25.7×
the Tokyo company’s net profit in the year Pokémon GO launched, from ¥619m to ¥15,921m. Both ends are filed.
- ¥15.5–22.2bn of that year’s pretax profit from GO, a model
- ¥18–21bn of Nintendo’s summer quarter that GO cannot explain
- 5.9–8.1× TPC’s growth since, depending on how the gap is read
In the summer of 2016 a mobile game made by someone else changed the size of The Pokémon Company in a single year. TPC did not build Pokémon GO and did not sell it. It licensed the name. Yet the licence fee alone turned a company that had never earned much more than two billion yen a year into one earning nearly sixteen. The same year left a hole in Nintendo’s accounts that is still there. Following both shows how the GO licence actually works: whom it pays, where it is taxed, and what it does not explain. The Pokémon GO and mobile page covers the deal as it stands today; this page is about the year it first landed.
What jumped, and what did not cause it
The gazette notice for that year is a bare balance sheet summary, with net profit noted at the foot. There is no income statement, no sales line, no tax line. Everything below is read from a handful of balance-sheet numbers.
Those numbers are consistent. Total assets rose ¥27.7bn. Of that, ¥15.9bn is new equity, which is almost exactly the year’s profit kept, and ¥11.8bn is new liabilities. Nothing else came in.
Not the merger. TPC absorbed a small sister company, Pokémon Communications, and the notice appeared in the gazette in January 2017. But the merger took effect on 1 March 2017, the first day of the next financial year. Even if it had landed inside the GO year, the absorbed company brought at most ¥580m of assets.
Not a new main-series game. The years of X and Y and of Omega Ruby and Alpha Sapphire, big launch years without GO, left the parent at $10.8m and $18.7m of profit. A game year alone does not move the Tokyo company much.
The recipient is named. When GO launched, Nintendo told the market that the Tokyo company, as the rights holder, receives a licence fee and a payment for helping develop and run the game. The original announcement carried a Tokyo dateline and named The Pokémon Company, Niantic and Nintendo, which “today announced their collaboration”. The Western arm, TPCi, was not named.
The fit is not perfect. The GO year was also the twentieth anniversary, and GO lifted Japanese cards too. Allowing for that, the non-GO part of the year’s pretax profit could be anywhere from ¥0.9–7.0bn, and GO the rest.
The step in Tokyo
The Tokyo company’s own net profit and current liabilities, ¥ billion, the year before GO and after.
- Tokyo company net profit
- Current liabilities at year end
Profit and liabilities jump together, then liabilities fall back the next year. That shape is the tax bill.
Built from .Show the numbers
| Year to February | Tokyo company net profit, ¥bn | Current liabilities at year end, ¥bn | Source |
|---|---|---|---|
| 2016 | |||
| 2017 | |||
| 2018 | |||
| 2019 |
Tokyo company net profit. The Tokyo company on its own, from the gazette notices: 25.7 times the year before in the year to February 2017. GO launched in July 2016, inside that year; this shows timing, not cause.
Current liabilities at year end. Same gazette notices, balance sheet at the end of February. They jump with the GO-year profit and fall back the next year, the shape of a tax bill accrued and then paid.
How a licence shows up in the accounts
With no income statement, the best evidence of what kind of money arrived is the tax. If the profit were a dividend from a subsidiary, Japanese rules would leave most of it untaxed. If it were licence income, it would be fully taxed. Liabilities tell the two apart.
They rose from ¥5.3bn to ¥17.2bn, then fell by ¥9.7bn the year after. That is what a big year-end tax bill looks like, paid off the following year. On the trainer’s notes’ model, tax due makes up 55–59% of the rise. The implied pretax profit is ¥22.5–23.1bn. At most ¥4.7–5.2bn of licence income could have been held back as deferred revenue.
A dividend from TPCi to Tokyo would not look like this. A new or repriced royalty from TPCi to Tokyo would, though, because it would be taxed the same way. That reading is not excluded. Nothing points to it either.
How much of GO landed in Tokyo. GO’s pretax contribution in the parent comes to ¥15.5–22.2bn, or 13–22% of what players spent on GO in vendor estimates. That is a model on a model. Vendor revenue, an assumed exchange rate and an assumed tax rate all go in.
The relationship held for three years. The parent earned ≈¥15–18bn of net profit for each billion dollars of GO player spending, in the launch year and in each of the two after it. A licence keyed to the game’s takings behaves like that. So does almost nothing else.
Could the Western arm have booked GO’s Western share instead? Tokyo’s step needs ¥15.5–22.2bn of GO income, which puts Tokyo’s share of TPC’s GO take at 44–100%. So TPCi booked at most a little over half of it, and perhaps none.
The owners took a little. The following year TPC paid out ¥2,989m, confirmed as a cash dividend by a rise in its legal reserve. That is 18.8% of the GO year’s profit. Nintendo’s share would be about ≈¥956m, but that is simply its stake times the total. No Nintendo document reports it. Nintendo’s own books only cap it at ¥1,363m. Game Freak and Creatures would each have received about ≈¥1.0bn, if the remaining stakes are equal.
The gap in Nintendo’s year
Nintendo owns 32% of TPC’s votes and books that share of its profit as it goes. In the GO year Nintendo booked ¥20.3bn. In the same filing it printed a summary table of its affiliates: ¥166bn of revenue across four companies. One of them was the Seattle Mariners’ operator, about 19% of that revenue.
