The Conclusion — ¥85.07mn Outside the Price on a ¥1.2bn Hotel, or 1.42 Years of First-Year NOI

The figures come from this article’s model property — a limited-service hotel in a regional core city, 100 rooms, purchase price ¥1.2bn, LTV 70%. Every assumption is set out in “Calculation Assumptions” at the end of the article. Three structural points matter here.

First, the breakdown of acquisition costs is essentially decided by brokerage commission and transfer taxes. In the ¥1.2bn model, brokerage commission is 3.31% and transfer taxes — real estate acquisition tax plus registration and licence tax on the ownership transfer — are 2.69%. Those two alone come to 6.00%, which is 85% of the ¥85.07mn total. The line items that appear on quotations, such as due diligence and judicial scrivener fees, are small in absolute terms. A small amount does not mean a small consequence, though. Misreading the legal form of a hot spring source in a ryokan acquisition, for instance, moves the valuation itself — a structure covered in detail in Who Owns the Hot Spring? Water Rights and Ryokan Valuation.

Second, the erosion barely changes across price bands. Across the three cases discussed below — ¥300mn, ¥1.2bn and ¥3.0bn — the cost ratio lands at 7.66%, 7.09% and 6.95%, a spread of less than one point. The main line items scale with price; the difference comes only from the relatively heavier weight of fixed-ish items such as due diligence and scrivener fees on smaller deals. On the acquisition-cost side, “buying bigger makes you more efficient” is a limited effect.

Third, transfer taxes can be compressed through a trust beneficiary interest, but not for free. Acquire the trust beneficiary interest rather than the real estate itself and no acquisition tax applies, while registration and licence tax is just ¥1,000 per property for the change-of-beneficiary registration. In the ¥1.2bn model that is a ¥32.22mn saving. Against that, a trust fee runs for the whole holding period: at ¥4mn a year it consumes the saving in about eight years. Short hold favours the beneficiary interest; long hold favours direct ownership.

Terminology — Gross, NOI and Net Yield: What Has Been Subtracted from Each

The usual stumbling block in a first hotel investment review is that the single word “yield” refers to three or more different calculations. If a property memorandum says only “yield 6.5%”, comparing it without first establishing what has been deducted makes any real investment judgement impossible.

Of these, the one investors use as common ground is the NOI yield. The “expected yield” published in the Japan Real Estate Institute’s investor survey likewise refers to single-year net operating income divided by price. Comparing “a property at 6.5% gross” against “a market at 4.5% expected yield” is therefore meaningless. To put them on the same footing you first have to deduct operating expenses and move to an NOI basis.

And this article’s subject lies one step beyond that. The NOI yield adjusts the numerator for operating expenses, but the amount actually invested is not the property price — it is the property price plus acquisition costs. How far the yield moves when the denominator is replaced with the real outlay is the true measure of what acquisition costs do to investment performance.

Build-Up 1: The Full Breakdown — ¥85.07mn on a ¥1.2bn Hotel

The model property is a 100-room limited-service hotel in a regional core city at a purchase price of ¥1.2bn (¥12mn per room). Land and building are split 35% / 65%, and debt is ¥840mn at 70% LTV. On these assumptions, here is everything that stacks up outside the price.

Brokerage commission at 3.31% and transfer taxes at 2.69% (acquisition tax 1.67% + ownership transfer registration 1.02%) account for 6.00% — 85% of total acquisition costs. Put the other way round: know the level of these two items and you can, in practice, forecast almost the whole acquisition cost.

Brokerage commission is capped by an MLIT public notice issued under the Real Estate Brokerage Act. On the portion of the price above ¥4mn, the cap is “sale price × 3% + ¥60,000” (consumption tax additional). At ¥1.2bn that is ¥36.06mn, or ¥39.67mn including consumption tax. A notice amendment dated 21 June 2024 raised the special cap for low-value vacant properties (¥8mn or less) to ¥330,000 including tax, but that virtually never applies to hotel deals. Because it is a cap rather than a fixed amount, whether the seller or the buyer bears the commission — and whether the agent works one side or both — is a matter for individual negotiation.

Real estate acquisition tax is a prefectural tax, and for a hotel the key point is that the rate splits between land and building. Land acquired on or before 31 March 2027 is taxed at 3% (statutory 4%), and for residential-zoned land the tax base is further compressed to half the fixed asset tax assessed value. Buildings, by contrast, qualify for the 3% concession only where they count as “residential” — hotels and ryokan are non-residential structures and are taxed at the full statutory 4%. That asymmetry is why the ¥15.60mn of acquisition tax on the building in the model is 3.5 times the ¥4.41mn on the land.

