Mitsubishi Heavy Industries
Long-cycle capital-equipment + long-tail aftermarket services. Sells large-ticket engineered systems (gas-turbine combined-cycle plants, nuclear reactors, defense platforms, aero-structures, industrial machinery) on multi-year contracts, then earns high-margin recurring service/parts revenue on an installed base. Order-backlog-driven; revenue recognized over project life.
Earnings, margins, COGS & capex
FY2025 (ended Mar 2026) set fresh records across order intake (JPY 7,653.6B, +20% on the restated continuing-ops base), business profit (JPY 432.2B, +21.8%, 8.7% margin), net income (JPY 332.1B, +35.3%), and free cash flow, with order backlog at a record JPY 13,237.6B (+JPY 3,001.3B YoY), led by gas-turbine combined-cycle (GTCC), nuclear, and defense. This built on a record FY2024 (revenue JPY 5,027.1B as-reported, business profit JPY 383.1B/7.6%, net income JPY 245.4B). Note the FY2025 figures exclude Mitsubishi Logisnext, now classified as discontinued operations, which restates FY2024 comparatives downward. FY2026 guidance (year ending Mar 2027) is JPY 6,800B order intake, JPY 5,400B revenue, JPY 540B business profit (10% margin), JPY 380B net income, JPY 25/share dividend - the former FY2026 medium-term-plan target now adopted as formal annual guidance.
Revenue trend
Margins
up from 7.6% (FY2024); FY2026 guide ~10%
up from 4.9% (FY2024); net income +35.3% YoY
FY2025 EBITDA higher YoY but exact figure not precisely verified here
reaching/above the 12% medium-term target; precise figure not verified here
COGS structure
Not separately disclosed as a clean line; cost base is dominated by materials (specialty steel/alloys, castings, forgings), long-cycle project labor, and supply-chain components. Margin expansion is being driven by price/mix on scarce gas-turbine and defense capacity plus operating leverage on a growing installed base, partly offset by input-cost and capacity-ramp expense.
Capex
Rising materially. MHI is investing to roughly double large-frame gas-turbine manufacturing capacity over ~2 years to meet AI-datacenter-driven GTCC demand, and expanding defense/aerospace production. Exact capex figures not verified here.
Latest earnings
Strong finish - FY2025 revenue (JPY 4,974.1B) beat the JPY 4,800B guide raised in Nov 2025, and business profit (JPY 432.2B) came in above the JPY 390B guide held through 1H, with net income and FCF at records.
FY2026 (ending Mar 2027, issued 12 May 2026): order intake JPY 6,800B, revenue JPY 5,400B, business profit JPY 540B (10% margin), net income JPY 380B, dividend JPY 25/share. This equals the former FY2026 medium-term-plan target (revenue JPY 5,400B / business profit JPY 540B / ROE 12%), now adopted as annual guidance.
- Order backlog
- JPY 13,237.6B (record, FY2025-end)
- FY2025 order intake
- JPY 7,653.6B (record, +20% restated)
- FY2025 business-profit margin
- 8.7% (JPY 432.2B)
- Large gas turbines booked
- 25 units FY2024 (Americas); 23 units in 1H FY2025 (N. America/Asia)
Growth drivers
- AI datacenter power demand pulling GTCC (gas-turbine combined-cycle) orders, especially North America - 25 large-frame units booked in FY2024 (majority Americas) and 23 units in 1H FY2025 alone (North America and Asia); capacity being doubled
- Japan defense budget ramp (~JPY 8.5T FY2025 domestic) with MHI as prime; leads the Global Combat Air Programme (GCAP) sixth-gen fighter with UK/Italy
- Nuclear tailwind — Japanese reactor restarts, SRZ-1200 next-gen PWR, and SMR investment aimed at carbon-free baseload for data centers
- Record JPY 13.2T backlog converting to revenue with improving margins
- Aftermarket/services on a large installed base of turbines, plants, and aero-structures
- Aero-structures recovery (Boeing/Airbus build-rate increases — MHI supplies 787 composite wing structure and other airframe work)
Bull & bear
MHI is a scarce-capacity winner of two simultaneous super-cycles - AI-datacenter power (gas turbines + nuclear) and allied re-armament - with a record JPY 13.2T backlog, margins that already stepped to 8.7% in FY2025 en route to a 10% FY2026 guide, and pricing power that a duopoly-plus supply structure protects for several years.
