# HAMAKYOREX independent model: assumptions and source lineage

Research cutoff: 5 November 2026. JPY millions, absolute shares, yen EPS. This is a new 9037 model; no previous 9037 executable was replaced. BEING's model remains a separate issuer's dated baseline. Executable: `earnings_model.py`; imported inputs: `actuals.json` and `assumptions.json`. `model_output.json` and `annual_cases.csv` are calculated outputs, not hand inputs. Management totals are a later comparator and are absent from imported inputs.

**Model certainty is limited by undisclosed contract economics and cash-cost behavior. The FY2028–2030 forecasts use constant pre-adoption lease economics, not an exact statutory EPS forecast under the new lease standard.** HAMAKYOREX states that it adopts the standard from the beginning of FY ending March 2028 and has not yet quantified the effect (annual securities report, company PDF p55/printed53). Existing operating rents remain in cash operating costs. Existing finance leases retain principal and interest schedules. Do not mix these OP/EV/return measures with post-adoption IFRS-style measures without a lease bridge.

## Source map and historical audit

| Source ID | Original and exact location | Uses |
|---|---|---|
| FY26 | [FY March2026 release, 12 May2026](https://www.hamakyorex.co.jp/ir/library/accounts/upload/tanshin_202603.pdf), PDF pp1–2,4,6–9,12,14–15 | Annual P&L, segment external sales/OP/D&A, cash capex, cash equivalents, shares, NCI, borrowings, exceptional gains/losses |
| H1 | [FY March2027 H1 release, 30 October2026](https://www.hamakyorex.co.jp/ir/library/accounts/upload/tanshin_202609.pdf), PDF pp1–2,4–6,8 | Current H1 actuals, September balance, customer starts, center count, declared interim dividend |
| YUHO | [FY March2026 annual securities report, EDINET S100YBXW, filed16 June2026](https://disclosure2dl.edinet-fsa.go.jp/searchdocument/pdf/S100YBXW.pdf). Company copy `houkoku55_4.pdf` has extra front matter; company PDF pp9–10/printed7–8; p23/printed21; p55/printed53; p62/printed59; p74/printed71 | Subsidiary ownership and Kinbutsu earnings; building investment pipeline; new lease standard; operating-lease commitments; borrowings/lease interest and maturities |
| DECK | [FY March2026 supplemental presentation](https://www.hamakyorex.co.jp/ir/library/presentation/upload/hosoku_shiryo_202603.pdf), pp19,24,26–32,39 | Labor and operating-cost mix, capital-investment scale, Kinbutsu strategy, operating discipline. Latest deck is FY2026, not current H1 |

Source originals, extraction and hashes are retained in the company input folder and its source manifest. Monetary extraction is rounded/truncated in the issuer's disclosures. FY26 segment external sales100,937+54,562=155,499 versus reported consolidated155,500; this JPY1m discrepancy is rounding. FY26 OP13,068+1,712−19=14,761. Current H1 OP6,933+1,196+30=8,159 versus consolidated8,160; use a JPY1m reconciliation, not a fabricated corporate expense. All disclosed OP adjustments are intersegment eliminations. No additional corporate-cost deduction is imposed.

The annual cash-flow statement reports depreciation6,857; segment depreciation3,534+3,106 plus central216 totals6,856 within rounding. Segment goodwill amortization87+92=179 is treated as additional noncash operating cost. The deck's D&A7,514 has a different scope and is **not** substituted. Annual cash purchases of tangible9,852 and intangible134 total9,986. Equipment investment14,687, capitalized additions13,666 and cash capex9,986 are different measures; none is silently treated as maintenance cash capex. No exact reconciliation of the deck's alternative D&A scope is claimed.

The latest H1 has no cash-flow statement; its half-year report is scheduled for10 November. September bank debt14,486+27,041=41,527 versus March24,810, while fixed assets and the consolidation perimeter grew. The model starts financing from the actual September balance. FY2027 cash-flow columns cover **October–March only**, and do not reconstruct or assert unknown H1 cash capex or acquisition consideration.

