, no lifetime cap (footnote 12).
What the 2-year clock runs from. The previous trigger date [S5]; the first of the month of the previous payment [S7]; the start date of the last hospitalisation [S10]. Composite: the trigger date — the only one needing no second date carried alongside it (footnote 13).
Condition on the second and later diagnosis payments. None [S5]; being under inpatient or prescribed outpatient treatment [S7]; a hospitalisation for cancer treatment [S10]. Composite: none, with the condition as a switch. The two differ by the whole conditional probability of being in treatment at the cycle date.
Chemotherapy and radiotherapy benefit. Monthly with a 60-month cap [S5]; monthly at 倍率 1 or 2 with a combined 120× cap counted across renewals [S7]; monthly in the main contract, uncapped [S10]; ¥100,000 per month with ¥50,000 for hormone-only months capped at 60 months [S11]. Radiotherapy sits inside the same monthly benefit at two carriers [S10] [S11] but is paid separately per treatment, once per 60 days, at a third [S7]. Composite: ¥100,000 per qualifying month, 60-month cap, radiotherapy folded in (footnote 16).
Surgery limits. Unlimited count [S1] [S2] [S6]; unlimited but 1 per 60 days for named procedures [S5]; 1 per 14 days for a 一連の手術 [S7]; fee-schedule de-duplication [S10]. Composite: unlimited at 20 × the base amount, with the re-payment bars as switches.
Outpatient benefit. Treatment-linked with no day limit at two [S1] [S7]; hospitalisation-linked and day-limited at three, on caps of 45 days per stay and 730 lifetime [S5], 120 days per post-discharge period [S10], and 60 days per 通院治療期間 [S1]. Composite: treatment-linked, unlimited (footnote 17).
Premium waiver trigger. Disability only at three [S1] [S5] [S6]; cancer diagnosis as a rider excluding in-situ at one [S10]; cancer diagnosis built into the contract at one [S11]; none stated in the 契約概要 of a sixth [S7]. Composite: cancer diagnosis excluding in-situ — the correlated trigger (footnote 8).
Surrender value and 自動振替貸付. None at three [S7] [S10] [S11]; none under 終身払 with 10 × the base amount after a completed 短期払 at one [S1]; a two-part formula after the premium term at another [S5]; and a conventional value throughout with APL at ≤ 8% p.a., 契約者貸付 and 払済保険への変更 at exactly one [S6]. Composite: none — so the lapse decrement has no cash-value offset and nothing delays it (footnote 22).
復活. 1 year at two [S1] [S8], one of which also bars it once any cancer benefit has been paid [S1]; 3 years at one [S6]; none at all at one [S11]. Composite: 1 year. It decides whether lapse is absorbing, and on this product it also re-runs the 90-day waiting period (footnote 24).
Grace, monthly. One month at three [S1] [S6] [S8]; two months at one [S11]. Composite: one month. クーリング・オフ likewise differs — 15 days [S1] against the statutory 8 [S11] REG-R36 — and is out of scope either way.
Death benefit. Stated absent in terms at one carrier — 「この保険に死亡保険金はありません」 [S1] — and not found in any other retrieved contract except one, where it is がん入院給付金日額 × a chosen 倍率 floored at the policy reserve [S6]. Composite: absent.
What does not vary. Every retrieved contract (i) has a waiting period before cancer cover starts and treats a diagnosis inside it as voiding the contract rather than merely excluding the claim [S1] [S5] [S6] [S10] [S13]; (ii) dates the diagnosis to the examination, not the consultation [S1]; (iii) requires 診断確定 by a Japanese-qualified physician or dentist on histopathological findings [S1] [S6] [S10] [S11]; (iv) pays the inpatient benefit with no day limit wherever it has one at all [S1] [S3] [S5] [S6] [S10] [S13] R11; (v) counts two simultaneous surgeries as one [S1] [S6] [S10]; (vi) defines the treatment benefits by reference to a public classification and carries a clause allowing the 支払事由 to be changed prospectively, with regulatory approval, if the public medical insurance scheme changes [S1] [S5] [S10]; and (vii) is 無配当 wherever the dividend basis is stated [S5] [S6] [S11]. These are the invariant core of the composite, and (i), (iv) and (vi) are exactly the three facts a reader coming from the medical product specification (医療保険) must not carry over unchanged.
