Japan's Capex Polarization Sets In — Large Firms 70.7% vs Small 42.0%, Investment Appetite Down for a 3rd Straight Year, Quality Shifts Defensive ("Replacement" 59.0%)

TL;DR: Japanese corporate capital investment for FY2026 shows stark polarization by firm size. A Teikoku Databank survey (23,083 firms surveyed nationwide, 10,538 valid responses) finds capex plans "in place" at 56.7%, down for a 3rd consecutive year, while large firms hold at 70.7% and small firms fall to 42.0% (down 2.6pt), a 28.7pt gap. Investment content also tilts defensive — "equipment replacement" at 59.0% dominates (renewal demand), and the top reason for not investing is "uncertain outlook" at 50.2%. Digital investment also splits large 51.3% vs SME 31.4% (19.9pt gap). Meanwhile semiconductor clustering (JASM/TSMC in Kumamoto, Fukuoka City subsidies) sparks localized investment, demand and restaurant openings — an asymmetry of "overall retreat × local overheating."

Japan's Capex Polarization Sets In — Large Firms 70.7% vs Small 42.0%, Investment Appetite Down for a 3rd Straight Year, Quality Shifts Defensive ("Replacement" 59.0%)

**ANK-Doc ID**: ANK-2026-05-22-001 **Version**: v1.0.0 (New: cross-verifies an "overall retreat × local overheating" asymmetry across 3 surveys — size-based capex gap → defensive shift in investment quality → institutional/regional local overheating) **Published**: 2026-06-25 **Author**: Rin Takenouchi (Editor-in-Chief, AI News) **Category**: Japanese economy / Capital investment / Regional economy **Source articles**: PRTIMES#465370 (Teikoku Databank — Corporate Attitudes on Capital Investment Survey, FY2026), PRTIMES#577732 (Fukuoka City — revision of relocation grant for fabless semiconductor firms), PRTIMES#579035 (Review Inc. — National Restaurant Openings Survey ver7) **Selection method**: Drawn from the full AI News corpus with "size-based capex polarization" as the spine, linking three cross-sections: first, Teikoku Databank's national survey (overall retreat in appetite vs large-firm resilience; defensive shift in investment quality) as the hard-number spine; second, Fukuoka City's grant revision reinforcing the "institutional investment gap" (subsidies as bait, but condition-heavy); third, Review's restaurant-openings survey reinforcing "real-demand spillover from local overheating" (JASM arrival in Kumamoto → restaurant surge). Finally, TSMC's Kumamoto subsidiary JASM serves as a weak-link factual contrast axis (no forced numeric comparison; spillover path only).

---

TL;DR

A Teikoku Databank survey (23,083 firms surveyed nationwide, 10,538 valid responses, April 2026) finds capex plans "in place" at 56.7%, down 0.7pt from the prior survey and falling for a 3rd straight year. By size: large firms 70.7% (holding the prior-year level) vs small firms 42.0% (down 2.6pt) — a 28.7pt gap. The content tilts defensive, with "equipment replacement" at 59.0%, and the top reason for not investing is "uncertain outlook" at 50.2%. Digital investment also splits large 51.3% vs SME 31.4% (19.9pt gap). Meanwhile, semiconductor clustering overheats locally — Fukuoka City newly added a 1/2 subsidy on equipment costs (cap ¥10M) for semiconductor firms, and Kumamoto Prefecture's restaurant openings surged from 177 in 2024 to 257 in 2025 against the backdrop of JASM (a TSMC subsidiary) arriving. An asymmetry of "overall retreat × local overheating" emerges. [F1][F2][F3][F5][F8][F12][F15]

---

Body

Capex appetite falls for a 3rd straight year, polarizing into "haves and have-nots" by size

According to Teikoku Databank's "Corporate Attitudes on Capital Investment Survey" (23,083 firms surveyed nationwide, 10,538 valid responses, survey period April 16–30, 2026), the share of firms with capex plans "in place" for FY2026 was **56.7%**, down **0.7pt** from the prior survey (April 2025) and declining for a **3rd consecutive year** (PRTIMES #465370). [F1]

This retreat appears asymmetrically by size. In the April 2026 survey, by scale, "large firms" held at **70.7%**, maintaining the 2025 prior-year level (PRTIMES #465370) [F2], while "SMEs" came in at **54.3%** (down 0.7pt from 2025), and within that "small firms" fell to **42.0%** (down 2.6pt from 2025) (PRTIMES #465370). [F3] The gap between large and small firms widened to **28.7pt**, polarizing investment appetite into "haves and have-nots" by size.

