Skip to main content
Japan IR Transcripts are the records of what a Japanese company’s management said, as the company published them on its own investor-relations website: results-briefing and conference-call transcripts, the speaker script published with a deck, and the Q&A record of briefings, small meetings and investor dialogues. Japanese companies rarely hold a US-style earnings call with a word-for-word transcript, so this is usually the only record of what was said and asked.
Plan: Free and above · Credits: 1 per call

What you can use Japan IR transcripts for

Read what management told analysts after results, in the company’s own published words. Find the questions investors asked at a segment small meeting. Compare this quarter’s answers on a topic with last quarter’s. Pair a transcript with the deck shown at the same briefing.

Verbatim or the company’s summary

Read subtype before quoting. verbatim is the words as spoken or scripted (transcripts, “Presentation Handout with Script”, “Material with Speech Text”). qa_summary is the company’s own condensed account of the questions and answers (要旨, “Main Q&A”, “Q&A Summary”) — about 70% of rows — and must be quoted as the company’s summary, not as a speaker’s words.
What a row is was decided from its title, not from the section of the website it was listed under — sites file Q&A transcripts under “earnings” and trade shows under “conference”. site_type keeps the section the site used.

Japan IR website coverage

Coverage is a fixed set of about 250 Japanese companies whose investor-relations websites are read continuously — not the whole market. coverage.companies_covered gives the current count; coverage.available_from and available_to give the date range for the company you asked about. History reaches the late 1990s where a site keeps an archive, and only recent years for most.
company.in_ir_web_universe: false means FocusAlpha does not read that company’s investor-relations website. It never means the company published nothing. Many Japanese companies post only slides — try Japan IR Presentation Decks.
The endpoint covers Japanese companies only: any other company returns 404. If the website source cannot be read, the endpoint returns 503 rather than an empty list, so an empty data array always means the source answered and nothing matched.

Key Japan IR transcript fields

  • title and title_language (en or ja) — as the site wrote it, untranslated.
  • kind is transcript; subtype is verbatim or qa_summary.
  • page_url is the page the entry was found on; media_urls are the linked files.
  • documents — each linked document’s format (pdf or html), url and chars. has_text says whether any text was extracted.
  • ticker — the listing the website is known by, which can be the ADR (Sony is SONY).
About 30% of website rows have no extracted text (has_text: false); the links are still served. That means “not extracted”, not “empty”. The response also carries company (company_id, ir_site_tickers, in_ir_web_universe) and counts: matched for your filters, in_this_tool for all of this company’s transcripts, by_subtype, undated_excluded_by_date_filter and truncated_at_source.

Japan IR website date convention

Dates are as the website wrote them, and only a real calendar day becomes date. date_status says what the site gave: date_as_written keeps the original. Rows are returned newest first: dated rows, then rows with only a period_label, then undated rows. date_gte and date_lte apply to date only; rows without one are dropped from a ranged answer and counted in counts.undated_excluded_by_date_filter.

Japan IR transcript text

Pass include_text=true to add each document’s extracted text, capped at max_chars per document, with text_truncated. Text extracted from PDFs loses its layout: quote it, but take reported figures from Japan Financials.

GET /v1/companies/{company}/jp/ir-transcripts

string
required
Tokyo symbol (6758.T or 6758), five-digit EDINET security code, EDINET filer code, ADR symbol (SONY), ISIN or company_id of a Japanese company.
string
verbatim or qa_summary. Any other value returns 400.
string
Case-insensitive substring of the title, English or Japanese.
string
Earliest date, YYYY-MM-DD; date_lte sets the ceiling. Undated rows are excluded and counted.
string
default:"false"
true adds each document’s extracted text.
integer
default:"20000"
Text cap per document when include_text=true, between 1,000 and 200,000.
integer
default:"50"
Rows per page, between 1 and 200. Defaults to 10 when include_text=true. Page through with cursor, passing back next_cursor unchanged.
See also Japan IR Presentation Decks for the slides, Japan IR Website Disclosures for everything else on the site, Management Guidance for Japanese forecasts, and Company Calendar (market=JP) for upcoming briefings.