Do luxury hotels run at higher or lower occupancy than the market as a whole? Lining up MetroEngines Research’s September 2026 estimates (OTA-listed-inventory basis) by prefecture for the “luxury” tier — properties whose average listed price is ¥100,000 or more — the urban markets came in below their all-property averages, at 83.6% in Tokyo and 83.7% in Kyoto (versus 89.8% and 84.0%). Meanwhile, the tiers that include small regional properties sat at around 90%: 90.7% in Nagano, 90.1% in Shizuoka and 92.7% in Hokkaido. In the finalized October figures, seven prefectures converged into an 88–94% band, including Tokyo at 92.1%, Kyoto at 89.7% and Kanagawa at 93.5%. This article first defines what this occupancy rate measures, then examines levels by tier for September and October, how bookings build from 45 days out, the volatility of small cohorts, and the relationship with pricing.
• ADR (Average Daily Rate) = an estimated settled rate (tax-excluded equivalent), calculated by applying segment-specific adjustment factors to the lowest-priced plan each property publishes on OTAs (per room, double occupancy, tax included). Checked against property-level results disclosed by listed hotel REITs (91 properties, latest 3 months), the median error is about 7%. These are estimates and differ from each property’s actual settled prices and accounting figures.
• Area-level ADR (prefecture / municipality) is the median of the estimated settled rates of the properties covered (i.e. the level of a typical property in the area).
• References to listed prices are per-room rates for double occupancy (tax included).
• OCC (occupancy) = rooms sold as a share of total rooms in the area (an estimate based on OTA sales inventory).
• Source: MetroEngines Research (continuously tracks OTA published prices for approximately 168,000 properties in Japan, of which approximately 27,000 have confirmed occupancy data)
OTA-listed-inventory occupancy = 100 − 100 × OTA-listed rooms remaining ÷ total rooms. The universe is a fixed cohort of properties that list at least 30% of their total rooms on OTAs (the same set of properties for each stay date). It is an estimate based on how OTA-listed inventory is consumed and is defined differently from actual room occupancy (rooms not listed on OTAs are counted as sold, so the figure skews high).
“Luxury” tier = properties whose average OTA listed price is ¥100,000 or more (per room, double occupancy, tax included). This is a different yardstick from the estimated settled ADR (tax-excluded equivalent). LT (lead time) = days before the stay date. Booking curves use only 45 days out through the day before.
Wherever estimated OCC or booking curves are shown, we note “N properties / M rooms (major booking sites listing basis)”, and tiers with fewer than 10 properties or fewer than 30 observed days are labeled as “reference values”.
- — September: Tokyo 83.6%, Kyoto 83.7% (September 2026, estimated occupancy, OTA-listed-inventory basis). Urban luxury came in below the all-property averages (Tokyo 89.8%, Kyoto 84.0%). Tokyo is a reference value based on 7 properties and 17 observed days.
- — Small regional luxury properties at around 90%: In September 2026, Hokkaido was 92.7% and Nagano 90.7% (both reference values), and Shizuoka 90.1% — 7.1pt above its all-property average.
- — October: seven prefectures converge at 88–94%: In October 2026, Tokyo was 92.1%, Kyoto 89.7% and Kanagawa 93.5%. The property cohorts differ from September, so this should not be read as a month-on-month change.
- — At 45 days out, luxury leads all properties in five prefectures (October 2026 average). In Tokyo, the gap opens in the last-minute build-up: +30.9pt for all properties versus +24.5pt for luxury.
- — Estimated settled ADR is up YoY in all seven prefectures (September and October 2026, all property types). Tokyo was ¥17,584 in October (+12.5%); Okinawa was +12.7%.