## Commercial drivers and cost architecture

**Recognition:** center services are recognized on performed handling work or elapsed service time; trucking on transport completion. Contract awards and centers are not revenue units. The center count rose194→218 and9 new customers/10 starts were reported in H1; acquired sites, different site sizes and multi-site clients prevent a clean fee-per-center calculation. No invented center fees, load rates, rental yield or relay-fee stream enters the model.

| Driver IDs / unit | Base and bear–bull bounds | Observed anchor, mechanism and base selection | Falsifier / affected outputs |
|---|---|---|---|
| H2_CENTER, H2_TRUCK / % versus observed seasonal revenue ratio | Base0%; bear−2%/−3%; bull+2%/+2% | H1 current sales scaled by FY26 H2/H1 **segment-specific** ratio. Actual seasonal cash costs separately preserve FY26 H2/H1 cost ratios, including weak Q4 margin. Current H1 starts and acquisitions are already in the base; no second M&A cohort is added. H2 incremental cost pressure0.2% base (−0.1% to+0.6%) | H2 contract losses, new starts, input prices or Q4 cost pattern differ; affects FY0 sales/OP and all starting bases |
| C_VOL, C_CHURN / annual same-book activity | Base0% physical proxy,1% lost/reduced work; bear−1%/2%, bull+0.5%/0.5% | Weak national freight volumes do not support a large physical-demand boom. Customer activity/contract retention are distinct from newly won work. No disclosed issuer retention statistic exists; bounds are house judgments | Existing customer exits, customer-volume decline, reduced outsourcing scope; center revenue and costs |
| C_NEW / new recognized revenue as fraction of prior-year center sales at prior rates | Base4.5%; bear2.5%, bull6.0%; rejection limit9% | Financial proxy, not a measured market share. FY26 center sales grew6.7% including M&A; H1 FY27 grew14.1% including M&A and customer starts. Base new-work revenue is below recent mixed-scope growth and does not assume future purchases. Incremental cash fixed capacity and startup costs are funded | New contract starts cannot sustain roughlyJPY5bn–6bn annual recognized additions, or mature customer revenue shrinks faster; affects capacity envelope, revenue, capex, OP |
| C_PRICE / annual realized price and mix | Base2%; bear1.2%, bull2.5% | Regulatory compliance raises client demand for coordination, but wages/subcontractors absorb much of repricing. Price is separate from activity and never added again as a margin plug. It is **not** an observed signed contract-wide price commitment | Renewal price recovery below costs; one-point annual miss is a principal sensitivity |
| T_VOL, T_PRICE / annual transport activity and realized rate | Base0%/2.2%; bear−1.5%/1.5%; bull+1%/2.7% | H1 trucking sales+5.5%, profit+63.8%, despite soft special-consolidated freight. Charter sales, terminal productivity and M&A contributed. Base assumes no aggregate volume boom and no continuing63.8% profit growth | Weak freight outweighs charter conversion, rate resistance, terminal losses; truck sales/OP |
| VARIABLE_FIXED / cash-cost shares | Center70% variable30% fixed; truck65%/35%; fixed-share sensitivity belongs in downside judgment | Honest cost-behavior proxy: reported labor52,361/total operating costs140,739≈37.2%; labor, subcontract transport, rent, fuel and site costs differ in flexibility.40% labor share of variable cost is a rounded mix proxy, not audited segment cost accounting. Cash fixed costs carry2% annual inflation and capacity for new work | Costs prove more rigid, temporary workers reclassified, rent escalation or staffing step changes; all margins |
| WAGE, OTHER_INPUT / annual variable cost inflation | Base3.5% wages,1.5% other; bear4%/1.8%; bull3.5%/1.5% | Labor scarcity supports continued wages. Repricing is offset by these cost paths. Other variable costs include fuel/subcontract fees as an aggregate; no duplicated surcharge benefit | Wages/subcontractor rates exceed these paths. Diesel disclosure40m annual EBIT perJPY1/liter is an independent gross scale before pass-through |
| C_PRODUCTIVITY, T_PRODUCTIVITY / annual variable cost reduction | Base1.4%/1.7%; bear0.8%/0.9%; bull1.8%/2.1% | Daily profitability management, terminal scheduling, charter utilization and warehouse automation support improvement. H1 center margin compressed12.45% from13.22%, so this is an assumption, not demonstrated current margin acceleration. Benefit applies once to variable unit costs | Automation fails to offset staffing, low-density routes or site ramps; OP/EPS sensitivity |
| STARTUP / one-time cash cost as % new revenue | Base2%, bear4%, bull1.5% | Recruitment, layout and ramp costs precede mature throughput; public cohort costs unavailable. Prior-year startup expense is removed before recurring costs roll forward | Cohorts remain loss-making; persistent losses need explicit revised cost, not automatic startup removal |

