How much did US tariffs hurt Japanese companies?
In 2025 the United States began imposing additional tariffs on imports from Japan. We counted every listed company that mentions tariffs under Business Risks in its annual securities report from FY2017, and AI read each paragraph to sort out how tariffs reach the company. Then we checked the impact companies report themselves and their operating profit over the year the tariffs applied.
It rose 2.4 times in a single year.
Listed companies mentioning tariffs under Business Risks grew slowly from 127 (3.75%) in FY2017 to 204 (5.17%) in FY2024. In FY2025 the count reached 484 (12.39%), 2.4 times the previous year.
In FY2026, 368 of the 2,788 reports filed so far (13.2%) mention tariffs. Among the same companies, however, the count fell from 391 to 362: 46 added it and 75 dropped it.
The growth is in tariffs on own trade and effects through the economy and customers.
AI read each paragraph mentioning tariffs and sorted it into the five types below.
1. Tariffs on the company's own trade
2. Through customers (lower production and demand)
3. Through the economy and markets
4. Tariff systems and customs (other than additional tariffs)
5. Other (name only)
| FY | Mentioning | 1 | 2 | 3 | 4 | 5 | 1 share |
|---|---|---|---|---|---|---|---|
| 2017 | 127 | 11 | 2 | 0 | 36 | 80 | 9% |
| 2018 | 146 | 14 | 3 | 0 | 39 | 93 | 10% |
| 2019 | 156 | 21 | 5 | 0 | 40 | 94 | 13% |
| 2020 | 187 | 40 | 5 | 4 | 44 | 99 | 21% |
| 2021 | 187 | 40 | 10 | 1 | 43 | 97 | 21% |
| 2022 | 193 | 42 | 9 | 1 | 47 | 100 | 22% |
| 2023 | 195 | 38 | 9 | 1 | 51 | 101 | 19% |
| 2024 | 204 | 49 | 12 | 1 | 48 | 100 | 24% |
| 2025 | 484 | 162 | 64 | 94 | 44 | 177 | 33% |
| 2026* | 368 | 134 | 32 | 53 | 30 | 163 | 36% |
* FY2026 is an early count (2,788 reports filed so far, mostly March year-ends).
From FY2024 to FY2025, type 1 rose from 49 to 162 companies, and companies writing type 2 or 3 rose from 13 to 154. Type 4, covering safeguards and customs, barely moved, from 48 to 44.
Many FY2025 reports were written while the course of the tariffs was still unclear. Among FY2026 reports filed so far, companies writing type 2 or 3 fell to 84 (23% of those mentioning tariffs), and type 1 stood at 134 (36%).
Half of steel companies, four in ten in transportation equipment.
The share writing it in FY2025 is highest in iron and steel (47.4%), transportation equipment (38.8%), electric appliances (34.8%) and machinery (29.1%), all export-heavy manufacturers. Banks rose from none in FY2017 to 15 (18.3%). Sumitomo Mitsui Financial Group estimated the effect of tariffs and booked an additional ¥46.5 billion of loan-loss provisions: tariffs reach banks through their borrowers.
| Industry | 2017 | 2025 | 2025 Share |
|---|---|---|---|
| Iron & Steel | 5/38 | 18/38 | 47.4% |
| Transportation Equipment | 6/78 | 31/80 | 38.8% |
| Electric Appliances | 20/215 | 80/230 | 34.8% |
| Machinery | 15/200 | 62/213 | 29.1% |
| Nonferrous Metals | 2/31 | 9/33 | 27.3% |
| Chemicals | 7/186 | 39/200 | 19.5% |
| Glass & Ceramics Products | 2/50 | 10/52 | 19.2% |
| Banks | 0/70 | 15/82 | 18.3% |
| Precision Instruments | 1/47 | 9/51 | 17.6% |
| Foods | 13/117 | 21/124 | 16.9% |
TSE 33 industry groups. Only industries with at least 30 companies in FY2025, sorted by the FY2025 share.
Tariffs stopped at the automakers. Most parts makers grew profit.
Tariffs hit almost all imports, but rates differ by product. Cars and parts faced 27.5% from April–May 2025, cut to 15% on September 16 under the Japan-US agreement. Steel and aluminum have faced 50% since June 2025. Most other goods faced 15% from August 2025, with pharmaceuticals and semiconductors exempt. These reciprocal tariffs were struck down by the US Supreme Court in February 2026, and a 12.5% rate under a different law now applies.
