Daikokuya Japan Buyback — What a Shop Actually Paid

Unofficial — independent tool, not affiliated with, endorsed by, or sponsored by Daikokuya. It reads only publicly visible pages. Support, reliability guarantees and the full disclaimer are at the bottom of this page.

What it does: Type a reference like 126610LN and get what Daikokuya in Japan actually paid for that model, from its own archive of dated buyback records.

You enter: A reference number as the shop prints it, or a brand name in Japanese. Example: 126610LN.

You get: One row — how many the shop bought, the typical amount and lowest–highest paid in yen, the median for every year on record, the first and last day it bought one, the amount it advertises for that reference today by condition, how much of that brand's price page is actually published, and the gap between the two.

Price: $0.02 per model. No results = no charge. +$0.002 per row if you also want the individual records and the brand's price page lines (off by default).

Example: enter 126610LN → 256 purchases between 2020-09-20 and 2026-08-28 · typical ¥1,955,000 · range ¥1,350,000–2,483,000 · year medians ¥1,375,000 (2020) rising to ¥2,110,000 (2026) · advertised ¥2,250,000 unused today · so it paid 93.8% of its own headline (real run, 2026-09-10)

Unofficial — not affiliated with Daikokuya. Reads public pages only.

Why the buy side is the point

Every other price on the Japanese second-hand luxury market is an ask — what a seller wants. Chrono24, eBay, KOMEHYO, Mercari, the auction sites: all of them publish the sell side. A buyback shop publishes the bid, and Daikokuya publishes it twice over:

	What it is	Where it comes from
Paid	Individual purchases, each with the day and the yen	The shop's buyback record archive
Advertised	The amount it offers per reference right now, by condition	The brand's own buyback price page

The two together give you the thing neither one gives you alone: the gap between the headline and the cheque. On 2026-09-10 the Rolex page advertised ¥2,250,000 for an unused 126610LN, and the shop's own 2026 records put the median it paid at ¥2,110,000 — 93.8% of the headline. That is offerVsPaid.paidShareOfOffer.

The dated records also carry six years of history, so byYear is a market trend rather than a snapshot: for 126610LN the median rose from ¥1,375,000 in 2020 to ¥2,110,000 in 2026, +53%, and the count per year (2 → 41) doubles as a liquidity signal — how many of that model actually walk into this shop.

Some brands publish no price at all — and you are told so

The advertised price page is not evenly filled in. A row can say 「お問い合わせ」, "ask us", instead of an amount. Measured 2026-09-10:

Brand page	Rows	Rows with every amount published	coverageRate
Louis Vuitton	42	42	1.0
Chanel	74	74	1.0
Rolex	32	26	0.8125
Omega	45	0	0.0
Patek Philippe	39	0	0.0

On the two 0% brands every single row declines to publish an unused price. offerNow comes back empty for those, never as a zero and never borrowed from a neighbouring model. coverageRate and offerTable.columns are what tell you the difference between "this shop pays nothing for it" and "this shop does not publish that number" — the second is true, the first would be a lie. Omega's used column is separately published on 44 of 45 rows, and offerTable.columns says that too.

None of this touches the dated records. Those carried an amount on 256 of 256 rows for 126610LN and 401 of 401 for エルメス バーキン.

Output

One row per run, plus one row per record and per price-page line when the individual rows are on.

Measured on 2026-09-10 (real run, model "126610LN")