What This Occupancy Rate Measures — Why It Runs Higher Than Actual Room Occupancy
Let us start by checking the yardstick. The occupancy rate in this article is an estimate derived by dividing the rooms remaining on major booking sites by total rooms. Because we aggregate only a fixed cohort of properties that list at least 30% of their total rooms on OTAs, the comparison holds the same set of properties for each stay date — but rooms not offered on OTAs (direct bookings, corporate contracts, member allocations, etc.) are counted on the “sold” side. As a result, the level runs higher than actual room occupancy. This gap tends to be larger for luxury properties, which typically have a higher share of direct sales, so figures such as 83% or 92% in this article cannot be placed side by side with occupancy figures published by properties or with the Japan Tourism Agency’s room occupancy rate.
So what is it useful for? Relative differences between tiers and regions measured on the same definition, and how bookings accumulate by lead time. Does urban luxury come in below the overall average? Do small regional properties reach around 90%? How full are they at 45 days out? For readers considering development or operations, this structure matters more than absolute levels. How to read these figures against the room occupancy rates in the Japan Tourism Agency’s “Overnight Travel Statistics Survey” (city hotel and resort hotel categories) is covered in detail in our earlier article on occupancy in Okinawa, Osaka, Yokohama and Fukuoka.
September by Prefecture — Tokyo 83.6% and Kyoto 83.7% Below Average, Small Regional Properties Around 90%
The table below lists September 2026 monthly averages for seven prefectures. Tokyo’s luxury tier was 83.6% across 7 properties and 1,193 rooms, but only 17 of the 30 days could be observed, so it is treated as a reference value. Kyoto was 83.7% across 13 properties and 589 rooms, and even within urban markets the deluxe tier (18 properties, 1,549 rooms) was 81.6% — all near or below the all-property average of 84.0%. In Tokyo, deluxe (20 properties, 3,454 rooms) was 88.3% and all properties (805 properties, 104,583 rooms) 89.8%, so the higher the price tier, the lower the monthly average.
The regional luxury tiers look different. Hokkaido (10 properties, 159 rooms) was 92.7% (29 observed days, reference value), Nagano (8 properties, 155 rooms) 90.7% (27 observed days, reference value), Shizuoka (13 properties, 151 rooms) 90.1% and Kanagawa (13 properties, 205 rooms) 89.1% — all around 90% — with Shizuoka and Kanagawa exceeding their all-property averages (83.0% and 84.2%) by 7.1 points and 4.9 points respectively. Okinawa (17 properties, 171 rooms) was 83.9%, closer to the urban markets, though only just under 3 points below its all-property average of 86.7%. A caveat: the regional luxury tiers consist of small properties with a dozen or so rooms each, and in tiers totaling 151–205 rooms, a single property adding or pulling inventory can move the monthly average by several points.
| Prefecture | Luxury | Deluxe | All properties | Luxury − All |
|---|---|---|---|---|
| Tokyo | 83.6% (reference value) 7 properties / 1,193 rooms · 17 days observed | 88.3% 20 properties / 3,454 rooms | 89.8% 805 properties / 104,583 rooms | -6.2pt |
| Kyoto | 83.7% 13 properties / 589 rooms · 30 days observed | 81.6% 18 properties / 1,549 rooms | 84.0% 566 properties / 31,989 rooms | -0.3pt |
| Okinawa | 83.9% 17 properties / 171 rooms · 30 days observed | — | 86.7% 708 properties / 36,220 rooms | -2.8pt |
| Hokkaido | 92.7% (reference value) 10 properties / 159 rooms · 29 days observed | — | 90.3% 679 properties / 53,149 rooms | +2.4pt |
| Shizuoka | 90.1% 13 properties / 151 rooms · 30 days observed | — | 83.0% 690 properties / 29,829 rooms | +7.1pt |
| Kanagawa | 89.1% 13 properties / 205 rooms · 30 days observed | — | 84.2% 385 properties / 30,855 rooms | +4.9pt |
| Nagano | 90.7% (reference value) 8 properties / 155 rooms · 27 days observed | — | 88.4% 507 properties / 18,893 rooms | +2.3pt |
Occupancy for September 2026 (by stay date; OTA-listed-inventory basis, estimated, monthly average). Property / room counts are fixed cohorts based on listings on major booking sites. “Reference value” = tiers with fewer than 10 properties or fewer than 30 observed days. The deluxe tier is shown only for Tokyo and Kyoto, where at least 10 properties are available.
