Lining up the September 2026 year-on-year change in estimated settled ADR (tax-excluded equivalent, settled month vs settled month) for business hotels across all 47 prefectures, 44 of the 47 came in above last year, with a median of +7.9%. The base of year-on-year gains had been widening gradually — June (24 prefectures, +0.1%), July (30, +1.2%), August (32, +2.7%) — and in September it spread across almost the whole country at once. The largest upside was Aichi at +53.2% (¥7,600→¥11,700, N=321), while only three prefectures were negative: Osaka at −34.0% (¥13,500→¥8,900, N=504), Shiga at −4.6% (N=80) and Kyoto at −2.4% (N=327). However, September 2026 contained a 5-day holiday from Saturday, September 19 to Wednesday, September 23, so the calendar differs from last year. This article sets out a framework for splitting September YoY into a calendar effect and underlying strength, examines Aichi’s surge and Osaka’s prior-year Expo effect, explains the step between settled values and estimates from October onward, and turns all of this into a procedure for reviewing October–November rate ladders and floors.

Scope: business hotels in all 47 prefectures (N=32–914 properties per prefecture; see the N column of each table). The price metric in this article is estimated settled ADR (the settled price level estimated from OTA and other sales data, tax-excluded equivalent); occupancy is an estimate based on OTA-listed inventory. Definitions are at the end of the article. Data as of: October 19, 2026.

44 prefectures above last year in September — how the base widened over the four months from June

Let us start with the national picture. Table 1 calculates, for each prefecture, the YoY change in estimated settled ADR between settled months (the same month in 2025 and 2026), and lists by month the number of prefectures with a positive change and the median YoY across the 47 prefectures. Business hotels improved in small steps — from 24 prefectures and a median of +0.1% in June to 30 and +1.2% in July and 32 and +2.7% in August — before jumping to 44 prefectures and +7.9% in September. The quartiles tell the same story: the lower quartile was +3.5% and the upper quartile +12.6%, meaning the whole center of the distribution moved up.

By contrast, city hotels, shown for reference, were positive in only 24 prefectures in September with a median of +0.1%. That is a recovery from 12 prefectures and −7.0% in August, but nowhere near as broad as for business hotels. Given that the city hotel median was negative from June through August, the September improvement appears to have happened first on the lower-priced business hotel side.

YoY for each prefecture is calculated between settled values. Business hotels N=32–914 properties per prefecture; city hotels N=2–110 (city hotel figures are for reference only, as some prefectures have N below 10).

Source: MetroEngines Research; compiled by HotelBank Editorial Team

So what is inside those 44 prefectures? Figure 1 ranks the September YoY for all 47 prefectures from highest to lowest. The distribution is 17 prefectures at +10% or more, 15 at +5% to +10%, 12 at 0% to +5% and 3 negative. Besides Aichi (+53.2%), the top of the ranking includes Ishikawa (+20.5%, N=110), Gifu (+20.3%, N=122), Kumamoto (+20.0%, N=122), Mie (+19.9%, N=93), Aomori (+18.8%, N=86) and Miyagi (+17.3%, N=142). Among the major metropolitan areas, Fukuoka was +9.9% (N=342), Hokkaido +8.9% (N=437), Kanagawa +6.5% (N=209) and Tokyo +2.2% (N=914) — all positive, but all below the median.

Figure 1: Business hotel estimated settled ADR, YoY change for September 2026 (by prefecture, settled months). Red = negative. The vertical line is the 47-prefecture median (+7.9%).

Source: MetroEngines Research; compiled by HotelBank Editorial Team

Separating the calendar effect from underlying strength — using August YoY as a yardstick

The calendar difference needs attention here. In September 2026, the 19th (Sat) through the 23rd (Wed, Autumnal Equinox Day) formed a 5-day holiday, whereas September 2025 was split into a 3-day weekend from the 13th (Sat) to the 15th (Mon, Respect for the Aged Day), the weekend of the 20th–21st, and a standalone holiday on the 23rd (Tue, Autumnal Equinox Day). Long holidays concentrate lodging demand and create more high-rate days within the month. It is therefore natural to assume that September YoY carries a nationwide “calendar lift” on top of each prefecture’s underlying demand.

A simple yardstick for separating the two is August YoY, when the calendar difference is small. In August, 32 prefectures were positive with a median of +2.7% — 5.2 points below September’s median. Taking “September YoY minus August YoY” for each prefecture gives a median of +6.3 points. Roughly speaking, about 6 points is a guide to the nationwide calendar effect, and anything well above that can be read as prefecture-specific factors (events, inbound demand, supply). Indeed, 13 of the 15 prefectures that were negative in August turned positive in September. For the Greater Tokyo prefectures — Kanagawa (August −4.8% → September +6.5%), Tokyo (−8.9% → +2.2%), Chiba (−1.8% → +3.9%) and Saitama (−3.9% → +0.4%) — the reasonable reading is that the calendar effect merely brought them back to last year’s level. Conversely, prefectures that were already above the median in August, such as Gifu (+7.3% → +20.3%), Miyagi (+6.5% → +17.3%), Aomori (+12.5% → +18.8%), Iwate (+7.0% → +14.4%) and Niigata (+5.4% → +12.6%), fit the pattern of underlying growth with the calendar effect stacked on top.