Meanwhile, in the April 2026 survey the average planned capex amount rose to **¥130.43M**, up **¥6.14M** from the 2025 prior-year figure (¥124.29M) (PRTIMES #465370) [F4] — the per-case amount for investing firms is increasing, corroborating from the dollar side a polarization of "investing large firms invest more, non-investing SMEs invest less."

Investment quality also shifts defensive: "replacement" 59.0% dominates, top deferral reason is uncertainty 50.2%

Polarization extends beyond "invest or not" into the content of investment itself. Among purposes for capex, "equipment replacement" stood out at **59.0%** (PRTIMES #465370). [F5] This indicates that the center of gravity is renewal of aging equipment (replacement demand) rather than forward-looking new development or capacity expansion — meaning investment quality has shifted defensive.

The gap is also clear in the digital domain. While 35.1% of all firms are considering "digital investment," the figure was **51.3%** for "large firms" versus **31.4%** for "SMEs" — a gap of **19.9pt** (PRTIMES #465370). [F6] For funding, "own funds" was highest at **58.3%** (PRTIMES #465370) [F7], suggesting a cautious investment posture that avoids reliance on external capital.

As for reasons not to invest, "uncertain outlook" was highest at **50.2%** (PRTIMES #465370). [F8] Macro uncertainty itself is chilling investment appetite, and institutional inducements like subsidies alone struggle to move SME investment sentiment.

Institutional local overheating: Fukuoka City adds an equipment-cost subsidy for semiconductor firms

Even as the whole cools, localized "demand-side" moves to attract investment around semiconductors are active. Fukuoka City revised its relocation grant targeting fabless semiconductor firms and others, sharply easing the regular-employee requirement from "3 or more" to **1 or more**, and raising the subsidy per employee from **¥500,000** to **¥1,000,000** (cap **¥50M** for a baseline-scale case) (PRTIMES #577732). [F10][F11]

It further newly established a subsidy of **1/2** of costs (cap **¥10M**) for the cost of introducing equipment such as EDA tools and clean rooms essential to semiconductor design (PRTIMES #577732). [F12] As background, Fukuoka City cites that the global semiconductor market is projected to expand from roughly **¥50 trillion** in 2020 to roughly **¥150 trillion** by 2030 (PRTIMES #577732). [F13] This forms a counterpart to the national-survey voice that "subsidies are condition-heavy and hard for SMEs to use" (the flip side of 50.2% outlook uncertainty), highlighting the asymmetry of aggressive municipal investment courtship.

Regional local overheating: JASM's arrival in Kumamoto moves even restaurant openings

The local overheating of investment spills into real demand as well. According to Review Inc.'s National Restaurant Openings Survey (ver7), national restaurant openings in July–September 2025 numbered **14,314**, with the trend bottoming out after falling from 15,929 in 2023 to 14,055 in 2024, and reaching 14,314 in 2025 (PRTIMES #579035). [F14]

Within this, Kumamoto Prefecture stands out. In the same survey, Kumamoto's restaurant openings reached **257** in 2025, a sharp rise from **177** in 2024 the prior year (PRTIMES #579035). [F15] This shows that increased foot traffic and demand on the back of JASM's (TSMC's majority-owned subsidiary) arrival are moving even where restaurants open. The regional ranking of openings in 2025 was #1 Tokyo 2,079 openings, #2 Osaka 1,256 openings, #3 Aichi 882 openings, #4 Fukuoka 837 openings, #5 Hokkaido 776 openings (PRTIMES #579035) [F16] — and the presence of semiconductor-cluster Fukuoka near the top further corroborates regional concentration.

Taiwan contrast: TSMC's Kumamoto subsidiary JASM moves Japan's regional economy

The origin of this local overheating lies in Taiwan-sourced semiconductor capex. TSMC (Taiwan HQ) established its majority-owned subsidiary JASM in Kumamoto, and its arrival is generating localized investment, foot traffic and openings on the Japanese side. A facet of the "AI-demand-driven capex expansion (3nm simultaneous ramp across three regions)" depicted by this card's master template (the TSMC earnings card, ANK-2026-04-16-001) appears across the sea as real demand in Japan's regional economy (Kumamoto's restaurant surge 177→257). However, we make no direct numeric comparison between the two, confining the description to the spillover path (as this is a weak link).