Source: Compiled by HotelBank Editorial Team from MetroEngines Research
Viewing the same figures as a bar chart makes the pattern clear: in urban markets (Tokyo, Kyoto, Okinawa) luxury falls below all properties, while in regional markets (Hokkaido, Shizuoka, Kanagawa, Nagano) it comes in above.
Occupancy in September 2026 (OTA-listed-inventory basis, estimated) — luxury tier vs all properties. Tokyo luxury is a reference value based on 7 properties and 17 observed days.
Source: Compiled by HotelBank Editorial Team from MetroEngines Research
Finalized October Figures — All Seven Prefectures Converge at 88–94%, Tokyo 92.1%, Kyoto 89.7%
October has a full 31 days of observations, and the luxury cohorts are also larger than in September. Tokyo was 92.1% (23 properties, 2,193 rooms), Kyoto 89.7% (30 properties, 769 rooms), Kanagawa 93.5% (31 properties, 655 rooms), Okinawa 92.3% (39 properties, 581 rooms), Nagano 93.1% (16 properties, 170 rooms) and Shizuoka 91.1% (19 properties, 242 rooms); including Hokkaido at 88.6% (11 properties, 128 rooms; 29 observed days, reference value), all seven prefectures fell within the 88–94% band. The all-property averages were also high, at 95.5% in Tokyo and 92.7% in Kyoto, indicating that tier differences narrow during the autumn demand season in October.
However, the cohort of properties differs between September and October (Tokyo luxury rose from 7 to 23 properties), so the difference between the two months cannot be read as “an 8-point improvement from the previous month”. The accurate reading is to treat each month as a separate snapshot: in October, Tokyo and Kyoto luxury still sat about 3 points below their all-property averages, while Kanagawa, Nagano and Shizuoka remained above theirs — the same direction as in September. In Kyoto, the deluxe tier (18 properties, 1,549 rooms) was 88.7%, almost level with luxury at 89.7%.
| Prefecture | Luxury | Deluxe | All properties | Luxury − All |
|---|---|---|---|---|
| Tokyo | 92.1% 23 properties / 2,193 rooms · 31 days observed | 94.7% 22 properties / 3,928 rooms | 95.5% 874 properties / 105,297 rooms | -3.4pt |
| Kyoto | 89.7% 30 properties / 769 rooms · 31 days observed | 88.7% 18 properties / 1,549 rooms | 92.7% 778 properties / 32,933 rooms | -3.0pt |
| Okinawa | 92.3% 39 properties / 581 rooms · 31 days observed | — | 90.1% 933 properties / 37,131 rooms | +2.2pt |
| Hokkaido | 88.6% (reference value) 11 properties / 128 rooms · 29 days observed | — | 87.5% 788 properties / 55,634 rooms | +1.1pt |
| Shizuoka | 91.1% 19 properties / 242 rooms · 31 days observed | — | 83.7% 887 properties / 32,607 rooms | +7.4pt |
| Kanagawa | 93.5% 31 properties / 655 rooms · 31 days observed | — | 90.7% 456 properties / 31,644 rooms | +2.8pt |
| Nagano | 93.1% 16 properties / 170 rooms · 31 days observed | — | 91.3% 696 properties / 22,461 rooms | +1.8pt |
Occupancy for October 2026 (by stay date; OTA-listed-inventory basis, estimated, monthly average). Property / room counts are fixed cohorts based on listings on major booking sites; the set of properties differs from September, so differences from September should not be read as month-on-month changes. “Reference value” = tiers with fewer than 10 properties or fewer than 30 observed days.
Source: Compiled by HotelBank Editorial Team from MetroEngines Research
Occupancy in October 2026 (OTA-listed-inventory basis, estimated) — luxury tier vs all properties. Hokkaido luxury is a reference value based on 11 properties and 29 observed days.