{
  "type": "buyback_summary",
  "model": "126610LN",
  "brand": "ロレックス",
  "brandSlug": "rolex",
  "category": "all",
  "status": "ok",
  "recordsFound": 256,
  "recordsRead": 256,
  "recordsComplete": true,
  "paidJpy": {
    "min": 1350000, "p25": 1840000, "median": 1955000,
    "p75": 2070000, "max": 2483000, "average": 1954687
  },
  "byYear": [
    {"year": 2020, "records": 2,  "median": 1375000},
    {"year": 2021, "records": 4,  "median": 1355000},
    {"year": 2022, "records": 5,  "median": 1750000},
    {"year": 2023, "records": 51, "median": 1870000},
    {"year": 2024, "records": 72, "median": 1905000},
    {"year": 2025, "records": 81, "median": 2000000},
    {"year": 2026, "records": 41, "median": 2110000}
  ],
  "dateRange": {"from": "2020-09-20", "to": "2026-08-28"},
  "offerNow": {"new": null, "unused": 2250000, "used": 2050000, "quoted": null},
  "offerBasis": "advertised-mail-in",
  "offerVsPaid": {
    "condition": "unused",
    "offerJpy": 2250000,
    "paidMedianJpy": 2110000,
    "year": 2026,
    "paidShareOfOffer": 0.9378
  },
  "buybackRateOfList": null,
  "coverageRate": 0.8125,
  "offerTable": {
    "brandSlug": "rolex",
    "url": "https://kaitori.e-daikoku.com/brand/brand/rolex.html",
    "updatedAt": "2026-09-09",
    "rowCount": 32,
    "quotedRows": 26,
    "modelMatched": true,
    "columns": [
      {"condition": "unused", "rows": 32, "priced": 26, "coverageRate": 0.8125},
      {"condition": "used",   "rows": 32, "priced": 32, "coverageRate": 1.0}
    ]
  },
  "hint": null,
  "sourceUrls": [
    "https://kaitori.e-daikoku.com/ex/archives/?category_key=none&freeword=126610LN&per_page=200",
    "https://kaitori.e-daikoku.com/brand/brand/rolex.html"
  ],
  "requestsUsed": 2,
  "disclaimer": "Unofficial - not affiliated with Daikokuya. ...",
  "fetchedAt": "2026-09-10T04:56:17.746524+00:00"
}

One dated purchase, returned only when Also list the individual records is on, at $0.002 each:

Measured on 2026-09-10 (real run, model "126610LN")

{
  "type": "buyback_record",
  "model": "126610LN",
  "boughtOn": "2026-08-28",
  "brand": "ロレックス",
  "modelName": "サブマリーナ デイト 126610LN",
  "modelRef": "126610LN",
  "paidJpy": 2050000,
  "disclaimer": "Unofficial - not affiliated with Daikokuya. ..."
}

One line of the brand's advertised price page, same toggle, same $0.002:

Measured on 2026-09-10 (real run, brand page "vuitton")

{
  "type": "offer_row",
  "model": "M41894",
  "brandSlug": "vuitton",
  "modelRef": "M41894",
  "offerNew": 103230,
  "offerUnused": null,
  "offerUsed": null,
  "offerQuoted": null,
  "askedConditions": [],
  "listPriceJpy": 111000,
  "rateOfList": 0.93,
  "offerBasis": "advertised-mail-in",
  "disclaimer": "Unofficial - not affiliated with Daikokuya. ..."
}

What the numbers are measured over

• paidJpy is money that changed hands, not an asking price and not an estimate: each record is one item this shop bought, on the day shown. There were no missing amounts on any search measured 2026-09-10.
• recordsComplete says true when the whole archive for that word was read in one call and false when there is more behind it. A reference number normally finishes (126610LN: 256 records, complete). A bare brand name does not — ロレックス passed 600 records in the most recent 3.3 months alone — so its figures are a recent sample, and the row says so.
• byYear buckets by the day of purchase. A record whose date cannot be read is left out of the year table and out of dateRange rather than landing in the newest year.
• offerNow is advertised, not agreed. offerBasis is advertised-mail-in because the shop states that the amounts it publishes are a mail-in guide and a branch may pay differently. Amounts printed as 「〜 ¥2,050,000」 are the top of a range; the number kept is that ceiling.
• offerVsPaid.paidShareOfOffer is paid ÷ advertised, so 0.9378 means the shop paid 93.8% of its own headline. It uses the newest year in byYear and the unused column where the brand publishes one; condition and year in the same object say exactly which two numbers were divided.
• buybackRateOfList is filled in only for the brands whose page prints a list price and a percentage on the same line — the bag and leather-goods brands. Louis Vuitton's page gives list ¥111,000 / 93% / ¥103,230 for M41894, and its whole table's median rate was 0.93 over 42 rows on 2026-09-10. It is null for every watch brand rather than computed from two numbers that were never meant to be divided.
• offerTable.updatedAt is the day the shop last edited that brand's page. All ten brands measured had been edited within the previous ten days (Rolex 2026-09-09, Vuitton 09-03, Chanel 09-01).
• A model that matches no line of the price page leaves offerNow empty and sets offerTable.modelMatched to false. Matching is on the reference only — two Rolex lines share the reference 126500LN under different dial names, and matching on the name would pick a watch a million yen apart.