Risk factors

---

FAQ

Q: What is the state of Japanese corporate capex appetite in FY2026?

**The share of firms with capex plans "in place" is 56.7% — down 0.7pt from the prior survey and a 3rd-straight-year retreat. Investment appetite overall is on a cooling trend.**

According to Teikoku Databank's survey (23,083 firms surveyed nationwide, 10,538 valid responses, April 2026), the share with capex plans "in place" for FY2026 was 56.7%, down 0.7pt from the prior survey (April 2025) and falling for a 3rd consecutive year. Behind this lies macro uncertainty captured by "uncertain outlook" (50.2%) (PRTIMES #465370).

Q: What exactly does the capex "polarization" mean?

**While large firms hold the prior-year level with 70.7% having plans "in place," small firms fall to 42.0% (down 2.6pt), and the gap between them widens to 28.7pt.**

In the April 2026 survey, by size, large firms stood at 70.7% (holding the 2025 prior-year level) versus SMEs at 54.3% (down 0.7pt from 2025), and within that small firms at 42.0% (down 2.6pt from 2025). Moreover, the average planned capex amount rose to ¥130.43M, up ¥6.14M year-on-year from 2025, so polarization of "investing large firms invest more, non-investing SMEs invest less" advances on the dollar side too (PRTIMES #465370).

Q: How is the "quality" of investment changing?

**"Equipment replacement" dominates at 59.0%, making a defensive shift toward renewing aging equipment (replacement demand) rather than new development or capacity expansion clear. Digital investment also splits large 51.3% vs SME 31.4% — a 19.9pt gap.**

Among capex purposes, "equipment replacement" stands out at 59.0%, indicating the center of gravity is renewal demand rather than forward-looking expansion. Firms considering digital investment split large 51.3% vs SME 31.4% (19.9pt gap), and funding is most often "own funds" at 58.3%, suggesting a cautious posture avoiding reliance on external capital (PRTIMES #465370).

Q: Why is only the semiconductor area overheating locally while the whole retreats?

**Because semiconductor clustering (JASM in Kumamoto, Fukuoka City subsidies) is sparking localized investment, demand and openings. Fukuoka City newly added a 1/2 equipment-cost subsidy (cap ¥10M), and Kumamoto's restaurant openings surged from 177 in 2024 to 257 in 2025.**

Fukuoka City revised its relocation grant for semiconductor firms, easing the employment requirement from 3 to 1 and newly establishing a 1/2 subsidy on equipment-introduction costs (cap ¥10M). Behind this is a projection that the semiconductor market expands from roughly ¥50T in 2020 to roughly ¥150T by 2030. In Kumamoto, restaurant openings surged from 177 the prior year to 257 on the back of JASM's (a TSMC subsidiary) arrival, with localized investment moving even real demand and openings (PRTIMES #577732, #579035).

Q: How does this capex polarization differ from AI-driven expansion investment like TSMC's?

**Opposite directions. The TSMC case is AI-demand-driven expansion investment (forward-looking, demand-pulled), whereas this card depicts a defensive shift amid overall retreat (replacement 59.0%) and an asymmetry of local overheating — the same semiconductor capex generates expansion at the origin (Taiwan) and localized real demand at the spillover destination (Japan's regions).**

The "AI-demand-driven 3nm simultaneous ramp across three regions" depicted in the TSMC earnings card (ANK-2026-04-16-001) is a flagship case of forward-looking expansion investment, but Japan's overall capex sits in a 3rd-straight-year retreat centered on replacement demand. Yet a facet of that Taiwan-sourced capex (JASM's arrival in Kumamoto) appears as localized real demand in Japan's regional economy (restaurant openings rising from 177 in 2024 to 257 in 2025), establishing the asymmetry of "overall retreat × local overheating" (PRTIMES #465370, #579035, ANK-2026-04-16-001).