Source: Compiled by HotelBank Editorial Team from MetroEngines Research
How Bookings Build — Okinawa 83% and Tokyo 68% at 45 Days Out, Reaching Around 90% by the Day Before
Monthly averages alone do not reveal when bookings come in. Looking at booking curves that average estimated occupancy from 45 days out to the day before across October’s 31 stay dates, distinct regional characters emerge. Okinawa luxury (average 39 properties, 898 rooms) had already reached 83.3% at 45 days out, building only 9 points to 92.3% by the day before. This reflects the fact that demand is centered on resort stays that are longer and booked early.
In contrast, Tokyo luxury (average 29 properties, 2,775 rooms) built 24.5 points from 67.6% at 45 days out to 92.1% the day before, and Kyoto (average 40 properties, 1,342 rooms) rose 15.9 points from 73.8% to 89.7%. Urban markets rely more heavily on last-minute demand, and the low level at 45 days out partly drags down the monthly average. Hokkaido (average 13 properties, 236 rooms; reference value) was nearly flat from 63.2% at 45 days out to 68.0% at 7 days out, then jumped to 88.6% the day before — a shape that should be seen as reflecting inventory being added and pulled, which is typical of small cohorts.
| Prefecture (Luxury) | 45 days out | 30 days out | 14 days out | 7 days out | Day before | Build-up over 45 days |
|---|---|---|---|---|---|---|
| Tokyo avg. 29 properties / 2,775 rooms | 67.6% | 71.4% | 79.1% | 83.0% | 92.1% | +24.5pt |
| Kyoto avg. 40 properties / 1,342 rooms | 73.8% | 77.9% | 83.5% | 84.1% | 89.7% | +15.9pt |
| Okinawa avg. 39 properties / 898 rooms | 83.3% | 85.2% | 86.1% | 86.7% | 92.3% | +9.0pt |
| Hokkaido avg. 13 properties / 236 rooms | 63.2% | 65.9% | 67.2% | 68.0% | 88.6% | +25.4pt |
| Shizuoka avg. 26 properties / 405 rooms | 72.3% | 78.1% | 82.4% | 82.9% | 91.1% | +18.8pt |
| Kanagawa avg. 37 properties / 839 rooms | 79.3% | 84.0% | 86.9% | 87.7% | 93.5% | +14.2pt |
| Nagano avg. 19 properties / 273 rooms | 75.6% | 79.5% | 82.1% | 81.1% | 93.1% | +17.5pt |
Booking curves averaged over the 31 stay dates of October 2026 (lead time 45 days out to the day before). Property counts are averages of each stay date’s cohort and do not match those in the monthly-average tables. For Hokkaido, 2 of the 31 stay dates did not meet the cohort requirements, so the figure is a 29-day average (reference value).
Source: Compiled by HotelBank Editorial Team from MetroEngines Research
Booking curves averaged over October 2026 (45 days out to the day before). Solid lines are the luxury tier; the grey dotted line is all properties in Tokyo (average 847 properties, 103,939 rooms).
Source: Compiled by HotelBank Editorial Team from MetroEngines Research
The reference line for all Tokyo properties (average 847 properties, 103,939 rooms) builds 30.9 points from 64.6% at 45 days out to 95.5% the day before, so in Tokyo both luxury and all properties share the trait of being a “market that fills at the last minute”. The difference lies in the end point: luxury reaches 92.1% against 95.5% for all properties, leaving a gap of just over 3 points. How to read booking curves in general is explained in our article using the curves of 541 Tokyo business hotels.
Comparing Curves With All Properties — Luxury Leads in Five Prefectures at 45 Days Out; the Gap Opens in the Last-Minute Build-Up
Placing the luxury tier’s curve alongside the all-property curve for the same prefecture shows where differences come from in a way monthly averages cannot. In estimated occupancy at 45 days out (OTA-listed-inventory basis), averaged across October 2026’s 31 stay dates, luxury exceeded all properties in five prefectures: Tokyo, Okinawa, Shizuoka, Kanagawa and Nagano. The lead was largest in resort and regional markets — Kanagawa +17.6pt, Okinawa +13.4pt, Shizuoka +12.5pt — and Tokyo was also ahead of all properties at +3.0pt. Only Kyoto (−1.4pt) and Hokkaido (−4.4pt, reference value) were below.
