Leisure, Hospitality & Parks Dashboard.

Four names, four archetypes, one verdict. Oriental Land (4661.T), Round One (4680.T), Resorttrust (4681.T) and Kyoritsu Maintenance (9616.T) share a TSE bucket and almost nothing else — a licensed destination park, a lease-financed entertainment rollout, a membership annuity carrying a float, and a twin-engine accommodation operator. Reconstructed part by part against the 17 July close, all four weighted fair values land under the price: Resorttrust at −2.7%, Oriental Land at −25.7%, Kyoritsu at −33.2%, Round One at −36.5%. Weighted by market capitalisation across ¥5,453bn, the bucket asymmetry is −25.2%. That is not four independent verdicts converging. It is a bucket verdict.

The quality scores do not sort the same way. They run from 12.5 to 17.0 out of 25, a spread of 4.5 points, and the ordering is close to the inverse of the valuation ordering. The best asset in the bucket carries the second-worst asymmetry ; the worst asset does not carry the worst. Quality buys no valuation protection here. Resorttrust is the one point where the two do not contradict each other, and even there the reconstruction lands marginally below spot rather than above. The corollary matters for anyone reading the sector on a screen: there is no quality-versus-price arbitrage available in 05a, only one symmetrical file and three of decreasing modellability.

One inherited ranking has been inverted at its head. The market-performance work made Resorttrust the most actionable dislocation in the bucket on a headline free cash flow yield of 8.6%. Cellular instruction establishes that the same figure reads 4.79% on the core flow, once ¥13,690m of deposits received and ¥10,366m of released property inventory are stripped out — below the cost of equity. What looked like a mispricing was a name correctly priced on incomplete information. The base of the ranking survives intact: Kyoritsu was flagged as a growth trap and comes out at −33.2%, with a bull case itself 12.4% below spot and restated cash conversion of −69.5% in its record year.

What follows sits in three layers. The conversion mechanics and the shared inputs describe what the four have in common — which is less than the bucket label suggests and more than the four business models imply. The archetype map and the names section sort them. The mispriced reads and the structural watchlist track what each consensus is getting wrong and what would force a reframing.

The decisive structural variable across this bucket is not demand and not margin. It is conversion — the step at which demand becomes a return on capital and then becomes cash. Consolidated revenue is a decoy for all four. Resorttrust grew revenue to ¥263.0bn and reports free cash flow of ¥32,468m, of which ¥17,610m is core ; conversion is 60.4% of operating profit on the year and 78.0% over three, not the 93% the sector work assumed. Oriental Land converts 62% of operating profit in FY March 2026 against 90% two years earlier, with the ¥1,000bn capital programme compressing it further through FY2028. Round One reports ¥30.0bn of free cash flow and −¥0.3bn after ¥30.3bn of lease principal — a conversion of −1% against the roughly 105% the published figures imply. Kyoritsu has generated −¥31,045m of cumulative free cash flow over eleven years against ¥221,223m of capital expenditure, and the record ¥7,862m of FY March 2026 restates to −¥17,263m once the ¥25,125m inventory release backed by ¥23,488m of reclassified property is removed. Four operators, four conversion regimes, one arithmetic.

Segment dispersion inside each operator is the second story, and it is wide enough to make consolidated multiples unusable. Resorttrust is the extreme case: Membership earns a 26.7% operating margin and an 18.0% return on capital, Medical 14.8% and 9.1%, Hotel & Restaurant 5.1% and 3.1% on ¥182,192m of assets — 34.7% of the balance sheet returning some 290 basis points below the cost of capital. The 2,160 basis-point spread between best and worst prices ¥182bn of assets returning 3% at the rate of ¥142bn returning 18%. Kyoritsu carries a comparable inversion in incremental terms rather than levels: the hotels convert 25.5% of incremental revenue into profit, the dormitories 3.9%, a 21.6-point spread that measures how differently two engines inside one ticker pass through Japanese wage inflation. Oriental Land's Theme Park segment margin has gone 27.2% to 25.4% to 23.0% across three years while Hotel has risen to 31.0% — the hotel leg is currently cushioning the park erosion at group level, which is precisely why consolidated revenue reads as a record.

