GUNOSY INC

Company Overview

Gunosy is a Japanese digital media company built around smartphone information distribution. Its legacy franchise is the Gunosy news-curation app, but today the group is broader: it also operates News Lite, the au Service Today portal app with KDDI, game-related media through Game8, a game publishing business through G Holdings, smaller B2B and commerce initiatives, and a meaningful investment portfolio that includes a stake in India’s slice Small Finance Bank.

The latest clean official annual base is FY2025. More recent data is partial, delayed, unaudited, or estimated. The audited FY2025 annual securities report was filed on 2025-08-27. The newest official company update is the FY2026 full-year earnings release dated 2026-07-15, but that is an unaudited earnings release, not yet an audited annual filing. Current market data in this report uses the supplied 2026-07-17 price of ¥350 and market capitalization of ¥8.38bn.

On the audited FY2025 base, Gunosy reported revenue of ¥6.10bn, operating income of ¥575m, and net income of ¥78m. On the FY2026 unaudited full-year earnings release, revenue rose to ¥6.44bn, but operating income fell to ¥186m and net income slipped to a ¥11m loss. The supplied deterministic snapshot marks net cash as unavailable, so I will not invent a precise net-cash figure. As a balance-sheet cross-check only, FY2025 audited cash was ¥3.99bn against ¥266m borrowings, and the FY2026 earnings release showed ¥5.44bn cash/deposits against only ¥71m borrowings.

This matters because Gunosy is no longer an earnings-led story. The stock sits between a weak operating franchise and a still-heavy asset base. That makes data freshness and source discipline more important than usual.

How the Company Makes Money

The core profit engine is still media and advertising. Gunosy uses its apps to gather user attention and then monetizes that attention through ad products, mainly CPC and CPM formats. Historically, the pitch was that its recommendation engine could deliver a more relevant audience than a generic display network. In practice, that means the business lives or dies on user habit, engagement, ad demand, and traffic acquisition efficiency.

The economics are now mixed. On the FY2026 unaudited segment view, the Media business generated ¥5.47bn of revenue and ¥1.16bn of segment profit. That is still a real business. But the rest of the group consumed most of it: G Holdings generated ¥944m of revenue and a ¥212m segment loss, New Businesses generated only ¥30m of revenue and a ¥198m segment loss, and corporate overhead took another ¥566m. Group operating profit therefore collapsed to ¥186m.

There are three important fragilities. First, the legacy news/app business is structurally dependent on distribution and user habit, not on hard-to-replicate intellectual property. Second, major-customer concentration still matters. KDDI was 17.2% of FY2025 audited revenue and about 11.9% of FY2026 unaudited revenue. Third, the group now behaves partly like a capital allocator: it owns large investment securities and has used M&A to buy adjacent businesses. That can create upside, but it also means weak operating returns can be masked for a while by balance-sheet strength.

Capital intensity in the media business itself is low. The problem is not heavy maintenance capex. The problem is that low-capex digital media with weak user lock-in is easy to erode. Gunosy is therefore not a high-return compounder. At best, it is a cash-rich, option-laden platform with one still-profitable segment and several areas that are consuming that profit.

Why the Stock Fell

The immediate trigger was straightforward. After the close on 2026-07-15, Gunosy reported FY2026 results and simultaneously disclosed that its business alliance with KDDI will end on 2027-03-31. On 2026-07-16, the stock hit ¥388, down 17.1% in one day. By the supplied market date of 2026-07-17, the price was down further to ¥350.

The reported facts justified a sharp repricing. FY2026 revenue rose 5.6%, but operating income fell 67.7% to ¥186m, and the company posted a small net loss. More damaging was the forward message: FY2027 company guidance called for revenue of ¥5.09bn to ¥5.26bn, an 18.4% to 20.9% decline, operating income of only ¥50m to ¥250m, another net loss, and a dividend cut to ¥13 from ¥22.

The market was also reacting to a pattern, not just a single print. In the Q1 FY2026 earnings release on 2025-10-15, management had guided to ¥7.89bn of revenue and ¥780m of operating income for FY2026. By the Q2 FY2026 release on 2026-01-13, that had already been cut to ¥6.45bn and ¥250m. Actual FY2026 revenue roughly matched that reduced sales guide, but operating income still came in below it and net income missed it outright. That sequence hurt management credibility.