The gap reverses in the last-minute build-up. The increase from 45 days out to the day before was +24.5pt for luxury versus +30.9pt for all properties in Tokyo, and +14.2pt versus +29.0pt in Kanagawa — luxury trailed all properties in six of the seven prefectures (the exception being Hokkaido, at +25.4pt versus +19.9pt). The reading most consistent with the numbers is that luxury’s monthly average in Tokyo and Kyoto falls short of the all-property average not because early bookings build slowly, but because the volume of last-minute bookings does not match that of all properties. For urban luxury, closing the gap with all properties is a question not only of raising the level at 45 days out, but also of which channels to use to sell out remaining rooms close to arrival without discounting.
| Prefecture | 45 days out (Luxury) | Gap at 45 days out | Day before (Luxury) | Gap on day before | 45-day build-up |
|---|---|---|---|---|---|
| Tokyo Luxury avg. 29 properties / 2,775 rooms All avg. 847 properties / 103,939 rooms | 67.6% All 64.6% | +3.0pt | 92.1% All 95.5% | −3.4pt | +24.5pt All +30.9pt |
| Kyoto Luxury avg. 40 properties / 1,342 rooms All avg. 761 properties / 32,688 rooms | 73.8% All 75.2% | −1.4pt | 89.7% All 92.7% | −3.0pt | +15.9pt All +17.5pt |
| Okinawa Luxury avg. 39 properties / 898 rooms All avg. 896 properties / 36,263 rooms | 83.3% All 69.9% | +13.4pt | 92.3% All 90.1% | +2.2pt | +9.0pt All +20.2pt |
| Hokkaido Luxury avg. 13 properties / 236 rooms(reference value) All avg. 786 properties / 57,007 rooms | 63.2% All 67.6% | −4.4pt | 88.6% All 87.5% | +1.1pt | +25.4pt All +19.9pt |
| Shizuoka Luxury avg. 26 properties / 405 rooms All avg. 893 properties / 32,628 rooms | 72.3% All 59.8% | +12.5pt | 91.1% All 83.7% | +7.4pt | +18.8pt All +23.9pt |
| Kanagawa Luxury avg. 37 properties / 839 rooms All avg. 455 properties / 31,186 rooms | 79.3% All 61.7% | +17.6pt | 93.5% All 90.7% | +2.8pt | +14.2pt All +29.0pt |
| Nagano Luxury avg. 19 properties / 273 rooms All avg. 733 properties / 22,909 rooms | 75.6% All 69.2% | +6.4pt | 93.1% All 91.3% | +1.8pt | +17.5pt All +22.1pt |
Booking curves averaged over the 31 stay dates of October 2026 (OTA-listed-inventory basis, estimated). “Gap” = luxury tier − all properties. Property / room counts are averages of each stay date’s cohort, and cohort sizes differ greatly between the two groups. Hokkaido luxury is a 29-stay-date average (reference value).
Source: Compiled by HotelBank Editorial Team from MetroEngines Research
Small Properties’ Rates Move With a Single Property’s Inventory — Looking at Daily Volatility
Before taking the “around 90%” figure for the regional luxury tiers at face value, it is worth checking daily volatility. In October, Tokyo luxury (23 properties, 2,193 rooms) stayed within a daily estimated occupancy range of 88.7–95.9%, with a standard deviation of 1.8 points. Hokkaido luxury (11 properties, 128 rooms; reference value), by contrast, moved within a range of nearly 25 points, from 74.3% to 99.2%, with a standard deviation of 7.4 points. Okinawa luxury (39 properties, 581 rooms) also had days at the end of the month — October 28–30 — that fell to 67–75%, giving a monthly range of more than 30 points.