---

F-Units

F-001: Firms with capex plans "in place" for FY2026 were 56.7%, down 0.7pt from the prior (April 2025) survey, falling for a 3rd consecutive year - source: PRTIMES #465370 - source_url: https://prtimes.jp/main/html/rd/p/000001344.000043465.html - source_type: PRTIMES - source_article_id: 465370 - confidence: high - basis: official_statement - period: Corporate Attitudes on Capital Investment Survey FY2026 (10th edition / April 2026) - caveat: Published value of a private Teikoku Databank survey

F-002: By size, "large firms" held at 70.7%, maintaining the prior-year level - source: PRTIMES #465370 - source_url: https://prtimes.jp/main/html/rd/p/000001344.000043465.html - source_type: PRTIMES - source_article_id: 465370 - confidence: high - basis: official_statement - period: Corporate Attitudes on Capital Investment Survey FY2026 (April 2026) - caveat: Published value of a private Teikoku Databank survey

F-003: "SMEs" were 54.3% (down 0.7pt), within which "small firms" were 42.0% (down 2.6pt) - source: PRTIMES #465370 - source_url: https://prtimes.jp/main/html/rd/p/000001344.000043465.html - source_type: PRTIMES - source_article_id: 465370 - confidence: high - basis: official_statement - period: Corporate Attitudes on Capital Investment Survey FY2026 (April 2026) - caveat: Private Teikoku Databank survey; gap vs large firms 70.7% is 28.7pt

F-004: Average planned capex amount was ¥130.43M, up ¥6.14M from the prior year (¥124.29M) - source: PRTIMES #465370 - source_url: https://prtimes.jp/main/html/rd/p/000001344.000043465.html - source_type: PRTIMES - source_article_id: 465370 - confidence: high - basis: official_statement - period: Corporate Attitudes on Capital Investment Survey FY2026 (April 2026) - caveat: Published value of a private Teikoku Databank survey

F-005: Among capex purposes, "equipment replacement" stood out at 59.0% - source: PRTIMES #465370 - source_url: https://prtimes.jp/main/html/rd/p/000001344.000043465.html - source_type: PRTIMES - source_article_id: 465370 - confidence: high - basis: official_statement - period: Corporate Attitudes on Capital Investment Survey FY2026 (April 2026) - caveat: Private Teikoku Databank survey; indicates a replacement-demand-centered defensive shift

F-006: Firms considering "digital investment" were 35.1% overall; large firms 51.3% vs SMEs 31.4%, a 19.9pt gap - source: PRTIMES #465370 - source_url: https://prtimes.jp/main/html/rd/p/000001344.000043465.html - source_type: PRTIMES - source_article_id: 465370 - confidence: high - basis: official_statement - period: Corporate Attitudes on Capital Investment Survey FY2026 (April 2026) - caveat: Published value of a private Teikoku Databank survey

F-007: The main funding method was "own funds," highest at 58.3% - source: PRTIMES #465370 - source_url: https://prtimes.jp/main/html/rd/p/000001344.000043465.html - source_type: PRTIMES - source_article_id: 465370 - confidence: high - basis: official_statement - period: Corporate Attitudes on Capital Investment Survey FY2026 (April 2026) - caveat: Published value of a private Teikoku Databank survey

F-008: The top reason for not investing was "uncertain outlook" at 50.2% - source: PRTIMES #465370 - source_url: https://prtimes.jp/main/html/rd/p/000001344.000043465.html - source_type: PRTIMES - source_article_id: 465370 - confidence: high - basis: official_statement - period: Corporate Attitudes on Capital Investment Survey FY2026 (April 2026) - caveat: Published value of a private Teikoku Databank survey

F-009: Survey scale — 23,083 firms surveyed nationwide, 10,538 valid respondent firms, survey period April 16–30, 2026 - source: PRTIMES #465370 - source_url: https://prtimes.jp/main/html/rd/p/000001344.000043465.html - source_type: PRTIMES - source_article_id: 465370 - confidence: high - basis: official_statement - period: Corporate Attitudes on Capital Investment Survey FY2026 (April 2026) - caveat: Sample base of the private Teikoku Databank survey

F-010: Fukuoka City eased the regular-employee requirement of its relocation grant from "3 or more" to "1 or more" - source: PRTIMES #577732 - source_url: https://prtimes.jp/main/html/rd/p/000000011.000182637.html - source_type: PRTIMES - source_article_id: 577732 - confidence: high - basis: official_statement - period: Fukuoka City revision of relocation grant for fabless semiconductor firms - caveat: Published value of a Fukuoka City municipal program

F-011: Raised the subsidy per employee from ¥500,000 to ¥1,000,000 (cap ¥50M for a baseline-scale case) - source: PRTIMES #577732 - source_url: https://prtimes.jp/main/html/rd/p/000000011.000182637.html - source_type: PRTIMES - source_article_id: 577732 - confidence: high - basis: official_statement - period: Fukuoka City revision of relocation grant for fabless semiconductor firms - caveat: Published value of a Fukuoka City municipal program