Build-Up 2: The Rates from Primary Sources — Statutory Rates, Concessions and Expiry Dates

The easiest mistake when estimating acquisition costs is applying residential concessions to a hotel. Take the rates straight from an explainer written for individual homebuyers and you will under-estimate the building’s acquisition tax by 1 point and its registration and licence tax by as much as 1.7 points. Below are the rates and expiry dates that actually apply to a hotel acquisition.

Two points on timing deserve attention. First, the land concession on registration and licence tax (2.0% → 1.5%) has been extended by three years to 31 March 2029. The extension came with the fiscal 2026 tax reform and covers both ownership transfer registration on land sales and registration of a trust of land ownership. Second, the 3% land rate for real estate acquisition tax and the half-value tax base for residential-zoned land both expire on 31 March 2027. Whether they are extended is not yet settled. If both lapsed, acquisition tax on the model property’s land would rise from ¥4.41mn to ¥11.76mn — an increase of ¥7.35mn. For deals whose completion straddles spring 2027, that gap is worth carrying as a schedule risk.

The other point commonly missed in practice is that there is no reduction for stamp duty on loan agreements. The stamp duty concession is limited to real estate transfer agreements and construction contracts; financing agreements are taxed at the statutory amount. The sums are not large, but budgeting on the assumption that “stamp duty is always reduced” creates a discrepancy.

Build-Up 3: By Deal Size — 7.66% at ¥300mn, 6.95% at ¥3.0bn. Little Room to Optimise

To answer “does buying a bigger property improve acquisition-cost efficiency?” with numbers, here are three cases on identical assumptions: ¥300mn (30 rooms), ¥1.2bn (100 rooms) and ¥3.0bn (150 rooms).

The three cases differ by only 0.71 points. The gap comes mainly from due diligence and judicial scrivener fees — items that do not scale with price and therefore weigh more heavily on smaller deals.

What this shows is that acquisition costs are hard to compress by changing deal size. Brokerage commission and transfer taxes, the dominant items, are both linear in price, so the ratio barely moves: 7.66% on a ¥300mn deal, 6.95% on a ¥3.0bn one. Economies of scale work on the operating expense side, not on acquisition cost.

Conversely, the levers that genuinely move acquisition cost are few. One is negotiating the brokerage commission (including off-market and direct-from-seller deals); the second is compressing transfer taxes through a trust beneficiary interest, discussed below; the third is the land / building allocation ratio. The allocation must be defensible for tax purposes and cannot be set arbitrarily, but its effect is not negligible. Comparing a 20% land share with a 65% one, the model property’s cost ratio moves from 7.22% to 6.82% — 0.40 points. The reason is that the building alone carries the statutory 4% acquisition tax and the statutory 2.0% transfer registration, while land is taxed at 3% on half the assessed value — an effective 0.75%.

How the Yield Is Eroded — from 6.50% Gross to 4.98% NOI to 4.65% Net

Now translate the figures so far into yields. The model property assumes a master lease to an operator on a fixed rent, with annual rental income of ¥78.0mn. Across 100 rooms that is ¥780,000 per room per year, or about ¥2,137 per room per day.

It is worth checking that rent level against market data. According to MetroEngines Research, the estimated settled ADR for business hotels in Fukuoka averaged ¥10,815 over the 13 months from July 2025 to July 2026 (N=329–341 properties per month, monthly median basis). At 80% occupancy (same period, July 2025–July 2026; this article’s assumption), RevPAR works out at roughly ¥8,600. Rent of ¥2,137 is about 25% of that — an unremarkable rent ratio for a limited-service master lease. Whether the lease is struck on a fixed or a variable rent changes both the volatility of income and how risk is shared with the operator, but that choice sits outside the acquisition-cost question and is not modelled here.

From a 6.50% gross yield, deducting operating expenses gives a 4.98% NOI yield (−1.52 points); adding the ¥85.07mn of acquisition costs to the denominator gives a 4.65% net yield (−0.33 points). Acquisition costs alone shave 6.6% off the yield.

0.33 points may look small. Seen from another angle the impression changes: ¥85.07mn of acquisition costs equals 1.42 years of the ¥59.77mn first-year NOI. Before you are handed the keys, cash equivalent to one year and five months of net income has already left the building.

Spread the burden across the holding period and the effect becomes clear. Assuming a three-year hold, the annualised acquisition cost is ¥28.36mn, or 47.4% of NOI; over five years ¥17.01mn and 28.5%; over ten years ¥8.51mn and 14.2%. The shorter the intended hold, the harder acquisition costs press on total return.