In a tier with only a hundred-odd rooms in total, a single property listing or withdrawing a few dozen rooms at once is enough to move that day’s rate by more than 10 points. So rather than concluding that “small luxury properties in Hokkaido run at 90% occupancy”, it is closer to reality to say that “the monthly average is around 90%, but daily figures range from the 70s to near full”. If this figure is used in development planning, the daily distribution — not the monthly average — is what should be examined.
Daily occupancy in October 2026 (OTA-listed-inventory basis, estimated). Tokyo luxury, 23 properties / 2,193 rooms vs Hokkaido luxury, 11 properties / 128 rooms (reference value; Oct 8–9 missing as cohort requirements were not met).
Source: Compiled by HotelBank Editorial Team from MetroEngines Research
Relationship With Pricing — Estimated Settled ADR Up YoY in All Seven Prefectures
Now let us overlay pricing on the occupancy picture. Prefecture-wide estimated settled ADR (all property types, median of covered properties, tax-excluded equivalent) exceeded the same month of the previous year in all seven prefectures in September 2026: Tokyo ¥13,700 (N=1,139, +1.9% YoY), Kyoto ¥14,200 (N=604, +1.1%), Okinawa ¥12,000 (N=515, +10.0%) and Hokkaido ¥11,500 (N=960, +4.5%). In October the gains widened, with Tokyo at ¥17,600 (N=1,137, +12.5%), Kyoto ¥20,600 (N=602, +6.9%) and Okinawa ¥12,200 (N=515, +12.7%) — rates and occupancy rising together in the autumn demand season.
Prices for the luxury tier itself can only be seen through listed prices (double occupancy, tax included, average of all plans), not estimated settled ADR. October monthly averages were Tokyo ¥145,700, Kyoto ¥147,400, Okinawa ¥148,600, Nagano ¥144,000, Kanagawa ¥141,000, Shizuoka ¥130,700 and Hokkaido ¥124,200, placing all seven prefectures in the ¥120,000–150,000 band. That urban and regional markets show different occupancy patterns despite no major difference in listed price levels shows that the luxury market cannot be discussed as a single block.
| Prefecture | Est. settled ADR, Sep 2026 (all types) | YoY | Est. settled ADR, Oct 2026 (all types) | YoY | Luxury tier listed price (Oct avg., tax incl.) |
|---|---|---|---|---|---|
| Tokyo | ¥13,660 N=1,139 | +1.9% | ¥17,584 N=1,137 | +12.5% | ¥145,657 |
| Kyoto | ¥14,168 N=604 | +1.1% | ¥20,564 N=602 | +6.9% | ¥147,431 |
| Okinawa | ¥12,040 N=515 | +10.0% | ¥12,228 N=515 | +12.7% | ¥148,555 |
| Hokkaido | ¥11,456 N=960 | +4.5% | ¥10,109 N=943 | +4.1% | ¥124,230 |
| Shizuoka | ¥12,990 N=862 | +3.0% | ¥12,683 N=859 | +0.4% | ¥130,715 |
| Kanagawa | ¥14,154 N=579 | +5.6% | ¥14,695 N=579 | +7.9% | ¥141,027 |
| Nagano | ¥11,841 N=818 | +3.7% | ¥12,160 N=823 | +3.9% | ¥144,044 |
Estimated settled ADR covers all property types in each prefecture (median of covered properties, tax-excluded equivalent). The luxury tier’s listed price is the October average of per-room rates for double occupancy (tax included) across that tier’s cohort, and is not ADR.
Source: Compiled by HotelBank Editorial Team from MetroEngines Research
Reading Against the Japan Tourism Agency’s Room Occupancy Rate
The room occupancy rate in the Japan Tourism Agency’s “Overnight Travel Statistics Survey” is an all-rooms statistic based on results reported by properties. The overall figure for July 2026 (first preliminary release) was 60.8%, and for March 2026 (second preliminary release) city hotels were 72.1% and resort hotels 58.8%. The estimates in this article come out in the 80–90% range because of the difference in definition (rooms not listed on OTAs are counted as sold) — it does not mean the market is running 30% higher than the official statistics. Note also that from January 2026 data, the survey changed its stratification basis from number of employees to number of rooms, so YoY comparisons include the effect of this revision.