F-012: Newly established a subsidy of 1/2 of costs (cap ¥10M) for equipment-introduction costs - source: PRTIMES #577732 - source_url: https://prtimes.jp/main/html/rd/p/000000011.000182637.html - source_type: PRTIMES - source_article_id: 577732 - confidence: high - basis: official_statement - period: Fukuoka City revision of relocation grant for fabless semiconductor firms - caveat: Fukuoka City municipal program (EDA tools / clean rooms etc. are eligible)

F-013: The global semiconductor market is projected to expand from roughly ¥50 trillion in 2020 to roughly ¥150 trillion by 2030 - source: PRTIMES #577732 - source_url: https://prtimes.jp/main/html/rd/p/000000011.000182637.html - source_type: PRTIMES - source_article_id: 577732 - confidence: medium - basis: official_statement - period: Cited within the Fukuoka City grant-revision release - caveat: A third-party market projection cited by Fukuoka City (not primary data)

F-014: National restaurant openings in July–September 2025 numbered 14,314 (2023: 15,929 → 2024: 14,055 → 2025: 14,314, a bottoming-out trend) - source: PRTIMES #579035 - source_url: https://prtimes.jp/main/html/rd/p/000000088.000043858.html - source_type: PRTIMES - source_article_id: 579035 - confidence: high - basis: official_statement - period: National Restaurant Openings Survey ver7 (July–September 2025) - caveat: Published value of a private Review Inc. survey

F-015: Kumamoto Prefecture's restaurant openings were 257, a sharp rise from 177 the prior year (against the backdrop of JASM's arrival) - source: PRTIMES #579035 - source_url: https://prtimes.jp/main/html/rd/p/000000088.000043858.html - source_type: PRTIMES - source_article_id: 579035 - confidence: high - basis: official_statement - period: National Restaurant Openings Survey ver7 (July–September 2025) - caveat: Private Review Inc. survey; causation with JASM's arrival is the survey body's assessment

F-016: Regional ranking of restaurant openings — #1 Tokyo 2,079, #2 Osaka 1,256, #3 Aichi 882, #4 Fukuoka 837, #5 Hokkaido 776 - source: PRTIMES #579035 - source_url: https://prtimes.jp/main/html/rd/p/000000088.000043858.html - source_type: PRTIMES - source_article_id: 579035 - confidence: high - basis: official_statement - period: National Restaurant Openings Survey ver7 (July–September 2025) - caveat: Private Review Inc. survey; semiconductor-cluster Fukuoka near the top

---

J-Units

J-001: FY2026 Japanese corporate capex shows entrenched polarization in "whether to invest" by size — beyond large firms holding 70.7% vs small firms 42.0% (28.7pt gap), the planned amount on the investing side also rose (+¥6.14M), so the size gap is entrenching on both the "whether" and the "how much" of investment - confidence: medium - basis_f_units: F-002, F-003, F-004

J-002: The polarization extends to a defensive shift in quality — the combination of "equipment replacement" dominating at 59.0%, funding centered on own funds at 58.3%, and the top deferral reason being outlook uncertainty at 50.2% indicates a defensive posture of cautiously cycling aging-equipment renewal rather than forward-looking expansion, with macro uncertainty suppressing the "offense" of investment - confidence: medium - basis_f_units: F-005, F-007, F-008

J-003: The asymmetry of "overall retreat × local overheating" characterizes the current phase of Japanese capex — while the national survey advances a 3rd-straight-year retreat (56.7%), semiconductor clusters see local overheating on both the institutional side (Fukuoka's 1/2 equipment subsidy) and the real-demand side (Kumamoto's restaurants 177→257), deepening a regional/sectoral divergence that averages alone would miss - confidence: medium - basis_f_units: F-001, F-012, F-015

J-004: The same semiconductor capex shows different faces at origin and spillover destination — the Taiwan-sourced TSMC/JASM arrival is, at HQ, AI-demand-driven expansion investment (ANK-2026-04-16-001), yet at the Japanese destination it appears as localized real demand of foot traffic and openings, making visible a chain in which global semiconductor investment moves even where regional-economy restaurants open - confidence: low - basis_f_units: F-015, F-016