Note that for tax purposes, real estate acquisition tax, registration and licence tax and stamp duty need not be capitalised into the acquisition cost of the fixed asset and may instead be expensed in the fiscal year in which they are paid (Corporation Tax Basic Circular 7-3-3-2). They are post-acquisition costs of a transfer-tax character, and a company may choose between capitalising them and expensing them. Brokerage commission, by contrast, is in principle capitalised as a cost directly required for the acquisition, and the land portion is not depreciated. Even within “acquisition costs”, the speed at which cash is recovered differs by line item.

Sensitivity — Assessed Value Alone Moves Costs from 6.70% to 7.87%

The largest source of uncertainty in the calculations so far is the level of the fixed asset tax assessed value. Both real estate acquisition tax and registration and licence tax are levied on the assessed value, not on the purchase price. The actual figure can only be confirmed from the property’s fixed asset valuation certificate, but at the pre-deal review stage an assumption is unavoidable.

Across both axes the cost ratio moves within a 1.57-point band, from 6.61% to 8.18%. The effect is asymmetric: moving the building assessment ratio from 40% to 70% swings the result by 1.17 points at a 35% land share — roughly three times the 0.40-point swing from moving the land share from 20% to 65% at a 50% assessment ratio. Only the building side carries the statutory 4% acquisition tax and the statutory 2.0% transfer registration; the land side is taxed at 3% on half the assessed value, so its effective burden is lighter. In pre-deal review, the building’s assessed value deserves more attention than the allocation negotiation.

The higher the building’s assessed value, the heavier the costs. Newer buildings, with less depreciation adjustment for age, carry larger acquisition and registration taxes — in other words, the newer the building, the heavier the tax burden at acquisition. Between 40% and 70%, the cost ratio moves 1.17 points, more than the 0.71-point difference between a ¥300mn and a ¥3.0bn deal. At the review stage, understanding the assessed value should take priority over deal size. How to think about building age and the timing of renovation capital is covered in Hotels at Year 10: The Renovation Gap Is +13.6% Below 100 Rooms.

Trust Beneficiary Interests — ¥32.22mn of Transfer Taxes Disappear, but a Trust Fee Arrives Every Year

The one element of acquisition cost that can be cut structurally is transfer tax. Rather than buying the real estate itself, taking assignment of the beneficiary interest in real estate held in trust by a trust bank changes the tax treatment substantially.

First, no real estate acquisition tax is levied. The transfer of trust property from settlor to trustee is a formal transfer of ownership and is exempt under Article 73-7(3) of the Local Tax Act. And because assignment of the beneficiary interest is not itself an acquisition of real estate, a change of beneficiary does not trigger acquisition tax.

Second, no ownership transfer registration arises, so registration and licence tax largely disappears. What is required is a change-of-beneficiary registration in the trust register, at ¥1,000 per property. For three land parcels and one building, that is ¥4,000 — two or more orders of magnitude below a direct transaction that pays 1.5% on land and 2.0% on the building.

This is not, however, free tax planning. Holding through a trust beneficiary interest means a trust fee payable to the trust bank for the entire holding period. The amount depends on property size and the terms of the trust agreement; this article models three levels — ¥3mn, ¥4mn and ¥5mn a year.

In the mid case (¥4mn a year), the break-even is about eight years. Sell within eight years and the trust beneficiary interest wins; hold beyond that and cumulative cost on direct ownership is lower.

What is interesting is that on a single-year yield measure the trust beneficiary interest looks worse. Cutting transfer taxes lowers the initial outlay (the denominator) once, whereas the trust fee reduces NOI (the numerator) every year. The mid case gives a net yield of 4.45%, 0.20 points below the 4.65% on direct acquisition. Judge on yield alone and you will misread which option is better on cash flow.

A few practical points are worth noting. Trading a trust beneficiary interest is a securities transaction under the Financial Instruments and Exchange Act, and any intermediary must be registered for Type II financial instruments business. The buyer does not need registration, but instead of the important-matters explanation required under the Real Estate Brokerage Act, it receives a pre-contract disclosure document under the Financial Instruments and Exchange Act — the procedural architecture changes. Listed REITs and private funds commonly hold in beneficiary-interest form, so if the exit is to a similar investor, staying in that form has advantages beyond the tax saved at acquisition.

Note also that placing physical real estate into a new trust incurs its own registration and licence tax on the registration of trust of ownership: 0.3% for land until 31 March 2029 (statutory 0.4%) and 0.4% for buildings, which for the model property is ¥0.88mn on the land and ¥1.56mn on the building, ¥2.44mn in total. Where the property is already in trust, that cost was borne by the seller in the past.