The dominant market narrative is therefore not “one bad year.” It is “the legacy distribution model is deteriorating, the KDDI shelter is ending, and AI is accelerating the erosion.” That narrative has teeth because management itself now describes AI-driven changes in information consumption as a central strategic issue.

What the Market Is Assuming

Some of the bad news is temporary. FY2026 included ¥69m of impairment losses, ¥60m of investment-security valuation losses, and higher upfront spend in the game business. Game title timing also slipped. Those items can reverse.

But the market is not paying attention only to one-offs. The medium-term record is poor. Revenue peaked at ¥15.0bn in FY2019, then fell to ¥6.10bn on the audited FY2025 base. FY2026 unaudited revenue improved to ¥6.44bn, but that recovery did not translate into durable group profit. Net income was ¥2.01bn in FY2019, then -¥1.15bn in FY2023, -¥1.19bn in FY2024, +¥78m in audited FY2025, and -¥11m in the FY2026 unaudited earnings release. The supplied deterministic 12-year block tells the same story in slower motion: revenue CAGR 6.67%, but net-income CAGR -5.41% and owner-earnings CAGR 1.25%.

The long-term threat is more serious: Gunosy’s original edge was curation and distribution. That edge weakens when information discovery shifts toward large platforms, social feeds, video, and AI-generated answers. Management is now explicit that news media may be under structural pressure from generative AI. The KDDI breakup reinforces that concern because it removes a privileged distribution and revenue relationship just as the broader model is being questioned.

There is also a valuation assumption embedded in the current price. Using the FY2026 unaudited earnings-release balance sheet, cash/deposits of ¥5.44bn plus investment securities of ¥4.89bn minus total liabilities of ¥1.75bn equals about ¥8.58bn. That is slightly above the current ¥8.38bn market cap. In plain English, the market is close to valuing the whole company at its financial assets net of liabilities and giving the operating business roughly zero value.

That does not automatically mean the market is wrong. It means the market is assuming two things at once: first, that the operating business is weak enough to deserve little or no franchise value; second, that the investment portfolio deserves a discount because it is not the same thing as cash in the bank. I think both assumptions are broadly rational.

Temporary or Structural?

The key distinction is this: the core news franchise has essence damage, but the company does not. The business can survive because the balance sheet is still strong, the media segment still generates profit, and the KDDI revenue stream does not vanish overnight. But the original moat is weak.

The time-as-a-moat test is unfavorable. In two years, a well-funded competitor can replicate a news app. In five years, AI assistants, large platforms, and video-first feeds can make simple curation even less differentiated. In ten years, there is no clear reason to believe the old Gunosy format is defensible. Game8 has somewhat better staying power because content depth, SEO, title-specific know-how, and editorial workflows matter more there, but even that is not a fortress. The overall group moat is weak.

Is the Market Wrong? By How Much?

The supplied deterministic block gives two very different signals. On worth-dead, FY2025-based NCAV is ¥5.30bn, only 63.2% of the current market cap. That is not a classic deep net-net. On worth-alive, FY2026-based book value is ¥10.75bn and P/B is 0.78x, but modelled owner earnings are only ¥187m, an owner-earnings yield of 2.23% with a model payback of nearly 45 years. That is too weak to support the equity on operating earnings alone.

So the stock is neither a clean liquidation bargain nor a cheap operating franchise. It is an asset-backed, low-moat, low-visibility situation. My base view is that the market is roughly right: it is correctly skeptical about the business but may be a bit too harsh in assigning almost no value to the still-profitable media operations and the option value in the investment portfolio.

Moat-and-mispricing score: 4/10. Weak moat, modest possible mispricing, limited confidence.

The valuation below is a hybrid asset-plus-franchise estimate. It uses the supplied deterministic block as the anchor for NCAV, book value, and owner earnings, and it uses the newer FY2026 earnings release only as a separate reconciliation for cash, investment securities, and liabilities. Because the authoritative deterministic snapshot marks net cash as unavailable, I do not use a separate net-cash adjustment; instead I explicitly build balance-sheet backing into each case.

The most important point is not the exact yen figure. It is the source of value. Almost all defensible value comes from asset backing, not from current earning power. If you want a business-led thesis, Gunosy does not offer it. If you want an asset-backed special situation with some optionality, the stock is arguable, but the margin of safety is thinner than the book-value discount suggests.

Key Facts, Estimates, and Judgments