---

P-Units

P-001: Whether capex appetite of 56.7% bottoms out or proceeds to a 4th-straight-year retreat — in the current 3rd-straight-year retreat, a further drop in small firms' 42.0% could directly drive structuralization of the productivity gap. Track whether the next (2027) survey reverses - status: open

P-002: Whether local overheating in semiconductor clusters persists or proves transient — the localized investment and demand from JASM in Kumamoto and Fukuoka subsidies depend on the semiconductor cycle, and the regional-economy backlash on a cycle reversal is unknown. Track the linkage between the semiconductor cycle and regional opening trends - status: open

P-003: Whether institutional inducements (Fukuoka's equipment subsidy, etc.) actually move SME investment sentiment — as the national survey's "outlook uncertainty 50.2%" shows, subsidies alone may struggle to spark appetite. Track the correlation between municipal subsidy uptake and the SME investment rate - status: open

---

同事件・三視角 / Three Perspectives on the Same Event / 同一イベント・三つの視点

---

Internal citation chain

Published ANK-Docs cited by this article: - **ANK-2026-04-16-001** (TSMC 2026 Q1 earnings call: AI-driven 3nm simultaneous ramp across three regions) → This article cites the "AI-demand-driven capex expansion" it reveals as a contrast axis: this card depicts the defensive retreat of Japan's overall capex (3rd straight year, replacement-59.0%-centered) and the asymmetry of local overheating, but the origin of that overheating — JASM's arrival in Kumamoto — is a spillover destination of the Taiwan-sourced AI-demand-driven expansion investment the TSMC case depicts. Juxtaposing the two reveals a chain in which the same semiconductor capex shows different phases at origin (Taiwan = expansion) and destination (Japan's regions = localized real demand).

---

Sources

1. [PRTIMES #465370] Teikoku Databank, Ltd., "Corporate Attitudes on Capital Investment Survey (FY2026)", 2026-05. https://prtimes.jp/main/html/rd/p/000001344.000043465.html 2. [PRTIMES #577732] Fukuoka City, "Revision of the relocation grant for fabless semiconductor firms", 2026-05. https://prtimes.jp/main/html/rd/p/000000011.000182637.html 3. [PRTIMES #579035] Review Inc., "National Restaurant Openings Survey ver7", 2026-05. https://prtimes.jp/main/html/rd/p/000000088.000043858.html 4. [ANK-2026-04-16-001] Rin Takenouchi, "AI Demand Drives TSMC to 30%+ Full-Year Growth: Q1 EPS NT$22.08 a Record, 3nm Ramps Simultaneously Across Taiwan, US and Japan", 2026-04-16. https://ainews.washinmura.jp/ainews/en/ank/ANK-2026-04-16-001