The two numbers do not agree. Taking the ball club out, the table gives TPC net profit of ¥41.3–43.3bn. What Nintendo actually booked implies ¥62.9–66.4bn. At Nintendo’s share, its stake in the table is worth ¥13.2–13.9bn, leaving ¥6.4–7.0bn unexplained. At the whole-company level, that is a block of ¥18.8–25.0bn with no identified source.
It sits in one quarter. Nintendo booked ¥36m in the spring quarter and ¥12.0bn in the summer quarter, when GO launched. Grossed up, that summer figure implies ¥37.5bn of TPC profit in three months. That is more than twice the ¥15,921m the Tokyo company earned in the whole year.
GO cannot fill it. Players spent about $0.42–0.46bn in GO’s first three months. One widely quoted first-month figure is revenue “in net revenue globally on the App Store and Google Play”, after the stores’ cut, so the quarter here is built on what players actually paid. Even so, TPC’s take for the quarter comes to ¥9.1–14.2bn, or ¥12.5–19.5bn if the quarter is counted a month later. At the highest share the trainer’s notes tests it is ≈¥26bn. Something worth ¥18–21bn sits in that quarter on top of GO.
Nintendo was asked. On its results call an analyst asked whether the summer jump was launch revenue booked all at once. Nintendo said it was Pokémon-related income, GO included but not only GO, and that it was not revenue recognised temporarily at launch. That rules out a one-off lump at TPC. It equally rules out a one-quarter item on Nintendo’s side, so it does not settle which is true. The quarterly report carries no note on its affiliates at all. The next year’s annual report reprints the table unchanged.
Three readings survive. The table may simply leave out some of TPC’s ordinary business; that is the weakest, because the money would have to come from somewhere, and the budget of plausible sources falls ¥4–9bn short, though on rough estimates whose error is wider than that. A licence-type item may sit at TPC outside the table. Or Nintendo may have booked something of its own in that quarter, for which no mechanism has been found.
GO does not fill the summer quarter
Nintendo’s implied TPC profit for July to September 2016, against TPC’s GO take at each tested share.
- TPC’s estimated GO take, 21–30% cut
- TPC’s estimated GO take, 40% cut
- TPC’s quarterly profit implied by Nintendo
Even the tallest GO bar stops well short of the quarter.
Built from , and .Show the numbers
| Case | TPC’s estimated GO take, 21–30% cut, ¥bn | TPC’s estimated GO take, 40% cut, ¥bn | TPC’s quarterly profit implied by Nintendo, ¥bn | Source |
|---|---|---|---|---|
| Jun–Aug 2016, 21–30% cut | (9.1–14.2) | – | – | |
| Jul–Sep 2016, 21–30% cut | (12.5–19.5) | – | – | |
| Jul–Sep 2016, 40% cut | – | – | ||
| TPC profit implied by Nintendo, Jul–Sep 2016 | – | – |
TPC’s estimated GO take, 21–30% cut. Estimates: app-analytics firms’ gross revenue at an assumed ¥103 per dollar, times an assumed share for TPC.
TPC’s estimated GO take, 40% cut. Even at 40% the take is ¥11.5bn short of the quarter’s profit.
TPC’s quarterly profit implied by Nintendo. Nintendo booked ¥12.0bn from TPC in July–September 2016 at its 32% holding, so ¥37.5bn at the full TPC level.
What the year says about the licence
The GO year is the only time the licence can be seen working. It shows three things. The money goes to the Tokyo company, the rights holder Nintendo named. It behaves like a royalty on what the game takes in, steady per dollar of player spend across three years. And it is taxed as ordinary income, not routed through a subsidiary.
That makes it hard to watch from outside now. In the first years after launch the parent’s whole profit capped how big a Tokyo-booked GO rate could be. From FY2/22 cards and TCG Pocket are so large that the cap no longer binds. If Scopely, which bought Niantic’s games business in 2025, renegotiated the fee, the change would show only in a document, not in TPC’s accounts. Everything here also assumes the fee is keyed to app-store takings, and in 2025 a large share of GO’s money did not go through the app stores at all.
Nor will Nintendo’s filings help. Its related-party note was empty in the GO year. Under Japanese rules it covers only Nintendo’s own dealings, and only those above a tenth of sales or costs, so it could never show a royalty between TPC’s Western arm and Tokyo.
The GO-year gap
Nintendo’s share of TPC’s profit, year to March 2017, ¥ billion: as booked, and as its own affiliate table implies.
- Unexplained remainder, not a business line
- Total
Nintendo booked several billion yen more from TPC than its own affiliate table implies.
Built from and .Show the numbers
| Year to March 2017 | Nintendo’s share of TPC’s profit, ¥bn | Source |
|---|---|---|
| Implied by the affiliate table | ||
| Unexplained gap | ||
| Booked by Nintendo |
Nintendo’s share of TPC’s profit. Booked is Nintendo’s filed equity-method income. Implied (estimate): the affiliate table’s profit less the Mariners club and Warpstar/PUX gives TPC ¥41.3–43.3bn; Nintendo’s 32% of that is ¥13.2–13.9bn. The gap is the difference.