Sample Underwriting Memo — 1K income condo, Kitahama area, Osaka

This memo is independent research prepared from Japan's public data sources for the property identified below. It is not investment advice, not a licensed real-estate service (宅地建物取引業), not an appraisal (不動産鑑定評価), and not legal or tax advice. The preparer performs no brokerage, introduction, mediation, or negotiation (売買・交換・媒介・代理のいずれも行いません), holds no stake in any transaction involving this property, and is compensated by flat fee only — never by commission or success fee. If a memo falls short of what's promised, the preparer's liability is limited to a refund of the fee paid for that memo. Figures derived from public records carry the limitations stated in the Method appendix. The buyer is responsible for their own due diligence and for engaging licensed professionals (宅建士, 司法書士, tax adviser) before transacting.

1 · The property, as listed

Item	As disclosed (原文 — original text)	English
Location	大阪市中央区東高麗橋（番地非公開）	Higashi-Koraibashi area, Chuo-ku, Osaka — Kitahama business district, 6 min walk Kitahama stn (Sakaisuji line), 10 min Tenmabashi stn (walk minutes per the 表示規約 advertising standard of 80m/min; crossings and waits excluded)
Price 価格	1,550万円	¥15,500,000 — seller-direct (売主): no buyer-side brokerage fee
Advertised yield 表面利回り	5.26%	Gross yield, claimed
Income 想定年収	816,000円（月68,000円）	¥816,000/yr projected — labeled 想定 (projected); actual contract rent not stated
Unit	1K・24.44㎡・バルコニー3.00㎡・8階/13階建	1K, 24.44㎡ + 3.00㎡ balcony, 8th of 13 floors
Building	2005年2月築・RC造・総戸数46戸・所有権	Feb 2005 — 21 yrs 5 mo at capture; the listing rounds to 築22年 (discrepancy noted). Post-1981 seismic standard. RC, 46 units, freehold
Monthly fees	管理費7,080円・修繕積立金6,110円	Management fee (管理費) ¥7,080 + repair reserve (修繕積立金) ¥6,110 = ¥13,190/mo — 19.4% of the claimed rent
Management	全部委託（巡回）	Fully outsourced, patrol-type building management — association health (minutes, arrears) unverified; question 3
Tenancy	賃貸中（オーナーチェンジ）・引渡相談	Tenanted, owner-change sale; delivery by consultation

Derived: ¥634,000/㎡ price · claimed rent ¥2,782/㎡/mo · repair-reserve rate ¥250/㎡/mo — below the MLIT guideline average of ¥335/㎡·月 for this building class (under 20 storeys, estimated under 5,000㎡ total floor area); adequacy rests on the 長期修繕計画 (long-term repair plan) and fund balance, both undisclosed. Listing registered 2026-07-29 — the capture date; zero observable days on market, no price-cut history yet.

2 · Price check — the public transaction record

All recorded pre-owned condominium transactions, Chuo-ku, 2024Q1–2025Q4 (MLIT reinfolib, 2,921 records), filtered to the subject's class:

Comp set	n	¥/㎡ p25	median	p75	Subject ¥634K/㎡ sits at
Kitahama/Tenmabashi walk zone, 18–32㎡, built 1998–2012 (32 of 39 are 1K)	39	640,000	700,000	800,000	21st percentile
Ward-wide, same class (context)	306	633,333	733,333	840,000	26th percentile
REINS rows only — actual closed prices reported through the brokers' transaction network, the highest-fidelity slice of the record	8	560,000	563,333	640,000	75th percentile (only 8 sales — treat as a secondary check)

Finding (factual placement, both bases reported): the two slices of the public record disagree, and this memo reports both rather than the more favorable one. Against the broad survey record, the asking price sits at roughly the 20th–26th percentile — below the middle of that record. The REINS closed-price slice points the other way: its median (¥563K/㎡) puts the same asking price near its 75th percentile, about 12% above the closed-price median. Eight sales are too few to settle the question, but the two bases disagree by more than the placement margin — so the placement is bracketed, not "below market."

A separate fact to hold alongside it: the seller is a licensed dealer selling its own inventory. If the price genuinely sits below the survey median, the causes worth ruling out are exactly the undisclosed items — facing, interior condition, the actual contract rent, and reason for sale (questions 1, 5, and 6 below). A discount is a fact with unresolved causes until disclosures rule them out.