Let us check how the calendar effect showed up in occupancy, using estimated OCC (based on OTA-listed inventory). Figure 2 shows daily estimated OCC for business hotels in Tokyo and Aichi in September 2026 (Tokyo N=538–561 properties / 78,348–80,960 rooms; Aichi N=171–175 properties / 24,654–26,210 rooms; both based on listings on major booking sites). Tokyo averaged 90.0% on September 1–14 and 92.6% during the 5-day holiday of the 19th–23rd, peaking at 98.0% on the 20th (Sun) before falling to 81.0% on the 23rd (Wed), the last day of the holiday. That clear peak followed by a pullback is the typical shape of a calendar effect. Our day-by-day review of Silver Week 2026 actuals, which checks how this 5-day holiday landed across 7 areas × 4 hotel types, is also useful for seeing how weekdays recover after the holiday.

Figure 2: Business hotel estimated OCC (based on OTA-listed inventory), daily, September 2026. Tokyo N=538–561 properties, Aichi N=171–175 properties (based on listings on major booking sites). Based on observations from 45 days before the stay date up to the stay date.

Source: MetroEngines Research; compiled by HotelBank Editorial Team

Aichi +53.2% — a late-September high that the 5-day holiday alone cannot explain

Aichi stands in contrast to Tokyo in Figure 2. Aichi business hotels started from a lower level than Tokyo, averaging 84.8% estimated OCC on September 1–14, then rose to an average of 96.9% on the 15th–18th, 97.4% during the 5-day holiday of the 19th–23rd, and held 97.1% even after the holiday on the 24th–30th. On the 20th (Sun) they reached 99.9%, close to fully booked. While Tokyo fell back to 88.7% after the holiday, Aichi stayed high through the end of the month — something the calendar effect alone struggles to explain.

A likely background factor is the 20th Asian Games, held in Aichi and Nagoya from Saturday, September 19 to Sunday, October 4, 2026 (per the Aichi Prefecture official tourism site and Nagoya City materials). The Games began on the first day of the 5-day holiday, and the schedule kept officials and spectators in town through the end of the month. Aichi’s estimated settled ADR had been below last year through August (−3.6%, ¥8,100→¥7,900), but jumped +53.2% in September (¥7,600→¥11,700). City hotels in Aichi, shown for reference, moved the same way at +53.7% (¥11,800→¥18,100, N=31). Our data does not directly measure the Games’ contribution, but the calendar overlap and the shape of late-September estimated OCC suggest that most of this rise is likely a temporary demand concentration.

Figure 3: Aichi business hotel estimated settled ADR (tax-excluded equivalent), monthly, year over year. 2025 from June and 2026 January–September are settled values; October–November are estimates based on current sales (dotted line). N=309–329 properties.

Source: MetroEngines Research; compiled by HotelBank Editorial Team

The year-over-year overlay in Figure 3 shows that Aichi’s monthly level stayed within a ¥7,300–¥8,400 range from June 2025 through August 2026. September’s ¥11,700 is a clear outlier. October (current-sales estimate ¥12,600, N=310) and November (same basis, ¥11,300, N=309) also look high, but these are pre-settlement estimates and should not be lined up alongside September’s settled value before the months close. For operators in Aichi, the priority is not to treat September’s +53.2% as a “new normal” but to check whether booking pace and rates return to the range from mid-October onward, after the Games. For October stay dates in Aichi, our comparison of Aichi’s October booking curves explains how city and business hotels build up bookings differently.

Osaka −34.0% — a comparison-base problem, with last year falling during the Expo

Of the three negative prefectures, Osaka’s −34.0% (¥13,500→¥8,900, N=504) is not a September-only phenomenon. Osaka business hotels have run more than 30% below last year for four straight months — June −32.8%, July −31.9%, August −33.6% — and September is a continuation of that trend. In 2025, the Osaka-Kansai Expo ran from April 13 to October 13 (per the Japan Association for the 2025 World Exposition), and September 2025’s ¥13,500 reflected the demand rush toward the end of the Expo. Osaka fell to ¥11,600 in November 2025 and ¥10,500 in December, so compared with post-Expo levels, the gap to September 2026’s ¥8,900 narrows. Osaka’s −34.0% should therefore be read not as “demand fell 30%” but as “last year’s comparison base was inflated by the Expo.” YoY will regain its normal meaning from November 2026, when post-Expo months can be compared with each other. For Osaka’s level when the comparison is reset to 2024, two years earlier, see our analysis of Osaka’s settled July–September ADR and resetting the baseline.