No particular future physical customer center has been assigned the new-revenue proxy. The named investment projects supply capacity and rationalization, but HQ and branch replacement do not automatically generate sales. The9% bound rejects an extreme unsupported growth case; it is not proof of physical maximum capacity. Rates, staffing, floor space utilization and contract margins remain material unanswered inputs.

## Funding, commissioning and accounting inputs

| Input | Value/path and range | Source/mechanism, linked effect and falsifier |
|---|---|---|
| Committed buildings | HQ2,680m, Narita4,000m, Atsugi1,622m, Komaki3,400m. Total11,702m excluding already paid HQ land1,100m | YUHO companyPDF23. CompletionSep2027/Mar2028/Sep2027/Nov2028. Baseline post-September cash reserve3,400m FY27H2,7,302m FY28,1,000m FY29 is a house spend schedule. Because unknown H1 spending may have prepaid part, reserving all March commitments again after September is conservative, not proof of cash still owed. No H1 profit is added twice. Updated project cash/progress falsifies schedule |
| Project depreciation |30-year building life; FY28 fractionsHQ0.5, Narita1/12, Atsugi0.5; KomakiFY29 4/12. Full periods afterward | Company's useful building lives15–31yrs anchor30. No D&A before commissioning. Generic additional equipment20-year composite life with25% nondepreciating land; sensitivity extra capex changes funding and D&A. Mixed additional investments need actual allocation when announced |
| Maintenance and discretionary growth | Annual maintenance6,000m then2% inflation, versus FY26 D&A6,857m; FY27H2 maintenance3,000m. Growth envelopes7,302/5,500/6,000m FY28/29/30 include named commitments. Bear/bull0.75×/1.35× applies only to discretionary balance | Maintenance/growth split is undisclosed and uncertain. Vehicles130/1,063m leased in FY27 fit within annual new finance leases2,100m; additional leases bound by observed annual principal2,109m. No cash capex double count of leased vehicles. Deteriorating replacement intensity or building overruns falsifies funding |
| Legacy D&A runoff | Nonlease7% annually; zero-runoff sensitivity; cash maintenance assets10yrs and discretionary growth composite20yrs. Finance-lease assets5yrs (4yr sensitivity), existing lease D&A proxy2,100m decays20% | Avoid holding all legacy D&A forever while adding replacements. Existing finance-lease D&A is an aggregate proxy anchored annual principal2,109m and mainly vehicles/software, not a separately disclosed depreciation amount. New finance-lease additions enter their own depreciation cohorts (including H2 FY27) and are not also cash capex. H2 additions receive one-quarter of annual D&A in that half-year, then full D&A; later full-year additions receive half-year D&A. No disclosed retirement schedule exists; the proxy has explicit uncertainty |
| Working capital | Receivables43.4days, payables17.5days, inventories0.5day of annual sales; other annual outflow200m | AnnualFY26 AR18,497/sales155,500=43.4days; AP7,474/sales=17.5days; inventory189≈0.44day. H2 moves actualSep balances to annual run-rate stock. Other accruals/tax cash timing remain200m proxy.10day AR stress shows funding requirement |
| Cash/debt | September deposits26,132 less unavailable-time-deposit proxy1,721=24,411; bank41,527; finance leases6,702 proxy |1,721 is March deposits minus equivalents, **assumed unchanged** in Sep; Sep restricted cash/leases undisclosed. Sensitivities unavailable deposits5,000, leases+5,000 plus annual principal+1,000 propagate. Bank repayments3,500/yr near observed FY26 repayments3,527, with refinancing/draw allowed when policy cash reserve15,000 breached. Reserve is an analyst liquidity assumption, no equity-price floor or guaranteed cash |
| Interest and other income | Bank1.5% base/bull,2% bear; lease0.259%; cash0.2%; other recurring750m/yr | FY26 loan rates≈1.1%; H1 interest241 and higher debt support1.5% case. Lease rate disclosed in YUHO debt schedule. Nonoperating750≈FY26 dividends165+net rent97+fees50+solar154+net miscellaneous281; cash interest separate. Exclude subsidy628 and routine/exceptional disposal gains. Higher borrowing incurs same-year half-period interest |
| Tax |33.5% base;35% bear,32.5% bull | H1 ETR34.23%; FY26 group31.04% affected by gains and mix. No automatic deferred tax windfall |
| NCI entity and residual pool | Kinbutsu standalone net367m at73% owned→99.09m FY26 NCI proxy; currentFY0 pool anchored H1 NCI574 plus prior-H2 pool551 grown4% base; future Kinbutsu net grows5% and residual pool4% | YUHO companyPDF9–10; groupFY26 NCI1,059. Remaining profit at SuperRex53.4%, HMK67.6%, Takatsuka70%, TokaiNyuka70%, Matsumoto93.4%, Keihan80% is undisclosed. Residual pool is an explicitly unresolved after-tax claim, not revenue×minority%. Kinbutsu subgroup net411 differs from standalone367; difference stays in residual pool. Greater minority attribution at fixed consolidated earnings is a sensitivity, not extra entity operating growth |
| Shares and dividends |74,050,222 shares outstandingSep; FY27 weighted74,043,772.88; FY28–30 unchanged; no new buyback or dilution. H2 cash dividend is declared interimJPY23/share; future cash30% parent profit; NCI payout28% | Current issued76,048,000 less treasury1,997,778; H1 weighted74,037,359. Actual split1:4Oct2024 already in comparatives. H2 dividend23 is declared/payableDec1, not imported full-year profit guidance. Future payout assumption near annual28.8%; timing normalized in full years, not board-approved dividend forecast |

