Japan's Cool Japan Fund Faces Abolition: The Numbers

According to METI materials published in February 2026:

The first thing to note is that although it is called a "public-private fund," the money is almost entirely public. Of ¥143.3bn in capital, ¥132.6bn is government money — 92.5%. The private ¥10.7bn is the combined contribution of 23 companies, capped as a rule at ¥500m each (¥100m for regional banks). Shareholders include ANA Holdings, Dentsu Group, Takashimaya, Bandai Namco, Toppan Digital and Isetan Mitsukoshi Holdings.

The fund was established in November 2013 under a dedicated act, starting with ¥30bn from the government and ¥8.5bn from the private sector, and grew through additional injections.

Despite the image of "Cool Japan" as anime and manga, the largest bucket by money is lifestyle, with media and content second. That gap matters for what follows.

METI publishes every disclosed deal. As of 17 February 2026 that is 65 companies across 78 investments. Below is the list by sector, with the business, the committed amount and the current status. Amounts are commitments (the maximum support figure), not the sum actually deployed; foreign-currency deals use the yen conversion given in the source, and multi-round deals show the total with its parts.

Media and content (16 deals / 15 companies)

Lifestyle (24 deals / 20 companies)

Food (17 deals / 15 companies)

Inbound tourism (16 deals / 10 companies)

Cross-sector (5 deals / 5 companies)

Exit dates follow the summary list in the METI document. The Kuala Lumpur department store (ICJ Department Store) is the one exception: the detail page of the same document states an exit in January 2019, while the summary list says July 2018.

One: the money was concentrated. The five largest companies account for ¥50.5bn, about 27% of the ¥188bn committed — Spiber ¥14bn, Ningbo Hankyu ¥11bn, Laugh & Peace Mother ¥10bn, Katana ¥8bn and SDI ¥7.5bn. This was not capital spread thinly across 65 companies. The ¥14bn in a single firm, discussed below, was the house style rather than an aberration.

Two: a large share went into funds, not operating companies. Eleven of the entries are investments into other investment or property funds, totalling roughly ¥43.3bn, about 23% of the whole — Japan Content Factory ¥5.15bn, Global Brain ¥5bn, Atona Impact Fund ¥5bn, Sinarmas-Spiral ¥6.2bn, and ¥11.4bn across two Japan Activation Capital vehicles. Rather than picking companies itself, the fund handed roughly a quarter of its money to other managers, and fund investments dominate the deals signed from 2024 onward.

Three: ten companies, about ¥34.4bn (18%), targeted Japan itself. Investments aimed at capturing inbound tourist demand fall into this category, so this is not inconsistent with the mandate. The two Japan Activation Capital vehicles (¥11.4bn combined) are worth noting separately: the source describes them only as "supporting long-term, sustainable growth and corporate value creation through investment in listed Japanese companies," with no stated link to overseas demand.

Four: new investments continued right up to the closure reports. The list above is dated 17 February 2026, but the fund's own site shows four further commitments after that.

METI settled on dropping the fund from its FY2027 budget request on 20 August 2026 — three months after the last of these. That is not in itself irregular, since the investment period runs to FY2028, but it shows the closure debate and the investment desk moving on separate tracks. As noted later, the fund is also under a duty to dispose of its holdings by 31 March 2034, so even these late commitments have to be sold within that window.

FY2024 produced a net profit of about ¥1.5bn — the first single-year surplus — and the cumulative figure beat plan. Then FY2025 delivered revenue of ¥4.4bn against a net loss of ¥15.6bn, pushing the cumulative loss to ¥54.0bn and missing the target by ¥11.4bn.

Here is the sharpest point. In its February 2026 materials, METI stated that the cumulative loss had "bottomed out in FY2024" and that recovery would begin in FY2025. One month later, in March 2026, the fund's largest single investment entered private restructuring and produced the biggest loss in its history. That outlook survived a single month.

The composition of the loss also deserves attention. For the −¥38.3bn as of end March 2025, the fund breaks it down as follows.

In other words, roughly 60% of the cumulative loss is not failed investment but the cost of running the organisation. Twelve years of personnel costs came to ¥10.1bn, about ¥840m a year; total operating costs averaged roughly ¥2bn a year.

The fund publishes average recovery multiples for exited deals by sector and by vintage (as of end March 2025, 29 exits). A multiple of 1.00 means the principal came back.