Kyoto’s −2.4% (¥12,000→¥11,700, N=327) is a sharp narrowing from −15.6% in August. Kyoto business hotels had seen the decline widen from June — −9.8%, −13.0%, −15.6% — but in September, helped by the 5-day holiday, they recovered to roughly last year’s level. The remaining negative, Shiga at −4.6% (¥7,900→¥7,500, N=80), partly reflects a pullback after a large +14.2% gain in August; given the relatively small N of 80 properties, it is best viewed as single-month volatility.

* N below 60 properties (Nara 32, Fukui 46, Saga 49, Tottori 53, Shimane 57, Akita 59); individual properties can sway the aggregate. N is the number of properties covered in September 2026. YoY is calculated between settled values.

Source: MetroEngines Research; compiled by HotelBank Editorial Team

The basis changes from October — do not read settled values and current estimates as one series

September in Table 2 is a settled value, but October onward are estimates based on current sales, so even though both are “estimated settled ADR,” the basis differs. Months that have not yet settled reflect the price level of remaining inventory still on sale, so they tend to show higher levels than settled values. As Table 3 shows, there is a large step between September settled and October estimated values: Tokyo ¥13,400 settled vs ¥20,200 current estimate for October (N=893), Kyoto ¥11,700 vs ¥21,400 (N=320), and Fukuoka ¥11,000 vs ¥15,300 (N=337). Reading this as “rates rose 50% in October” would be wrong; the two cannot be joined as a month-on-month series.

October values are estimates based on current sales and may change as the month settles. Wait for month-end settlement before comparing directly with October 2025 settled values.

Source: MetroEngines Research; compiled by HotelBank Editorial Team

To sum up, three distinctions are needed to read September YoY. First, compare settled months with settled months (Tables 1 and 2). Second, set August YoY alongside as a guide to the calendar effect. Third, compare October-onward estimates with last year again only after they settle. Following these three rules turns the headline “44 prefectures up” into information you can use for your own pricing.

For revenue managers running business hotels nationwide — implications and action plan

Summary — three yardsticks for reading September YoY

Business hotel estimated settled ADR in September 2026 was up YoY in 44 of 47 prefectures, with a median of +7.9% — the broadest year-on-year gain since June. To use this figure in your own decisions, however, you need three yardsticks. First, the gap to August YoY: at the national median, about 6 points is a guide to the calendar effect. Second, last year’s comparison base: Osaka’s −34.0% is measured against a year that fell during the Expo, so look again using settled values from November onward. Third, the distinction between settled values and estimates: do not read Tokyo’s September settled ¥13,400 and October current estimate ¥20,200 as one series. For outliers such as Aichi’s +53.2%, keep in mind that they may be a temporary demand concentration overlapping a major sporting event, and confirm the “reversion” through booking pace in the following months before reflecting them in your rate ladder and floors.

Frequently asked questions

Q. Up to which month can estimated settled ADR YoY be compared using settled values?
A. As of this article’s data date (October 19, 2026), values are settled through September 2026. October onward are estimates based on current sales; they can be compared with last year’s settled values once each month settles.

Q. Can Aichi’s +53.2% be attributed definitively to the Asian Games?
A. The data in this article does not directly measure the Games’ contribution. Because the Games period (September 19 – October 4) coincides with the sustained high estimated OCC in late September, we read it as likely a temporary demand concentration, but we do not state it definitively.

Q. Is the same trend seen at city hotels and ryokan?
A. City hotels, shown for reference, were positive in 24 prefectures in September with a median of +0.1% — not as broad as business hotels. Ryokan and resorts are outside the scope of this article.

About the data

Source: MetroEngines Research; compiled by HotelBank Editorial Team

References and sources

■ Data source

OTA public price and listed-inventory data continuously tracked by MetroEngines Research. Estimated settled ADR is the prefecture-level median of estimated settled rates (tax-excluded equivalent), obtained by applying segment factors to each property’s lowest-plan level (2 guests, per room); estimated OCC is occupancy estimated from OTA-listed inventory. Data as of October 19, 2026.

■ Calculation assumptions

YoY is calculated between settled months in 2025 and 2026 (business hotels N=32–914 properties, city hotels N=2–110). The calendar-effect guide is the median across prefectures of “September YoY minus August YoY” (+6.3pt). Estimated OCC uses a fixed cohort of properties listing at least 30% of total rooms on OTAs: September 2026, Tokyo N=538–561 and Aichi N=171–175 properties. October–November values are estimates based on current sales.

■ Limitations and caveats

These are estimates and differ from each property’s actual settled prices and room occupancy. Estimated OCC counts unlisted rooms as sold, so it runs higher than actual occupancy. In prefectures with N below 60 properties (Nara, Fukui, Saga, Tottori, Shimane, Akita), individual properties can sway the aggregate. The contributions of the Asian Games and the Expo are not measured directly by our data; they are inferred from the overlap in timing.