Five limitations of the public record, so you can weigh this finding appropriately: (1) the government database reports each sale's location only down to the district (町) — not the street or building — so these comparables are "same district, same class of unit," not "same building"; (2) recorded prices are rounded to two significant figures — a ¥15.5M sale appears as ¥16M — which blurs fine price distinctions; (3) sales are published roughly 4–6 months after they close, so the record trails today's market; (4) coverage is partial — the record comes from a voluntary survey capturing a fraction of all transfers; and (5) the data contains no adjustment for floor level, view, or interior condition — a renovated high-floor unit and a tired low-floor unit of the same size and age read identically. None of these change the placement finding; they are why this memo treats it as a placement check, not a valuation.

3 · Achievable rent — quoted as a range

Nine live rental listings in two 2007 15-story RC buildings bracket the subject's location: one two blocks away in the same Kitahama micro-district, one in southern Chuo-ku. Method: comparables within ±4㎡ / ±7 build-years (7 of 9 qualify); predictor = median comp ¥/㎡ × subject area. Rents exclude management fees where separable.

	Result
Predicted market rent (method central)	¥67,700/mo — the listing's projected ¥68,000 is +0.5% vs. this: consistent with market
Range (typical comp error, ±8.8%)	¥61,700 – ¥73,600/mo
Tail band (90th-percentile error)	¥55,200 – ¥80,200/mo
Micro-market bracket	Same-district building comps imply ~¥79,600; southern-ward comps ~¥67,200 — the method median sits at the conservative end

Why a range: this comp method was blind-tested on 450 listings in this exact segment (Osaka Chuo/Kita/Nishi 1R/1K). Typical error is small, but tail cases cannot be identified in advance from comp statistics — a single point estimate would overstate precision. Comps are asking rents; achieved rents typically settle somewhat lower.

4 · What the yield looks like after the full cost stack

All-in cost basis ≈ ¥15.9M (price + estimated registration, acquisition tax, stamp, 司法書士 judicial-scrivener fees; no brokerage — seller-direct). Assumptions stated: property management 5% of collected rent; turnover allowance = AD/leasing (広告料) 1.5 months combined — this absorbs the letting agent's fee — + ~¥80,000 restoration/refresh per re-let, amortized over an assumed 3-year tenancy (typical for the 1K segment); vacancy 1 month per 2 years; property tax estimated ¥45,000/yr (verify against 課税明細, the property-tax assessment notice); insurance ¥10,000/yr (fire only — earthquake cover is a separate optional premium, buyer decision).

Annual line	Low rent (¥61.7K)	Central (¥67.7K)	High (¥73.6K)
Gross rent	740,768	812,246	883,724
− Building fees (管理費+修繕積立金)	158,280	158,280	158,280
− Property management (5%)	37,038	40,612	44,186
− Turnover allowance (AD + restoration, amortized)	57,517	60,517	63,467
− Vacancy allowance	30,865	33,844	36,822
− Property tax (est.)	45,000	45,000	45,000
− Insurance (est.)	10,000	10,000	10,000
Net operating income	402,068	463,993	525,969
Net yield on all-in cost	2.53%	2.92%	3.31%

The central finding: the projected rent matches comps within half a percent, the price placement is bracketed (p20–26 of the survey record, p75 of the small closed-price slice — see §2), and the advertised 5.26% gross is arithmetically correct as labeled. After the full cost stack the listing does not model, the same deal nets approximately 2.5–3.3%. Building fees alone consume 19.4% of gross rent; turnover costs (AD plus restoration each re-let) take roughly another 0.4 points. None of this is a recommendation in either direction; it is the drag a buyer must price before wiring funds.

Two facts about the hold, stated without forecast: the rent comps are 2007-build stock, so over a typical hold the subject leaves the age band its rent evidence came from; and the repair-reserve rate (¥250/㎡/mo) is a current figure that sits below the MLIT guideline average of ¥335/㎡·月 for the class — reserve rates commonly step up when a building's 長期修繕計画 is revised. The fee line is a floor, not a constant. One reserve sensitivity, stated as arithmetic: at the guideline-average rate instead of the current ¥250, the net-yield range reads 2.37% / 2.76% / 3.15% — a −0.16-point shift. That is a benchmark substitution, not a forecast.

Thresholds for the seller's answer to question 1, same cost stack, so no reworking needed: actual contract rent of ¥65,000 → ≈2.74% net · ¥63,000 → ≈2.61% · ¥60,000 → ≈2.42%. Arithmetic only — what any figure means for the purchase is the buyer's call.

Vacancy sensitivity, same stack (net yield at low / central / high rent): half the assumed vacancy (0.5 mo per 2 yrs) → 2.63% / 3.02% / 3.42% · base (1 mo per 2 yrs) → 2.53% / 2.92% / 3.31% · double (2 mo per 2 yrs) → 2.33% / 2.71% / 3.08%. Benchmark substitutions, not forecasts.