The principle for reading them is simple. Use the Japan Tourism Agency statistics when discussing long-term trends, YoY comparisons or actual property-wide occupancy; use the estimates in this article to read near-term-to-future booking progress, relative comparisons between tiers and regions, and build-up by lead time. The agency’s property-type categories do not include “luxury” — figures are reported separately for city hotels and resort hotels — so they do not map one-to-one onto this article’s price-based tiers. Our article examining the 27.5pt gap between the Japan Tourism Agency’s 61.8% and the OTA-inventory-based 89.3% for July, which tracks the same definitional difference month by month for Okinawa alone, is a useful worked example.
Implications for Readers Considering Development and Operations
Putting all this together, three opportunities emerge. First, urban luxury is a market where occupancy at 45 days out is low — 67.6% in Tokyo and 73.8% in Kyoto — and builds on last-minute demand to around 90% by the day before (Tokyo 92.1%, Kyoto 89.7%). Product design for multi-night and long stays that broadens the early-booking segment, and operations that raise the share of direct sales through corporate and member channels, have room to grow by pulling the curve forward. Second, small regional luxury is at a high level of around 90% on a monthly average basis, but daily swings are large, and the scale is such that a single property’s inventory decisions move the regional rate. Anyone considering entry should look at the daily distribution and direct-sales share rather than the high monthly average.
Third, 2026 has seen a run of international luxury openings in Kyoto, Nagoya and elsewhere, and large projects targeting 2028 have been announced in Hamamatsucho and Yokohama. The concentration of entry into the high-price segment while new room supply nationwide remains limited needs to be read together with the current situation in which Tokyo and Kyoto luxury fall below their all-property averages. Whether occupancy can be maintained as supply increases depends on how well the front half of the curve can be filled, and the metrics in this article serve as a yardstick for tracking that progress monthly. Price tiers within 1km of Gion are covered in our article on the 11 properties in Higashiyama Ward listed above ¥100,000, and the economics of a large Tokyo project in our required-ADR estimate for the 130-room Hamamatsucho project.
Frequently Asked Questions
Q. Do luxury hotels have lower occupancy than other property types?
A. In the September 2026 estimates, the luxury tiers in Tokyo (83.6%; 7 properties, 17 observed days, reference value) and Kyoto (83.7%; 13 properties, 589 rooms) were below their respective all-property averages (Tokyo 89.8%, Kyoto 84.0%). Meanwhile, tiers that include small regional properties — such as Nagano 90.7%, Shizuoka 90.1% and Hokkaido 92.7% — were around 90%, on par with or slightly above the all-property averages. In October, all seven prefectures converged at 88–94%, so it cannot be said across the board that “luxury means lower occupancy”.
Q. Why does this occupancy rate differ from the Japan Tourism Agency’s room occupancy rate?
A. The occupancy rate in this article is estimated from how inventory listed on OTAs is consumed; rooms not offered on OTAs are counted as sold, so it runs higher than actual room occupancy. The room occupancy rate in the Japan Tourism Agency’s “Overnight Travel Statistics Survey” is an all-rooms figure based on results reported by properties, and the overall figure for July 2026 was 60.8%. Rather than comparing levels directly, use it as an indicator for relative differences between tiers and regions and for how bookings accumulate.
Q. How should figures for regions with small samples be read?
A. In tiers covering only a dozen or so properties and a hundred-odd rooms — such as Hokkaido (11 properties, 128 rooms in October) or Nagano (8 properties, 155 rooms in September) — a single property adding or pulling inventory can move the monthly average by several points. This article labels tiers with fewer than 10 properties or fewer than 30 observed days as “reference values” and also shows daily volatility (Hokkaido ranged from 74.3% to 99.2% in October). It is safer to read them as a structural difference between urban markets and small regional properties rather than as levels in themselves.
