Japanese small-cap earnings, in plain English — results vs. the company's own forecast.
For the Standard & Growth small-caps that never get an English release, Hayadex turns each new earnings filing into a clear English scoreboard: actual results next to the company's own forecast, year-over-year, and progress toward the full-year plan — read from the official filing data, not guessed by an AI. The wording? Hand it to your own ChatGPT or Claude. The numbers are us. No more machine-translating PDFs one by one, every quarter.
Join the waitlist — it's free → Tell us what you'd pay. 2 quick questions.
A real filing, in English: actual results next to the company's own guidance — beat or miss, YoY change, and (for quarterly reports) how far they are through the full-year plan. Company guidance = the full-year forecast the company itself publishes, not analyst consensus.
…and your whole watchlist in one view: every holding's latest earnings side by side — vs. last year and vs. each company's own forecast, sortable in a click to surface the biggest beats and misses.
The problem
Since April 2025, large Japanese companies (Prime Market) must disclose earnings in English — about 99% now do. But the small- and mid-cap value stocks overseas investors hunt for — the 1,500+ companies on the Standard & Growth markets — have no English-disclosure requirement. Their earnings often arrive late in English, or only in Japanese.
JPX's own 2025 overseas-investor survey flags exactly this: dissatisfaction with delayed English disclosure for small & mid-cap stocks.
Why not just TDnet + ChatGPT?
You can do it for free: watch TDnet, open the right Japanese PDF, machine-translate it, and paste it into ChatGPT. It works — but every quarter, for every name, it is slow, and you still have to line up the numbers against what the company itself had forecast.
Hayadex does that number part for you: it watches your list, reads the official figures from each filing, and lays them out in English against the company's own guidance. Your AI handles the words; Hayadex handles the numbers. Translation stays with your own ChatGPT or Claude — we never charge you for it.
How Hayadex works
- Add your watchlist — the Japanese tickers you follow (including Standard & Growth names).
- We read the numbers for you — for each new earnings filing (TDnet), Hayadex pulls the official figures from the filing data and lays them out in English: actual vs. the company's own forecast, YoY, and progress to plan.
- Your own AI handles the words — one click opens the filing in your ChatGPT or Claude to translate & summarize. No API key needed — it just opens your own AI in a new tab; we never see your conversation or keys, and never bill you for it.
The figures are read from each filing's official structured data (XBRL) — not guessed by an AI. We aim for accuracy, but always check anything important against the original filing. We link to official public disclosures and don't resell anyone else's translations.
See it in action
Click 🤖 AI on any filing — Hayadex pulls the disclosure text, wraps it in a translation & summary prompt, and opens it in your own ChatGPT or Claude. The figures scoreboard up top comes from Hayadex; the translation is your AI.
Free to start
Follow a few tickers for free. For your full watchlist there's one paid plan — and we're setting its price together with early users. That's what the question in the form is for.
Join the waitlist
We're validating demand before launch. Leave your email to get early access — and tell us what it's worth to you. Two quick questions; the rest is optional.
Built by an independent developer testing whether this is worth making — not a big company, no spam, just honest early-access updates. You can ask to be removed anytime.
We'll be in touch about early access. Mind sharing this with a fellow Japan-stock investor?