A cheap stock behind a language wall

Kabu Research · A worked example

Somewhere on the Tokyo Stock Exchange, a profitable company earns money, pays a dividend, and trades near half of book value. Below is the shape of the work we did on it, shown before the subscriber-only name and call.

In most markets, a stock like that would have analysts arguing over it. This one has none that we can find in English.

The gap exists because nobody can read the filings

This company's primary documents — its annual securities report (the 有価証券報告書, or Yuho), its mid-term plan, its daily exchange disclosures — are published in Japanese, and only in Japanese. No issuer English version exists today. That is the opening.

It is not obscure because it is bad. It is obscure because almost nobody outside Japan can read the documents that would tell you what it is worth. That language gap is not a small inconvenience. It is a structural wall, and behind it sits a large cohort of profitable, cash-generating, under-followed Japanese small- and mid-caps that the global market has simply never priced — what we call the orphan cohort.

The cheap headline is the invitation. Anyone with a screener can see a number that looks like half of book. What the screener cannot tell you is whether that discount is a mistake — a profitable business the world can't read — or a verdict — a balance sheet quietly hiding a problem. Answering that question is the whole job. And you cannot answer it from a screen. You have to go read four years of Japanese filings.

So we did.

What the work looks like

We won't tell you the name or the numbers here — that's the point of this piece. But we'll show you the shape of the work, because a publisher without a public track record yet should not ask you to take "trust us" on faith.

The process starts with original Japanese filings, not summary data. Kabu reads the source, compares the years, flags what changed, and keeps each claim tied to the underlying document. The goal is simple: get from "I've never heard of this company" to "I know whether this is worth my time" without asking you to trust a black box. Here's what that looked like on this one company:

• Four straight fiscal years, pulled from the company's 有価証券報告書 and stacked — not a one-year screen, and not a quick translation pass. Reading four years changed the story. The single-year snapshot every screener shows was misleading in both directions — one year made it look like a melting ice cube, another like a runaway rebound. Neither was the truth. Only the four-year shape was.
• We read the capital-allocation record from the cash-flow statement and the exchange disclosure feed — not from a story management tells. We found something specific about what this company does with its money, and, just as importantly, something specific it conspicuously does not do — an absence that turns out to be one of the most important facts in the whole analysis. (In Japan's current governance-reform wave, what a cheap company is not doing on the tape is often the tell.)
• We tested how a competitor would attack the business. Instead of admiring the business, we looked at the playbook against it, the company's counter, and what the record says about who's winning. That showed exactly where returns are fragile, and it wasn't where most people would guess.
• We ran a supply-side stress test on the real inputs that drive this business, asking what a sharp adverse move in each would do to the thesis. That test relocated the risk from a vague "it's cyclical" to a single, monitorable sentence about what would actually break it.
• Every figure in the full note ties to a specific filing — by disclosure ID and date. Every retrieved piece of evidence is pinned to the exact document, passage, and retrieval record, so any claim can be traced back to source and re-checked later. Where a number could not be verified to a primary document, we marked it a gap rather than print it. We don't retype figures off a summary site and hope.

That is the difference between a screen hit and a decision. The cheap number is free and on every screener on earth. Reading four years of primary filings in the original language, naming the attacker, stressing the cost side, and citing every figure to a disclosure ID — that's the part that's hard, and that's the part nobody else is doing on these names.

So — is it a buy?

Here is the honest answer, and it's the most important thing on this page: a disciplined process says no and not yet far more often than it says yes. A half-of-book price is where the work starts, not where it ends. Sometimes the discount is a gift. Sometimes it's the market being right for a reason you only find on page 40 of a Japanese filing.

We reached a conclusion on this specific company — a clear call, with the named conditions that would change it, and a full fair-value range checked more than one way. That call, the fair value, the name, and the ticker are what subscribers get.