Regulatory context#
Classification. がん保険 is named by the regulator as one of the three 第三分野 lines, alongside 医療保険 and 介護保険 R3, writable under either a 生命保険業免許 or a 損害保険業免許 by 保険業法 第3条第4項第2号 and 第5項第2号 REG-R1. That is not a formality here: six of the seven carriers in this composite hold a life licence and the seventh, writing the same risk as an expense indemnity, holds a non-life one [S13].
Prudential — and the fact that shapes this whole document. The FSA says of 第三分野 business
that 「標準死亡率、参考純率といったスタンダードな指標が存在しておらず、公的なデータや各社の実績等から給付事由ごとその発生率を見込まざるを得ない」 R3. There is
no standard cancer-incidence table and no reference pure premium. Every insurer’s incidence
rate (kiken hasseiritsu, 危険発生率) is its own, sits in the unpublished 算出方法書 REG-R2,
and is set from public data plus that insurer’s experience. What the regulator supplies
instead of a table is a test: annual 事後検証 of the assumed incidence rates; a ストレステスト
each reporting period over a 10-year horizon comparing future benefit outgo on the
pricing rate P against outgo on 危険発生率A, covering 99% of incidence risk, and 危険発生率B,
covering 97.7%, with contingency reserve (kiken junbikin, 危険準備金) established where P
falls short; and a 負債十分性テスト R3. The 監督指針 requires the calculation be performed under
平成10年6月8日大蔵省告示第231号 R4 REG-R13, with the tests run in principle per 契約区分 sharing a
common 基礎率 under 保険業法施行規則 第69条 R4 REG-R8. The reserve itself is the standard policy
reserve (hyōjun sekinin-junbikin, 標準責任準備金) chain: 保険業法 第116条 requires the reserve and
delegates the method REG-R4, 施行規則 第68条 says which contracts are in scope REG-R7, and
平成8年大蔵省告示第48号 sets 平準純保険料式 with the table vintages and the standard valuation interest rate
(hyōjun riritsu, 標準利率) reset machinery R4 REG-R10. jplib implements the
sensitivity — a re-runnable incidence basis — and not the statutory stress, whose
own notification text was not retrieved REG-R13; the 99% / 97.7% levels quoted above come
from the FSA policy paper R3, not from the 告示.
The mortality basis, by contrast, is prescribed and public. For contracts written from
April 2018 the third-sector valuation mortality is 第三分野標準生命表2018 R1 R2 REG-R11
REG-R18, published in full and freely downloadable — the sharpest contrast with uklib,
where the CMI tables cannot be read at all without a subscription. It is deliberately the
lighter table: male q(30) is 0.00041 against 0.00068 on 生保標準生命表2018（死亡保険用）, and male
q(90) 0.11657 against 0.15760 REG-R18. That is the correct direction of conservatism for a
morbidity product — survival prolongs benefit payment, so death releases the liability —
and it is precisely why a best-estimate cancer model must not adopt the valuation table as its
mortality decrement without a stated std adjustment. The table is not best estimate in any
case: built on the 国民表 (第21回生命表, 2010), improved forward at 2.5% p.a. for five years then
1.0% for three, then loaded by a 数学的危険論 margin sized to hold adverse deviation to about
2.28% (2σ) on an assumed 1,000,000 lives per sex, floored at 70% and capped at 85% of
the unadjusted rate, and constructed on a nearest-birthday age (hoken-nenrei, 保険年齢) basis
while the contracts age on attained age with the fraction discarded (man-nenrei, 満年齢) R2
REG-R20. jplib cites these tables and does not ship them: the 日本アクチュアリー会’s site
terms prohibit reproduction and transmission without written consent REG-R21, so the
model’s mortality input is a std construction whose provenance column points at the IAJ
entries REG-R18 REG-R19, and the cancer-incidence input is a std table built from
全国がん登録 downloads carrying the attribution string those datasets require R5 REG-R29. The
IAJ acts as 指定法人 in this chain under 保険業法 第122条の2 REG-R23.