An empty result is an answer, and it is free

A word this shop has never bought comes back as status: "not_found" with a hint, no amounts, and no charge. That is a real fact about a shop that has been buying for six years.

It is deliberately not the same thing as a failure. The site answers an empty search with a complete page carrying お探しの実績は存在しません — "the record you are looking for does not exist" — and that phrase, not a page size or a row count, is what this tool treats as an empty search. A page with neither that marker nor any record has changed shape, and the run fails and says so instead of reporting an empty market that was never measured.

Cost and load

• Two calls for a normal run: the record archive, then the brand's price page. A word the shop has never bought stops after the first, and is not charged.
• Calls are 2.0 seconds apart and capped at 12 per run, in code, with no input that can lower either.
• No login, no cookie, no browser, no proxy. 256 MB is enough.
• No image is ever fetched, and no image address is ever returned. The record photographs are named after the branch that made the purchase, and the shop's copyright page refuses use of its photographs; the parser has no image field at all.
• No prose is copied. What comes out is facts and arithmetic — brand, model name, reference, yen, date — and the statistics computed from them. No descriptions, no page text.
• There is no person in this data. The buyback archive is the shop's own record of what it bought; it has no seller field, no name, no contact detail, and this tool reads none of the site's customer-voice pages.

If it stops working

Everything here is one host, kaitori.e-daikoku.com, over plain HTTPS with no login and no cookie, so a failure is the site or the network rather than an expired session. The failure messages separate the three cases that need different answers: the site did not answer at all, the site refused the call and wants less load, or the page changed shape. The last one names which anchor went missing — the record container, the empty-search marker, or the price table — so it can be checked against the live page directly.

Disclaimer

Unofficial, independent tool — not affiliated with, endorsed by, or sponsored by Daikokuya. Brand names and references only say which item the numbers are about. Every figure is read from Daikokuya's own public pages at the moment shown in fetchedAt and is not an offer to anyone. Daikokuya states that the amounts it publishes on the web are a mail-in guide and that a branch may pay differently, and that condition, papers and accessories move the amount. Check before you act on any of it.

If something goes wrong

• Wrong number or a failed run? Open a ticket on the Issues tab. I read every one and reply within 2 business days (Japan time).
• You never get a fake "empty" result. If the site can't be read, the run fails and says so.
• No results = no charge. You only pay for results you actually get.
• Checked every week. An automatic test runs this tool weekly; if the site changes, I fix it.
• Public pages only. No login, no personal data, and it goes easy on the site.

More tools by the same author

• Mercari Japan Sold Prices — What Items Really Sell For
• Yahoo! Auctions Japan Sold Prices — Median, Range, Bids
• Hifido Japan Used Audio — What It Actually Sold For
• Japan Used Camera & Lens — Dealer Price vs Sold Price
• Japan Figure & Gunpla — New Price vs Used Sold Price
• Used iPhone & Android Prices Japan — Dealer vs Sold
• Yahoo, Mercari & PayPay Japan — Same Item, 3 Prices
• Mercari & Yahoo Japan Demand — Sold vs Still Listed

All tools (Japan marketplaces, real estate, jobs, racing, prediction markets): https://apify.com/jpmarketdata

Disclaimer

Unofficial, independent tool — not affiliated with, endorsed by, or sponsored by Daikokuya. Product names and logos belong to their owners and only say where the data comes from. Data is read from public pages, for market research; check before you act on it.