Higashiyama-ku and Sakyo-ku share a border. They share the same prefecture, the same train operators, and the same overall Kyoto cultural brand. The investment math behind them is not the same. Higashiyama trades at a median GRM of 328x with a station spread of barely 1.4x. Sakyo trades at 263x with a 3x station spread. Most foreign buyers treat Kyoto as a single premium market. The transaction data tells a more useful story.
This article walks through 7,194 active rental listings and 441 sales transactions across the two wards, station by station, and explains where the real arbitrage sits.
Two wards, one prefecture, two different markets
Kyoto City is divided into eleven administrative wards. The two we look at here sit on the eastern side of the city, separated by Sanjo-dori and the Kamogawa river. They are easy to confuse on a tourist map and impossible to confuse on a yield map.
Higashiyama-ku covers the densest stretch of the historic tourist core. Kiyomizu-dera, Yasaka Shrine, the Gion geisha quarter, and the Higashiyama hillside walk all sit inside this ward. The resident population is small, the housing stock is constrained by preservation rules, and demand for short-term accommodation is global.
Sakyo-ku is larger, residential, and structurally different. It runs north from the Imperial Palace area up to Mount Hiei and the Kurama corridor. Kyoto University sits in the southern half. The Eizan Electric Railway runs along its spine. There are pockets of expensive villa real estate (Shugakuin, Matsugasaki) and pockets of student rental supply (Mototanaka, Ichijoji, Demachiyanagi).
The yield curve across the two wards reflects this directly.
The dataset
We pulled active listings from Suumo, Lifull HOME's, and AtHome, refreshed weekly. After deduplication and outlier filtering, the working set for this analysis is:
- Higashiyama-ku: 2,014 rental listings, 132 sales transactions
- Sakyo-ku: 7,060 rental listings, 309 sales transactions
GRM here is calculated as median price per m2 divided by median monthly rent per m2. All multiples are monthly basis, not annual. To compute station-level multiples, we require at least 20 active rental listings and 10 active sales transactions at the station. Stations below those thresholds are excluded rather than smoothed in.
Higashiyama-ku: one-dimensional demand
Three stations in Higashiyama clear our liquidity threshold. The full station table:
| Station | Listings | Sales | Rent ¥/m² | Price ¥/m² | GRM | Yield |
|---|---|---|---|---|---|---|
| Gion-shijo | 269 | 41 | 2,416 | 754,821 | 312x | 3.21% |
| Kiyomizu-gojo | 543 | 49 | 2,316 | 730,435 | 315x | 3.17% |
| Higashiyama | 197 | 14 | 2,351 | 993,418 | 423x | 2.36% |
The total range across the ward is 312x to 423x. That is a 1.4x spread. For comparison, a single Tokyo ward like Shibuya runs from roughly 170x to 350x, and Sumida runs from 186x to 215x. Higashiyama compresses harder than either.
The structural reason is simple. Demand in Higashiyama is one-dimensional. Buyers want walk-to-Kiyomizu, walk-to-Gion, or walk-to-the-Higashiyama-hillside. Every property in the ward competes for the same pool of capital, which is mostly inbound investors and Airbnb operators paying for tourist proximity. There is no second axis of demand (no major university, no commuter business cluster, no industrial employer) that would create internal price dispersion.
When demand is one-dimensional and supply is constrained by preservation rules, prices flatten upward and yields flatten downward. Higashiyama is a yield-compressed market. The data confirms it.
Sakyo-ku: a real city inside one ward
Ten stations in Sakyo clear our liquidity threshold. The picture is completely different:
| Station | Listings | Sales | Rent ¥/m² | Price ¥/m² | GRM | Yield |
|---|---|---|---|---|---|---|
| Yase-Hieizanguchi | 31 | 20 | 1,312 | 217,757 | 166x | 6.02% |
| Chayama | 633 | 52 | 2,091 | 439,011 | 210x | 4.76% |
| Mototanaka | 825 | 20 | 2,131 | 470,088 | 221x | 4.53% |
| Keage | 91 | 18 | 1,916 | 447,942 | 234x | 4.27% |
| Ichijoji | 656 | 26 | 1,857 | 503,728 | 271x | 3.69% |
| Demachiyanagi | 1,147 | 41 | 2,222 | 875,274 | 394x | 2.54% |
| Jingu-marutamachi | 481 | 30 | 2,242 | 904,863 | 404x | 2.48% |
| Matsugasaki | 439 | 19 | 2,033 | 923,703 | 454x | 2.20% |
| Shugakuin | 390 | 26 | 1,702 | 856,563 | 503x | 1.99% |
The total range is 166x to 503x. That is a 3x spread within a single ward.