---

📊 引用級事實單元（F-Units）

Firms with capex plans "in place" for FY2026 were 56.7%, down 0.7pt from the prior (April 2025) survey, falling for a 3rd consecutive year
F-001 · Confidence: high · Basis: official_statement PRTIMES #465370 Corporate Attitudes on Capital Investment Survey FY2026 (10th edition / April 2026)
By size, "large firms" held at 70.7%, maintaining the prior-year level
F-002 · Confidence: high · Basis: official_statement PRTIMES #465370 Corporate Attitudes on Capital Investment Survey FY2026 (April 2026)
"SMEs" were 54.3% (down 0.7pt), within which "small firms" were 42.0% (down 2.6pt)
F-003 · Confidence: high · Basis: official_statement PRTIMES #465370 Corporate Attitudes on Capital Investment Survey FY2026 (April 2026)
Average planned capex amount was ¥130.43M, up ¥6.14M from the prior year (¥124.29M)
F-004 · Confidence: high · Basis: official_statement PRTIMES #465370 Corporate Attitudes on Capital Investment Survey FY2026 (April 2026)
Among capex purposes, "equipment replacement" stood out at 59.0%
F-005 · Confidence: high · Basis: official_statement PRTIMES #465370 Corporate Attitudes on Capital Investment Survey FY2026 (April 2026)
Firms considering "digital investment" were 35.1% overall; large firms 51.3% vs SMEs 31.4%, a 19.9pt gap
F-006 · Confidence: high · Basis: official_statement PRTIMES #465370 Corporate Attitudes on Capital Investment Survey FY2026 (April 2026)
The main funding method was "own funds," highest at 58.3%
F-007 · Confidence: high · Basis: official_statement PRTIMES #465370 Corporate Attitudes on Capital Investment Survey FY2026 (April 2026)
The top reason for not investing was "uncertain outlook" at 50.2%
F-008 · Confidence: high · Basis: official_statement PRTIMES #465370 Corporate Attitudes on Capital Investment Survey FY2026 (April 2026)
Survey scale — 23,083 firms surveyed nationwide, 10,538 valid respondent firms, survey period April 16–30, 2026
F-009 · Confidence: high · Basis: official_statement PRTIMES #465370 Corporate Attitudes on Capital Investment Survey FY2026 (April 2026)
Fukuoka City eased the regular-employee requirement of its relocation grant from "3 or more" to "1 or more"
F-010 · Confidence: high · Basis: official_statement PRTIMES #577732 Fukuoka City revision of relocation grant for fabless semiconductor firms
Raised the subsidy per employee from ¥500,000 to ¥1,000,000 (cap ¥50M for a baseline-scale case)
F-011 · Confidence: high · Basis: official_statement PRTIMES #577732 Fukuoka City revision of relocation grant for fabless semiconductor firms
Newly established a subsidy of 1/2 of costs (cap ¥10M) for equipment-introduction costs
F-012 · Confidence: high · Basis: official_statement PRTIMES #577732 Fukuoka City revision of relocation grant for fabless semiconductor firms
The global semiconductor market is projected to expand from roughly ¥50 trillion in 2020 to roughly ¥150 trillion by 2030
F-013 · Confidence: medium · Basis: official_statement PRTIMES #577732 Cited within the Fukuoka City grant-revision release
National restaurant openings in July–September 2025 numbered 14,314 (2023: 15,929 → 2024: 14,055 → 2025: 14,314, a bottoming-out trend)
F-014 · Confidence: high · Basis: official_statement PRTIMES #579035 National Restaurant Openings Survey ver7 (July–September 2025)
Kumamoto Prefecture's restaurant openings were 257, a sharp rise from 177 the prior year (against the backdrop of JASM's arrival)
F-015 · Confidence: high · Basis: official_statement PRTIMES #579035 National Restaurant Openings Survey ver7 (July–September 2025)
Regional ranking of restaurant openings — #1 Tokyo 2,079, #2 Osaka 1,256, #3 Aichi 882, #4 Fukuoka 837, #5 Hokkaido 776
F-016 · Confidence: high · Basis: official_statement PRTIMES #579035 National Restaurant Openings Survey ver7 (July–September 2025)

❓ FAQ

What is the state of Japanese corporate capex appetite in FY2026?

**The share of firms with capex plans "in place" is 56.7% — down 0.7pt from the prior survey and a 3rd-straight-year retreat. Investment appetite overall is on a cooling trend.** According to Teikoku Databank's survey (23,083 firms surveyed nationwide, 10,538 valid responses, April 2026), the share with capex plans "in place" for FY2026 was 56.7%, down 0.7pt from the prior survey (April 2025) and falling for a 3rd consecutive year. Behind this lies macro uncertainty captured by "uncertain outlook" (50.2%) (PRTIMES #465370).

What exactly does the capex "polarization" mean?

**While large firms hold the prior-year level with 70.7% having plans "in place," small firms fall to 42.0% (down 2.6pt), and the gap between them widens to 28.7pt.** In the April 2026 survey, by size, large firms stood at 70.7% (holding the 2025 prior-year level) versus SMEs at 54.3% (down 0.7pt from 2025), and within that small firms at 42.0% (down 2.6pt from 2025). Moreover, the average planned capex amount rose to ¥130.43M, up ¥6.14M year-on-year from 2025, so polarization of "investing large firms invest more, non-investing SMEs invest less" advances on the dollar side too (PRTIMES #465370).

How is the "quality" of investment changing?

**"Equipment replacement" dominates at 59.0%, making a defensive shift toward renewing aging equipment (replacement demand) rather than new development or capacity expansion clear. Digital investment also splits large 51.3% vs SME 31.4% — a 19.9pt gap.** Among capex purposes, "equipment replacement" stands out at 59.0%, indicating the center of gravity is renewal demand rather than forward-looking expansion. Firms considering digital investment split large 51.3% vs SME 31.4% (19.9pt gap), and funding is most often "own funds" at 58.3%, suggesting a cautious posture avoiding reliance on external capital (PRTIMES #465370).

Why is only the semiconductor area overheating locally while the whole retreats?