Yield benchmark — where 5.26% sits in this segment: applying this memo's rent method across the recorded 2024–2025 transaction record for the three central wards (1,739 sold 1R/1K units scored, ≥3 rent comps each) gives an implied gross-yield distribution with a median of 4.42% (p25 3.90% / p75 5.39%). The subject's advertised 5.26% sits at the 72nd percentile of that distribution — and at the 60th percentile of age-matched Chuo-ku stock (built within ±7 years of the subject, n=268, median 4.93%). In plain terms: the advertised gross yield runs somewhat above what the segment's recorded prices and market rents typically imply — which also means typical segment stock, run through the same cost stack, nets less than this deal. Basis note: the subject's figure rests on asking price and the comp-consistent projected rent; the benchmark rests on recorded sold prices with method-predicted rents — if the subject transacts below asking, its yield rises further. No cheap/dear conclusion is drawn; this is where the advertised number sits.

5 · Red flags & checks

Severity labels: HIGH = affects the deal's core numbers until resolved · MEDIUM = material, resolvable by disclosure · INFO = factual consideration to plan around · LOW = minor · CLEAR = checked, nothing found.

Severity	Item	What resolves it
HIGH	Income is labeled 想定 (projected) on a tenanted unit — the deal's most load-bearing number is soft. Comps make ¥68,000 plausible; only the contract makes it actual.	Request the 賃貸借契約書 (tenancy contract)/rent roll: actual rent, contract type (普通借家 = standard renewable / 定期借家 = fixed-term), start date, tenure
MEDIUM	Repair-reserve fund balance and long-term repair plan undisclosed — and the monthly rate (¥250/㎡) sits below the MLIT guideline average of ¥335/㎡·月 for the building class, so the low fee load flatters the yield	修繕積立金 balance, arrears rate, 長期修繕計画 (long-term repair plan), any planned 一時金 (special assessments) — via the 重要事項調査報告書 (the management company's standard disclosure report; question 2)
MEDIUM	Exit-pool wedge: the unit's 24.44㎡ exclusive area sits below the ~25㎡ floor many Japanese lenders apply to residential mortgages — the future resale buyer pool is effectively investors (cash or investor loans), on a building that will be past 30 years old at a typical exit horizon	Buyer's own exit planning; factual constraint, no valuation implied
INFO	2005-build RC (21 yrs): a long amortization carries the building deep into age; resale-horizon consideration	Buyer's own financing/horizon analysis
INFO	Tenancy transfer: no サブリース (master-lease sublease) indication — owner-change with delivery by consultation is consistent with a direct tenancy — but the inherited management structure (集金代行 = rent-collection agency vs サブリース = sublease), PM-contract transfer, 敷金 (tenant security deposit) assignment at closing, and 保証会社 (rent-guarantor company) coverage of the sitting tenant are all undisclosed	Question 8; contract review with 司法書士 (judicial scrivener)/counsel
INFO	Overseas-landlord mechanics: non-resident owners require a 納税管理人 (tax agent); corporate tenants must withhold 20.42% of rent paid to a non-resident; confirm the property manager accepts overseas landlords. The sale-side mirror applies at exit: a buyer purchasing from a non-resident seller generally withholds 10.21% of the purchase price — a cash-flow-at-exit fact to plan for	Items to arrange with a tax adviser and the PM — flagged, not advised
INFO	Seller is a licensed real-estate dealer selling as principal (取引態様: 売主) — read from the Japanese original. Dealer-sold units typically carry statutory buyer protections not present in individual-seller transactions, and no buyer-side brokerage fee applies	Confirm the specific protection scope in the sale contract with 司法書士/counsel
INFO	Tenancy regime (借地借家法, the Land and Building Lease Act): if the sitting contract is 普通借家 (standard renewable), the tenancy is strongly protected — rent increases and non-renewal are difficult, and personal use is effectively unavailable while the tenant remains. No conflict with this memo's client thesis (income only) — stated because a buyer cannot infer it from the contract-type question alone	Question 1; legal specifics referred out
INFO	Short-term-rental (民泊) permissibility depends on the building's 管理規約 (bylaws) — not disclosed	Request the 管理規約 with question 3
LOW	Unit facing/direction (方角) not disclosed — the floor plan itself is provided as a listing image; reason for sale not stated	Ask items 5–6 below
LOW	Listing discrepancy: the listing's own 築22年 age label overstates — computed age at capture is 21 yrs 5 mo (rounding). All other listing arithmetic checks out (yield, fee, and area figures are internally consistent)	Noted; no action needed
CLEAR	Hazard maps: no flood or liquefaction flag at the subject block (MLIT hazard portal, checked 2026-07-29)	—
CLEAR	事故物件: no incident record for the subject building (Oshimaland, checked 2026-07-29). An adjacent building shows 2 incident pins — incidents in neighboring buildings attach no disclosure obligation to the subject and typical market impact is marginal; noted for awareness only, not a legal conclusion	—