The bottom of the table (Yase-Hieizanguchi, Chayama, Mototanaka) is the student rental belt along the Eizan line. Buildings are small, the rent per m2 is moderate, and prices have not been bid up by tourism capital. Yields cluster between 4.5% and 6%.
The top of the table (Shugakuin, Matsugasaki, Demachiyanagi) is the premium villa and access-to-temples belt. Shugakuin in particular sits next to the Imperial Detached Villa and carries a heavy premium per m2 that is not matched by rental absorption. Yields cluster between 2% and 2.5%.
A property bought at Yase-Hieizanguchi at the median price clears its rent in roughly 14 years of gross income. The same property bought at Shugakuin clears in 42. Both are in Sakyo-ku.
Why the spread matters for foreign buyers
The standard mistake is to read a ward-level yield number for Kyoto and underwrite a deal against it. Sakyo-ku's median yield is 3.17%. That number is technically correct and operationally meaningless, because no single property in Sakyo trades at the median. Properties trade at the station level, and the difference between the right and wrong station inside Sakyo is two or three percentage points of gross yield, which compounds heavily over a ten-year hold.
For a foreign buyer evaluating Kyoto from abroad, the practical implication is to do the analysis at station level before drawing any conclusion about ward attractiveness. A Sakyo property near Mototanaka and a Sakyo property near Matsugasaki are entirely different investment products. Treating them as one bucket because they share an administrative boundary will produce systematically poor underwriting decisions.
For Higashiyama, the implication is different. The station spread is narrow, so the choice is less about "which station inside Higashiyama" and more about "is Higashiyama the right ward for this thesis." If the thesis is short-term rental cash flow at scale, the math is tight. If the thesis is capital preservation in a globally branded micro-market, the math works.
How to use this on a deal
When you are screening a Kyoto property in either of these wards, the workflow we recommend internally is:
- Identify the closest station and check its position in the table above.
- Compare the asking GRM (price per m2 divided by expected monthly rent per m2) to the station median.
- If the asking GRM is below the station median, the deal is potentially worth deeper diligence. If it is above, you are paying a premium that needs a clear thesis (renovation upside, layout differentiation, off-market discount on resale).
- For Sakyo specifically, compare your station's GRM to neighboring stations on the same line. The Eizan line in particular shows large yield gaps between adjacent stops.
This is a standard station-level deal screen and it removes most of the guesswork that comes from city-level or ward-level averages.
Related reading
- [Why station-level data matters](/blog/why-station-level-data). The structural argument for analyzing Japan at station rather than ward level.
- [Ward vs station pricing in Tokyo](/blog/ward-vs-station-tokyo-pricing). Same logic applied to Tokyo's 23 wards.
- [Kyoto: what a 512x demand ratio reveals](/blog/kyoto-real-estate-in-what-a-512x-demand-ratio-reveals-2026). The Kitayama analysis that originally surfaced Kyoto's intra-city dispersion.
FAQ
What is a good GRM in Kyoto?
In our framework, anything below 192x is a strong yield, 192x to 264x is normal for residential Kyoto, 264x to 320x signals a tightening market, and above 320x signals premium or compressed yield territory. Sakyo's student belt sits in the strong band. Higashiyama's tourist core sits at the upper end of the compressed band.
Why is Sakyo's spread so much wider than Higashiyama's?
Sakyo has multiple independent demand drivers (students, families, premium villa buyers, temple proximity). Higashiyama has one (tourism plus historic preservation). Multi-driver markets always produce wider station-level dispersion than single-driver markets.
Are these gross yields or net yields?
Gross. They do not account for property tax, building management fees, vacancy, or income tax for non-residents. Realistic net yields in Kyoto residential typically run 1.5 to 2.5 percentage points below the gross figure.
Is the data from a single point in time or a rolling window?
The active listings are pulled fresh weekly. Sales transaction medians use a rolling window of the most recent 12 months of confirmed transactions in the dataset.
Can I get the same analysis for other Kyoto wards?
Yes. We track all eleven Kyoto wards in our internal database. Nakagyo, Shimogyo, Kita, and Fushimi all show distinct station-level dispersion patterns. If you are evaluating a specific ward, contact us directly.
Tokyo Insights is an independent, fee-only real estate advisory firm. We do not earn commissions. Our analysis is built from a proprietary database of active rental listings and sales transactions across Tokyo, Fukuoka, Yokohama, and Kyoto, refreshed weekly.