Cars took the biggest blow. Exports to the US fell 4.1% in 2025, and cars fell 11.4%. Comparing the same companies over the year the tariffs applied, all eight automakers saw profit fall, and their combined operating profit dropped from ¥7.7 trillion to ¥4.4 trillion. Yet among 48 parts makers, only 12 saw profit fall, and median profit growth was 20.9%. Operating margin fell from 7.4% to 4.1% at automakers but rose from 5.0% to 5.4% at parts makers.
| Group | Cos. | Fell | Growth | Profit (¥tn) |
|---|---|---|---|---|
| Automakers | 8 | 8 | -33.5% | 7.7→4.4 |
| Parts makers | 48 | 12 | +20.9% | 1.1→1.3 |
Fiscal years ending December 2025 to July 2026 against the previous year. Fell: companies whose operating profit fell. Growth: median, only for companies profitable in the previous year. Profit: total operating profit in each group.
The effects companies state in their reports show the same gap. Toyota reports a ¥1.38 trillion effect from US tariffs, more than its ¥1.0293 trillion fall in profit. Mazda absorbed ¥154.9 billion of tariff costs and still made ¥51.6 billion of operating profit. Honda's tariff effect was ¥346.9 billion; most of its decline came from ¥1.4536 trillion of EV-related losses.
| Company | Effect | Profit | Share |
|---|---|---|---|
| Mazda | ¥154.9bn | ¥51.6bn | 75% |
| Toyota Motor | ¥1,380bn | ¥3,766bn | 27% |
| F.C.C. | ¥2.2bn | ¥18.9bn | 10% |
| Takeuchi Mfg. | ¥3.2bn | ¥37.7bn | 8% |
| Isuzu Motors | ¥16.0bn | ¥203.7bn | 7% |
| NTN | ¥2.1bn | ¥31.0bn | 6% |
| Yamaha | ¥1.8bn | ¥29.3bn | 6% |
| NHK Spring | ¥1.7bn | ¥45.8bn | 4% |
| Sumitomo Electric | ¥5.0bn | ¥418.2bn | 1% |
Effect: FY2026 tariff effect as written in each annual report; wording differs by company. Profit: operating profit. Share is effect ÷ (operating profit + effect), a rough guide to how much of the profit without tariffs was lost.
Parts makers pay tariffs too. In August 2025 Denso foresaw a hit of up to ¥130 billion, and F.C.C. reports a ¥2.2 billion effect. The main reason their profit held up is probably that they recovered the burden from automakers. Nikkei xTECH reports that Denso, Aisin and others said they would recover all of the remaining effect. In August 2025 Toyota booked ¥400 billion for parts, including some support for suppliers, within its expected ¥1.4 trillion tariff effect.
Annual reports show traces of this recovery. Denso cites one-off cost recovery among the reasons for higher North American profit, Tirad says customers came to better understand the tariff burden, and Nihon Plast mentions some progress in recovering higher tariff costs. Ikuyo describes the industry as a whole: automakers are struggling, while parts suppliers are supporting earnings with price pass-through and restructuring. Recovery is not the only reason for higher profit, of course: at F.C.C., lower product warranty provisions also helped.
Annual reports cover only listed companies, however. Even among listed parts makers, Nihon Plast's sales fell in part because customers cut production. Whether unlisted second- and third-tier suppliers recovered costs the same way cannot be told from this count.
Steel took the hit through a different route. Its 50% rate is the highest, yet median profit growth across 38 steel companies was minus 18.1%, and 55.3% saw profit fall. JFE Holdings names the US as a market where it exports little, then worries that steel shut out of such markets will flow into its main export regions. Nippon Steel writes that US tariff policy, among other factors, could worsen global steel supply and demand and lower prices. Tariffs probably arrived as falling prices more than as a direct cost. Both companies also cite Chinese overproduction, so how much of the decline comes from tariffs is unclear.
Across other industries the effect fades further. Of the 145 companies writing about tariffs on their own trade, 34.5% saw profit fall, the same as the 34.7% of the 2,699 that did not mention tariffs. Median profit growth was 6.4% against 10.2%, so tariffs were not without effect, but most companies probably absorbed them.