**Because semiconductor clustering (JASM in Kumamoto, Fukuoka City subsidies) is sparking localized investment, demand and openings. Fukuoka City newly added a 1/2 equipment-cost subsidy (cap ¥10M), and Kumamoto's restaurant openings surged from 177 in 2024 to 257 in 2025.** Fukuoka City revised its relocation grant for semiconductor firms, easing the employment requirement from 3 to 1 and newly establishing a 1/2 subsidy on equipment-introduction costs (cap ¥10M). Behind this is a projection that the semiconductor market expands from roughly ¥50T in 2020 to roughly ¥150T by 2030. In Kumamoto, restaurant openings surged from 177 the prior year to 257 on the back of JASM's (a TSMC subsidiary) arrival, with localized investment moving even real demand and openings (PRTIMES #577732, #579035).

How does this capex polarization differ from AI-driven expansion investment like TSMC's?

**Opposite directions. The TSMC case is AI-demand-driven expansion investment (forward-looking, demand-pulled), whereas this card depicts a defensive shift amid overall retreat (replacement 59.0%) and an asymmetry of local overheating — the same semiconductor capex generates expansion at the origin (Taiwan) and localized real demand at the spillover destination (Japan's regions).** The "AI-demand-driven 3nm simultaneous ramp across three regions" depicted in the TSMC earnings card (ANK-2026-04-16-001) is a flagship case of forward-looking expansion investment, but Japan's overall capex sits in a 3rd-straight-year retreat centered on replacement demand. Yet a facet of that Taiwan-sourced capex (JASM's arrival in Kumamoto) appears as localized real demand in Japan's regional economy (restaurant openings rising from 177 in 2024 to 257 in 2025), establishing the asymmetry of "overall retreat × local overheating" (PRTIMES #465370, #579035, ANK-2026-04-16-001). ---

🧠 編輯判斷（J-Units）

FY2026 Japanese corporate capex shows entrenched polarization in "whether to invest" by size — beyond large firms holding 70.7% vs small firms 42.0% (28.7pt gap), the planned amount on the investing side also rose (+¥6.14M), so the size gap is entrenching on both the "whether" and the "how much" of investment
Confidence: medium · Based on: F-002, F-003, F-004
The polarization extends to a defensive shift in quality — the combination of "equipment replacement" dominating at 59.0%, funding centered on own funds at 58.3%, and the top deferral reason being outlook uncertainty at 50.2% indicates a defensive posture of cautiously cycling aging-equipment renewal rather than forward-looking expansion, with macro uncertainty suppressing the "offense" of investment
Confidence: medium · Based on: F-005, F-007, F-008
The asymmetry of "overall retreat × local overheating" characterizes the current phase of Japanese capex — while the national survey advances a 3rd-straight-year retreat (56.7%), semiconductor clusters see local overheating on both the institutional side (Fukuoka's 1/2 equipment subsidy) and the real-demand side (Kumamoto's restaurants 177→257), deepening a regional/sectoral divergence that averages alone would miss
Confidence: medium · Based on: F-001, F-012, F-015
The same semiconductor capex shows different faces at origin and spillover destination — the Taiwan-sourced TSMC/JASM arrival is, at HQ, AI-demand-driven expansion investment (ANK-2026-04-16-001), yet at the Japanese destination it appears as localized real demand of foot traffic and openings, making visible a chain in which global semiconductor investment moves even where regional-economy restaurants open
Confidence: low · Based on: F-015, F-016

🔮 待驗證假設（P-Units）

Whether capex appetite of 56.7% bottoms out or proceeds to a 4th-straight-year retreat — in the current 3rd-straight-year retreat, a further drop in small firms' 42.0% could directly drive structuralization of the productivity gap. Track whether the next (2027) survey reverses
Status: open
Whether local overheating in semiconductor clusters persists or proves transient — the localized investment and demand from JASM in Kumamoto and Fukuoka subsidies depend on the semiconductor cycle, and the regional-economy backlash on a cycle reversal is unknown. Track the linkage between the semiconductor cycle and regional opening trends
Status: open
Whether institutional inducements (Fukuoka's equipment subsidy, etc.) actually move SME investment sentiment — as the national survey's "outlook uncertainty 50.2%" shows, subsidies alone may struggle to spark appetite. Track the correlation between municipal subsidy uptake and the SME investment rate
Status: open

Related Analyses