Only "lifestyle and other" returned the principal. Media and content — the very definition of Cool Japan — came last at 0.74x. The fund itself states that its overseas content distribution platform business returned ¥7.2bn less than it invested.

After a recovery to 1.32x in the middle period, the eight deals decided in 2019–2021 returned 0.30x — seventy percent gone. The pandemic hit that vintage, so the fund cannot be blamed alone, but the numbers show a stretch where investment judgement did not keep pace with a deteriorating environment. Deals decided in 2015 alone account for a shortfall of ¥7.3bn.

By size, deals under ¥1bn returned 0.76x (16 deals), ¥1bn–4bn returned 0.55x (9 deals) and ¥4bn–15bn returned 1.10x (4 deals): mid-sized deals performed worst, which runs against intuition.

Across all exits, ¥48.6bn invested returned ¥42.7bn — a shortfall of ¥5.9bn (0.88x), plus ¥2.4bn of dividend income from limited-partnership holdings.

The FY2025 loss came mainly from an investment in an advanced fibre materials start-up (Spiber). METI records an initial ¥3bn in November 2018 and a further ¥11bn in September 2021 — ¥14bn in total, an unusually large concentration.

The company's position deteriorated under pandemic conditions and rising input costs driven by the weak yen. Debt problems surfaced in December 2025 and it entered private restructuring in March 2026, with the business transferred at a low price.

Note where this sits: inside the lifestyle bucket, the largest by money. The biggest hit came not from anime or food but from a fibre materials venture. The policy banner and the destination of the money were not the same thing.

METI's materials contain a candid assessment of a department store project in Kuala Lumpur (¥1.07bn committed, announced September 2014, exited January 2019).

The phrase "by prioritising the policy significance" identifies the structural difficulty of any public-private fund. When judgement swings between policy purpose and financial return, neither gets achieved. This reads as a problem of institutional design rather than of any individual.

The Board of Audit made a related point in a May 2025 report. Under the revised plan, cumulative P&L in the final year (FY2033) is projected at a profit of only ¥1bn, against a cost of capital on the government's industrial investment of roughly ¥15bn. Even if everything goes to plan, the return falls ¥14bn short of what the public money cost.

• Private capital leverage of 2.1x as of end March 2025, already above both the 1.7x milestone for 2029 and the 1.7x target for 2034. Cumulative co-investment and lending from private companies reaches approximately ¥386.8bn.
• Policy KPIs met. The number of companies using services of portfolio companies reached 7,827, above the 7,037 milestone for March 2029. Business matching reached 139 companies, also above target.
• FY2024 net profit of about ¥1.5bn, the first single-year surplus, with cumulative investment P&L improving from −¥8.6bn to −¥3.5bn.
• On the ¥23.8bn of operating costs, the fund argues that "in private funds, necessary management costs are generally secured separately from investment P&L through management fees of around 2%." Its case is that it discloses inside its P&L what a private fund charges investors separately.
• Cost cutting has happened: office space was reduced by 40% in FY2021, servers moved to the cloud, and outsourced research narrowed.

How to read 2.1x is contested. Mobilising ¥386.8bn of private money is a real policy effect, but whether the fund acted as a catalyst and whether the public money comes back are separate questions. The Board of Audit asked the second one.

Headlines often say taxpayers lost ¥54bn. That is not accurate.

The ¥132.6bn of government capital came from industrial investment in the Fiscal Investment and Loan Program special account. The main revenues of that account are dividends on the government's holdings of NTT and JT shares and payments from the Japan Bank for International Cooperation — not income tax or consumption tax revenue routed from the general account.

That does not make it painless. Industrial investment income comes from state-owned assets, and it is money that could otherwise have funded different policy or been transferred to the general account to lighten the public burden. That opportunity cost is exactly what the Board of Audit measured with its cost-of-capital yardstick. See Japan's national burden ratio in international comparison for the wider frame.

• It was always time-limited. The statutory life is twenty years (FY2013–FY2033), new investment ends in FY2028, and the act requires the fund to endeavour to dispose of all holdings by 31 March 2034. An exit schedule existed regardless of the abolition debate.
• Live investments remain. As of end March 2025, 43 deals worth ¥88.2bn were still held, and nine were already behind their assumed exit timing.
• Closing early may reduce recovery. Reporting notes that abolition could leave part of the government's money unrecovered. Rushing sales invites low prices; waiting adds operating costs. Both paths cost money.
• Nothing is decided. The expert panel reports by year end. The Prime Minister's Office is reported to accept abolition, but the formal conclusion is still ahead.