Summary
Occupancy for the luxury tier — properties with an average listed price of ¥100,000 or more — (OTA-listed-inventory basis, estimated) in September 2026 was below the all-property averages in Tokyo at 83.6% (7 properties, 1,193 rooms; 17 observed days, reference value) and Kyoto at 83.7% (13 properties, 589 rooms), while tiers including small regional properties were around 90%: Nagano 90.7%, Shizuoka 90.1% and Hokkaido 92.7%. In October, seven prefectures converged at 88–94%, with Tokyo at 92.1% (23 properties, 2,193 rooms), Kyoto 89.7% (30 properties, 769 rooms) and Kanagawa 93.5% (31 properties, 655 rooms). The booking curves confirmed that Okinawa fills early at 83.3% at 45 days out, while Tokyo is a market that builds at the last minute, from 67.6% to 92.1% by the day before. Small cohorts move 25–30 points day to day, so they should be read as distributions rather than monthly averages; and on the understanding that these are estimates defined differently from the Japan Tourism Agency statistics, we encourage readers to use them as indicators of the structure across tiers and regions.
Related Reading
- Hotel Occupancy Sep 2026: Okinawa 86.7%, Osaka 80.4%, Kanagawa 84.2%
- Tokyo City Hotel ADR: −10.3% Sep to −0.7% Oct; Business Hotels +14.0%
- What Is a Booking Curve? Tokyo Business Hotels Go 61.6% to 95.8%
- Gion 1km: 11 Hotels, 562 Rooms Above ¥100k as Imperial, Capella Open
- Raffles Tokyo 2028: 130 Rooms in Hamamatsucho, ¥96,200 Required ADR
- Estimated OCC vs Sell-Out Rate: Tokyo Business 77.7% vs Capsule 93.8%
- Okinawa Occupancy 2026: 61.8% Official vs 89.3% OTA, Why a 27.5pt Gap
References & Sources
■ Data Sources
MetroEngines Research estimated occupancy by prefecture × property tier (OTA-listed-inventory basis, stay dates in September and October 2026), booking curves on the same basis (average of October 2026’s 31 stay dates, lead time 45 days out to the day before), and estimated settled ADR by prefecture (September 2025 – October 2026, all property types). The Japan Tourism Agency’s “Overnight Travel Statistics Survey” was referenced only to illustrate the difference in definitions.
■ Calculation Assumptions
Occupancy = 100 − 100 × OTA-listed rooms remaining ÷ total rooms. A fixed cohort is constructed for each stay date from properties that list at least 30% of their total rooms on OTAs. “Luxury” = properties whose average OTA listed price is ¥100,000 or more (per room, double occupancy, tax included). Estimated settled ADR is the median of covered properties across all property types in each prefecture (tax-excluded equivalent); YoY compares against the same aggregation for the same month of 2025. Differences and build-up figures in Tables 1, 2 and 4 are differences between the displayed values.
■ Limitations and Caveats
Rooms not listed on OTAs are counted as sold, so figures run higher than actual room occupancy. The sets of properties covered differ between September and October, and between monthly averages and booking curves, so differences cannot be read as month-on-month changes. Tiers with fewer than 10 properties or fewer than 30 observed days (luxury in Tokyo for September, Hokkaido, and Nagano for September) are reference values. Prices for the luxury tier are listed prices, not ADR.
■ References
- MetroEngines Research — Estimated occupancy by prefecture × property tier (OTA-listed-inventory basis, September / October 2026), booking curves (October 2026, lead time 45 days out to the day before), estimated settled ADR (September 2025 – October 2026)
- Japan Tourism Agency, “Overnight Travel Statistics Survey” press releases (July 2026 first preliminary, August 2026 first preliminary, and others)
- Travel Voice, “Where are luxury hotels headed in 2026? Six notable hotels opening in Japan this year” (January 7, 2026)
- Real Estate Asia “Why hotel supply will remain limited in Japan in the near term”