Unit economics complete the picture, and they are where the four separate most cleanly. Oriental Land raised park revenue per visitor 3.0% to ¥20,642 and saw operating profit per visitor fall 7.0% to ¥4,740, because the derived cost per visitor rose 6.4% — the incremental yen of guest spending now carries a negative operating contribution, on an attendance base capped roughly 15% below the FY2019 peak and not expected back above it before FY2030. Round One's mature Japanese facility produces something like ¥180m of operating profit on the available proxy while a recently opened American site produces far less, the right-of-use amortisation landing in full from opening day ; the Japanese base is paying for the American expansion and consolidated growth of 8.8% presents the transfer as validation. Kyoritsu's Dormy Inn raised its average daily rate 5.1% to ¥16,450 while occupancy rose 1.35 points — the only pricing power in the bucket demonstrated in the strict sense, and 67% of enterprise value against 34% of revenue.

The thread tying this together is that in 05a a consolidated multiple prices an average of businesses that are not comparable to each other, and that the average is systematically flattering. Three of the four carry a headline metric that inverts under instruction: an 8.83% free cash flow yield that reads 4.79%, a 9.0% headline free cash flow yield that reads negative after lease service, and a record free cash flow year that restates to a deficit. The fourth, Oriental Land, converts genuinely — and is the most expensive name in the bucket by a factor of two.

The first input is the one the four have most in common and the one least discussed: the discount rate. Every reconstruction in this bucket rests on a doctrinal cost of capital of roughly 6.0% for the three domestic names and 6.75% for Round One on its American exposure. Only one file has priced the sensitivity. Kyoritsu establishes that a direct capital-asset-pricing construction on certified parameters — a dated government bond yield, a 0.66 beta, a sourced Japanese equity risk premium — plausibly lands near 4.0%, that today's price is consistent with 4.93%, and that 130 basis points move the sign: base fair value is ¥2,010 at 6.25% and ¥3,311 at 5.00%. The same doctrinal rate underwrites the other three valuations. If it sits systematically 150 to 250 basis points above a certified construction, the bucket's −25.2% average asymmetry is a parameter artefact rather than a market fact. This is the single most important observation the cross-reading produces, and it is the reason a shared cost-of-capital build is prescribed ahead of any individual modelling cycle.

The second input is Japanese wage inflation, and its distinguishing feature here is asymmetric pass-through. Personnel is the structural cost line for all four, with no hedge, no lag and no downward flexibility: ¥85,226m at Resorttrust, 32.4% of revenue, is the mechanical cause of a 1.93x operating leverage that cost 760 basis points of margin on a 12.5% revenue decline. Only one operator has demonstrated pass-through in the strict sense — a price increase that cost no volume — and it is Kyoritsu's Dormy Inn at +5.1% on rate with occupancy up 1.35 points across all four quarters. Oriental Land has pricing power on revenue and none on margin: the composition of its 3.2% increase in spend per guest is diagnostic, with ticketing at +2.4% against food and beverage at +6.2%, growth migrating from the discretionary pre-entry decision toward captive in-park spend. Resorttrust reports tariff revisions on three lines and the test does not confirm them — revenue up 5.49%, personnel up 4.88%, headcount up 4.76%, cost per head up 0.11%. The margin held because unit labour cost did not move, which is a different fact. Round One is the exception by structure rather than by execution: as an anchor tenant generating mall footfall rather than borrowing it, rent is the one critical cost line in the bucket that is bargained rather than borne.

The third input is the set of obligations that sit outside the debt line. Round One carries ¥138.9bn of capitalised leases against ¥333.8bn of market capitalisation, a fixed claim that levers any move in enterprise value 1.40x onto the equity and takes return on capital from about 24% on own capital to 9.3% once restored. Kyoritsu pays ¥12,767M of rent a year — 51.4% of consolidated operating profit — against ¥119,594M of future minimum commitments that Japanese standards leave off the balance sheet entirely ; capitalise them and invested capital moves from ¥236,359M to ¥355,953M and the return falls from 7.36% to 5.5–5.7%, crossing below the cost of capital. Resorttrust carries the mirror image: ¥149,685m of member guarantee deposits, 28.5% of the balance sheet, interest-free, funding a ¥213.6bn estate on ¥35.2bn of gross debt. Invert the price-to-book under a 6.0% cost of equity and the market pays for a sustainable return on equity of 11.18% against 11.05% adjusted for a 3.0% charge on those deposits — thirteen basis points apart. The market is not ignoring the free financing. It is billing for it, and the open question is whether the bill is right. Note finally what this bucket does not carry: it is domestic by construction, immune to the translation illusion that dominates the rest of the coverage, and the only normalisations that bite are accounting ones.