## Earnings anchors and normalization

Reported FY ended31 March2026 EPS is **JPY144.83**, basic, split adjusted. Reported parent earnings10,720m; weighted shares74,013,556. The latest reported TTM parent numerator is10,720+5,084−4,831=10,973m. Daily weighted shares use365FY days less183old-H1 plus183new-H1:74,027,643.49; TTM basic EPS≈JPY148.23 for1 October2025–30 September2026. Do not sum quarterly EPS.

Signed FY26 normalized parent adjustments are calculated in `model_output.json/anchors/normalization_components`: remove fixed-asset special gain941, negative goodwill120, securities gain54 and other gain12; add exceptional losses126. Use observed group tax rate5,303/17,083 except assumed tax-free negative goodwill. Net adjustment−727.515m; estimated normalized EPS135.01. The941 gain is also in the parent-only P&L, confirming parent100% attribution. Remaining adjustments assume parent100%; a conservative bound applies46.6% to the gross absolute amounts of those subsidiary-uncertain adjustments, not to the parent-confirmed941 gain (aboutJPY1.59EPS). This is a bound, not a known minority claim or confidence interval. Company amounts truncate, so annual reported EPS and calculated EPS differ by underJPY0.01.

FY27 recurring H1 adds back settlement182×(1−2,945/8,604)≈119.70m, assumed parent100%. If46.6% of this after-tax loss belonged to minorities, the parent addback falls≈55.8m (aboutJPY0.75EPS); apply this only to the normalization comparison. Acquisition-related operating costs remain in H1 actual and are not generically removed. No future negative-goodwill or asset-sale gains are assumed.

The FY0→FY3 CAGR spans31 March2027→31 March2030, exactly3 years, both house estimates. FY26 actual→FY30 spans4 years. TTM midpoint/end-period comparison is a distinct scale comparison. Yen EPS bridge contributions show their percent of the named FY26 reported basicEPS144.83; they are model effects, not promised price returns.

## Weakest inputs and decision tests

The forecast earns its improvement only if price recovery, contract retention and variable productivity exceed wage/subcontract inflation after funding capacity. Same-book customer activity, gross wins/losses, realized center pricing, wage/subcontract cost escalation, new-site margins, NCI allocation, September capex/acquisition cash and lease adoption are the most useful follow-up questions. A high center count or regulatory deadline alone does not answer them. No future acquisition premium, acquisition profit, relay fee or government subsidy is required by the base case.