For some, the damage was smaller than feared. Mazda's expected hit of over ¥230 billion came in at ¥154.9 billion, and Honda's initial ¥650 billion became ¥346.9 billion. Of the 123 companies writing in FY2025 about effects through the economy or customers, 45 (37%) no longer mention tariffs in FY2026. Among the 128 writing about their own burden, only 15 (12%) did so.
After the February 2026 ruling, refunds of reciprocal tariffs have begun, and Nomura Securities estimates refunds to Japanese companies at ¥400–500 billion. That calculation excludes autos and steel, so it does not cover the tariffs automakers paid, which were the largest burden.
- Source:JETRO, US tariff measures (reciprocal rates and exemptions, August 22, 2025, Japanese)
- Source:US Federal Register, implementing tariff elements of the US-Japan agreement (cars and parts at 15%, September 16, 2025)
- Source:JETRO, US Section 301 additional tariffs (12.5% in total for Japan, July 24, 2026, Japanese)
- Source:nippon.com, Japan's 2025 trade statistics (based on Ministry of Finance data; Japanese)
- Source:Nikkei xTECH, how 16 auto parts makers plan to handle tariffs (August 26, 2025, Japanese)
- Source:Nikkei, Toyota's ¥1.4 trillion tariff effect for the year to March 2026 (August 7, 2025, Japanese)
- Source:Newswitch, full-year outlook of seven major Toyota group parts makers (August 4, 2025, Japanese)
- Source:Newswitch, Honda's tariff effect cut from ¥650 billion to ¥450 billion (August 8, 2025, Japanese)
- Source:Car Watch, Honda FY ended March 2026 results (tariff effect ¥346.9 billion, EV-related losses ¥1.4536 trillion; Japanese)
- Source:Nomura Securities, estimate of tariff refunds to Japanese companies (May 22, 2026, Japanese)
How we counted.
- Scope: listed companies (with a securities code) that filed an annual securities report for the fiscal year. Fiscal year is the year the period ends. Source: annual reports on EDINET (Japan FSA).
- We counted whether the Business Risks section contains the word 関税 (tariff or customs duty). Mentions of only trade policy or trade friction are not counted. Customs law and customs practice do count as the word, and are separated by type.
- FY2026 covers reports filed by mid-October 2026, mostly March year-end companies. December year-end companies are not yet included.
- AI read each relevant paragraph and assigned one of the five types; a company with several paragraphs can count in several types. A second AI re-read 60 FY2026 paragraphs: 43 were reasonable, 12 could be read either way, and 5 were misses. Types 2 and 3 are hard to separate, so the insight section treats them together.
- For the effects in yen, we extracted by machine every passage in any section with an amount near the word, then selected with AI. Table figures were checked against the reports. This is not a complete list of companies.
- The profit comparison covers companies with fiscal years ending December 2025 to July 2026 and operating profit for the previous year. Automakers are the eight: Toyota, Honda, Nissan, Suzuki, Mazda, Subaru, Mitsubishi Motors and Isuzu. Parts makers are the 48 left after removing automakers and companies whose main business is not auto parts (motorcycles, shipbuilding, rail, aerospace, special-purpose vehicles and so on) from the TSE transportation equipment group. AI looked up each company's business to choose those removed. Profit changes include factors other than tariffs and do not isolate their effect.
- The comparison with companies writing about their own tariff burden uses the AI types. Types reflect the wording in Business Risks, so companies with large effects are not necessarily in type 1.
- The FY2017 company count (3,391) is lower than the actual number of listed companies; some companies that later delisted may be missing.
Find this year's companies that wrote the same words.
Katana screens every listed Japanese company by words and conditions in its latest annual securities report. AI reads each report and judges qualitative conditions too.
Open the screenerJapan through annual reports
Japanese companies citing labor shortage as a risk rose eightfold in nine years Japanese companies mentioning AI in their risk disclosures rose eighteenfold in nine years Japanese companies citing rising interest rates as a risk rose 10% in a single year Has Japan fixed its low-PBR problem? Japanese companies hurt by a weak yen now outnumber those it helps See AI screening examples