Prudential — solvency. From 31 March 2026 insurers are supervised on the
economic-value ESR (経済価値ベースのソルベンシー規制), a three-pillar regime valuing liabilities at 現在推計
plus MOCE, re-measured at each 基準日 on assumptions re-set then, with required capital
calibrated at 99.5% and early corrective action below an ESR of 100% REG-R15. It
supersedes the ソルベンシー・マージン比率 trigger at 200% REG-R17, and the two are not comparable.
jplib computes neither. What it owes both regimes is that its projections are re-runnable
on a re-set assumption basis at a stated 基準日 — which, for a third-sector product, is exactly
the capability the ストレステスト demands of the insurer anyway.
Conduct. 監督指針 II-4-2-2 fixes what the 契約締結前交付書面 must contain — 商品の仕組み, 保障の内容 with the main 支払事由 and 免責事由, 付加できる主な特約, 保険期間, 引受条件, 保険料, 配当金に関する事項 and 解約返戻金等の水準 in the 契約概要; and クーリング・オフ, 告知義務, 責任開始期, the main non-payment cases and 保険料の払込猶予期間・契約の失効・復活 in the 注意喚起情報 REG-R14. That list is, item for item, the source set this specification is built from. クーリング・オフ is the eight-day dispatch-rule right of 保険業法 第309条 REG-R36, contracted wider at one carrier [S1] and scoped out here. Non-disclosure remedies run under 保険法 第55条, whose ceiling is five years from inception with a one-month clock from discovery REG-R35; the two-year contractual windows observed [S1] [S6] narrow that in the policyholder’s favour and are permitted. On insurer failure, contracts are compensated up to 90% of the 責任準備金 at the failure date through the 生命保険契約者保護機構 REG-R40, the rate set by ordinance under 保険業法 第270条の3 REG-R41; one carrier’s booklet adds that the 基礎率 (予定利率, yotei riritsu, the assumed interest rate; 予定死亡率; 予定事業費率) may be changed on transfer and an 早期解約控除 may apply [S1].
Tax. A がん保険 premium falls in the 介護医療保険料控除 basket of the post-2012 three-basket 生命保険料控除, which covers contracts concluded on or after 2012-01-01 paying on 疾病 or 身体の傷害 under 医療費支払事由 R8. The income-tax deduction is the full premium up to ¥20,000; premium × 1/2 + ¥10,000 to ¥40,000; premium × 1/4 + ¥20,000 to ¥80,000; and a flat ¥40,000 above — capped at ¥40,000 per basket and ¥120,000 overall R9 REG-R43. The anchor cell pays ¥36,000 a year, sitting in the second band for a deduction of ¥28,000 = ¥36,000 × 1/2 + ¥10,000: real, and second-order. The 住民税 caps are not stated on the 国税庁 pages retrieved and are unverified R9. On the benefit side, one carrier states that its cancer treatment, first-diagnosis, inpatient, surgery, discharge, advanced-medicine and outpatient benefits are in principle non-taxable where the payee is the insured, a spouse, a lineal relative or a 生計を一にする親族 [S1]. Benefits are not modelled net of policyholder tax.
Professional standards. Every life insurer appoints an appointed actuary (hoken keirinin, 保険計理人) under 保険業法 第120条 REG-R5, who confirms at each 決算期 whether the 責任準備金 is accumulated on sound actuarial principles and submits an 意見書 under 第121条 REG-R6. The 日本アクチュアリー会 実務基準 turns that into the 1号収支分析 — a forward income-and-outgo analysis over 「少なくとも将来10年間」 by 区分経理 segment, with sufficiency judged over the first five years REG-R22 — which is the same ten-year horizon the third-sector ストレステスト uses R3, and the shape these projections take. Statutory reserving under J-GAAP, the ESR economic balance sheet and IFRS 17, which is voluntary in Japan REG-R47, are three measurement bases fed by one set of projected cash flows, which is why this library keeps product cash flows basis-agnostic. This chassis is 無配当, so the surplus-distribution methods of 施行規則 第30条の2 REG-R9 and the 三利源 framing that goes with them do not apply to it.