6 · Questions to send the seller, as-is

Please provide the current tenancy contract details: actual monthly rent, contract type (普通借家 = standard renewable / 定期借家 = fixed-term), start date, and tenant tenure (賃貸借契約書 or rent roll).
Please provide the building's 重要事項調査報告書 from the 管理会社 (management company) — the standard disclosure containing the repair-reserve balance, arrears rate, current 長期修繕計画 (long-term repair plan), any special assessments levied or under discussion, and the management form. (This memo verifies the individual fields against it on receipt.)
Most recent management-association AGM minutes (総会議事録) — any disputes, planned fee increases, or deferred maintenance? Please also provide the 管理規約 (building bylaws), including any short-term-rental (民泊) restrictions.
Any rent arrears or late payments by the sitting tenant in the past 24 months?
Reason for sale (譲渡理由)?
Unit facing/direction, renovation history, and any 告知事項 (matters requiring disclosure — defects, incidents)?
Delivery is listed as "consultation" — what determines the timeline, and is the tenancy assigned as-is at closing?
What is the current management arrangement (集金代行 rent-collection agency, or サブリース master-lease)? Does the management contract transfer at closing, how is the tenant's 敷金 (security deposit) handled, and is the tenant covered by a 保証会社 (rent-guarantor company)?

7 · What this memo did not assess

Registry verification (登記事項証明書, the certified registry record: ownership matches seller · 抵当権 mortgages/encumbrances · 所有権 freehold vs 借地権 leasehold) — Deep tier, or buyer-obtained; legal title conclusions in any tier · buyer-specific tax position · building physical condition (no inspection) · financing availability for non-resident buyers · currency considerations · yield placement against competing current for-sale listings' advertised yields (asking basis; portal data member-gated at capture — the transaction-based yield benchmark in §4 covers the segment question) · any recommendation to enter or not enter the transaction.

Method appendix — every input auditable

Sources & pull dates: listing captured 2026-07-29 (major investment-property portal, seller-direct listing; 情報登録日 (listing registration date) 2026-07-29 per the listing record) · MLIT reinfolib Chuo-ku pre-owned-condo transactions 2024Q1–2025Q4, 2,921 records · 9 SUUMO rental listings (ward-level rental search, Chuo-ku, 1R/1K layout filter; captured as print PDFs 2026-07-29) · MLIT hazard portal & Oshimaland checked 2026-07-29 · MLIT マンションの修繕積立金に関するガイドライン (guideline average ¥335/㎡·月, buildings under 20 storeys / under 5,000㎡ floor area; floor-area class estimated from 46 units × 1K).
Comp rules (applied mechanically, set before analysis): price comps = walk-zone/ward, 18–32㎡, built 1998–2012; rent comps = ±4㎡/±7yr around the subject, widen once if fewer than 3 qualify (not needed: n=7); predictor = median comp rent/㎡ × subject area; rents exclude management fee where separable.
Calibration: the rent-comp method was blind-tested on 450 one-per-building listings in this segment (Osaka Chuo/Kita/Nishi 1R/1K): rent error MAE 8.8%, 90th-percentile 18.5% — the ranges in §3 are those measured errors, not judgment calls. The subject is inside the calibrated segment.
Yield benchmark (§4): computed from the same public datasets — the reinfolib sold record × the 450-listing rent pool, per-unit comp matching (same ward, 1R/1K class, ±4㎡/±7yr, widen once, ≥3 comps required; 1,739 of 1,790 sold units scored). The distribution median reproduces the calibration run's observed segment yield (4.42%) exactly.
Extraction QA: all listing and comp figures dual-pass verified against source captures; this memo's extraction error count: 0.
Known limitations: rent comps are asking (not achieved) rents; portal rent figures may fold in common-area fees (direction: makes the prediction conservative); the 9 rent comps come from 2 buildings (mitigated by their location bracket); transaction prices are district-grain, 2 significant figures, ~4–6 months lagged; the two price bases (survey vs REINS closed-price) disagree and are reported bracketed, not averaged.
Stated allowances (not measured values): turnover = AD 1.5 months + ¥80,000 restoration per re-let over a 3-year assumed tenancy; vacancy = 1 month per 2 years; PM = 5%. Net-yield sensitivity to each is linear — a buyer can substitute their own assumptions directly into the §4 table.
Memo standard version: 2026-07-30 · integrity: sha256 of the delivered file recorded at delivery · frozen-protocol calibration record retained, available on request.