The Easily Missed Item — ¥70.91mn of Consumption Tax on the Building, and Bridging It

It rarely appears on an acquisition-cost schedule, but the largest single cash-flow item is consumption tax. Transfers of land are exempt, but transfers of buildings are taxable. In the model property the allocated building amount of ¥780mn (tax-inclusive) breaks down into a base price of ¥709.09mn and consumption tax of ¥70.91mn — 5.91% of the ¥1.2bn property price.

What matters to a hotel-acquiring business is that this consumption tax is in principle creditable as input tax. Input tax credit is restricted for “residential rental buildings” acquired on or after 1 October 2020 (Article 30(10) of the Consumption Tax Act), but that restriction excludes buildings clearly not to be used for residential letting, and ryokan and hotels fall into that exclusion. A taxable enterprise can therefore take the credit.

The credit or refund is only realised when the return for the taxable period is filed, however, while ¥70.91mn must be paid in cash on the completion date. How to bridge the gap between completion and refund is always an issue on a first acquisition. Tax-exempt enterprises and those using the simplified taxation regime cannot take the credit at all, so this 5.91% is added directly to acquisition cost. It is a point to confirm particularly on acquisitions by high-net-worth individuals.

Market Context — In a World of 4.1–5.3% Expected Yields, How Far Does the Net Figure Fall?

Now set the build-up against market levels. According to the Japan Real Estate Institute’s 54th Japanese Real Estate Investor Survey (as of April 2026), expected yields for limited-service hotels are 4.1% in Tokyo, 4.5% in Kyoto, 4.5% in Osaka, 4.9% in Nagoya, 4.7% in Fukuoka, 5.0% in Sapporo, 5.0% in Naha and 5.3% in Sendai. Against the previous survey (the 53rd, as of October 2025), four districts — Tokyo, Nagoya, Kyoto and Fukuoka — fell by 0.1 points, while the rest were unchanged.

Expected yield is the ratio of NOI to property price and excludes acquisition costs. Layering on the same 7.09% used in this article’s ¥1.2bn model takes Tokyo from 4.1% to 3.83%, and even Sendai from 5.3% to 4.95%.

A falling expected yield means investors are willing to pay a higher price for the same NOI. And because both transfer taxes and brokerage commission scale with price, a rising market inflates the absolute amount of acquisition costs at the same time. Limited-service hotels in Tokyo had compressed to 4.1% as of April 2026; on a net basis including costs that is 3.83% — well below the nominal expected yield in regional core cities. For the ADR level at which new development becomes viable as funding costs rise, Policy Rate 1.00%: Break-Even ADR Across Japan’s 8 Hotel Markets works the calculation backwards across eight major markets.

So what do hotels actually trade for? Aggregating the property data disclosed by six listed hotel REITs (285 domestic properties, ownership master as of 17 August 2026), the median acquisition price per room is ¥23.11mn. The first quartile is ¥11.92mn and the third quartile ¥39.60mn — a spread of more than three times. The median acquisition price itself was ¥3.10bn.

The six investment corporations covered: Hoshino Resorts REIT, Ichigo Hotel REIT, Japan Hotel & Residential REIT, Japan Hotel REIT, Mori Trust REIT and Invincible Investment Corporation.

For a first-time hotel buyer this distribution is a set of coordinates. This article’s model property (¥12mn per room) sits near the first quartile, Case C (¥20mn per room) near the median, and Case A (¥10mn per room) in the bottom quartile. Work back from acquisition price per room to deal size, then add roughly 7% of acquisition costs on top — that is the first step of a funding plan. The procedure for backing out deal size from room count is also covered in Own a Hotel from ¥100,000 — Security Tokens Open a Third Funding Route.

Summary — Seven Numbers to Hold On To at the Review Stage

Calculation Assumptions

Related Reading

• Own a Hotel from ¥100,000 — Security Tokens Open a Third Funding Route
• Who Owns the Hot Spring? Water Rights and Ryokan Valuation
• Japan’s Second Cities: 21 Prefectures Where the Capital Isn’t #1
• Hotels at Year 10: The Renovation Gap Is +13.6% Below 100 Rooms
• Hotel Supply Pressure Ranking 2026-2027: Ozu Kumamoto Tops at 30.8%
• Station Premium Ends at an 8-Minute Walk: 8 Stations, 1,012 Hotels
• Policy Rate 1.00%: Break-Even ADR Across Japan’s 8 Hotel Markets
• Corporate Retreat Conversions: 8.27% GOP Yield Across 5 Onsen Areas