- Large gas-turbine capacity is sold out multi-year; MHI is one of three credible suppliers and is doubling capacity into demand it can price - 25 FY2024 + 23 1H-FY2025 unit bookings validate the datacenter-power thesis
- Defense is a durable second engine: GCAP fighter lead plus a rising Japanese budget give a decade-long, government-funded backlog with low cyclicality
- Margin story is real, not just volume - business profit margin moved 7.6% (FY2024) to 8.7% (FY2025), with net income +35% and a credible 10% / 12% ROE FY2026 guide now formalized
- Nuclear optionality (restarts, SRZ-1200, SMR) is a call option on carbon-free baseload for AI that the market underprices in a conglomerate
- Record FCF is de-levering the balance sheet (interest-bearing debt down >20% in FY2025, equity ratio 37.3%) and funding both growth capex and rising dividends - a re-rating catalyst if the conglomerate discount narrows
A late-cycle heavy-industrial priced for a super-cycle whose peak may already be visible: turbine capacity is being added industry-wide, margins remain structurally sub-Western-peer, execution risk on mega-programs is the company's own worst historical enemy, and U.S. investors can only access it through an illiquid ADR and yen translation.
- Buying at record orders, record backlog, and a ~38x trailing P/E is buying the top of a cycle - GE Vernova and Siemens Energy are also doubling capacity, and by ~2028-30 the turbine scarcity premium can compress
- MHI's margin is still ~8-9% vs. cleaner double-digit Western pure-plays; the conglomerate structure dilutes the best assets and has historically destroyed value on new platforms (SpaceJet write-off, nuclear/plant overruns)
- The capacity build-out is a large fixed-cost bet; if datacenter power tilts to renewables+storage or SMRs faster than gas, MHI over-builds into a fade
- FY2026 order-intake guidance (JPY 6,800B) sits below the FY2025 record actual (JPY 7,653.6B) - conservative as usual, but a reminder that record order intake may be cresting even as revenue/profit still grow
- For a USD-first mandate, direct ownership is impractical (thin OTC MHVYF, yen translation risk); the cleaner ways to express the turbine/defense theme are U.S.-listed
What it is worth
Sum-of-the-parts sanity check + backlog-coverage framing (no formal DCF here). Market cap ~JPY 12.73T (~$79B) on FY2025 continuing-ops revenue JPY 4,974.1B (~$32B), business profit JPY 432.2B and net income JPY 332.1B implies roughly ~2.6x sales, ~29x business profit, and ~38x trailing P/E (~33x on FY2026 guided net income of JPY 380B) - a growth-cycle multiple for a heavy industrial, reflecting the turbine/defense/nuclear super-cycle already substantially priced in.
Peak-cycle orders + industry-wide capacity additions compress turbine pricing by ~2028-30, an execution stumble on a mega-program (the historical failure mode) hits margins, and yen strength cuts reported earnings - de-rating a stock priced for continued records.
Backlog converts steadily, margins grind from 8.7% toward the guided 10%, but FX and conservative order-intake guidance temper the growth optics; the current ~high-30s trailing / ~33x forward P/E is roughly fair for the visible cycle.
If MHI delivers the FY2026 JPY 540B business-profit / 10% margin / 12% ROE guide and the turbine cycle extends through the decade, earnings compound off the FY2025 record base and the conglomerate discount narrows - supporting a higher multiple on materially higher profit.
The stock has re-rated hard on the power + defense + nuclear thesis; a record JPY 13.2T backlog underwrites revenue but the ~38x trailing multiple leaves little margin for execution slips or an earlier-than-expected turbine-supply catch-up. Direct U.S. ownership is impractical (thin OTC MHVYF, yen translation). Not financial advice.
SWOT
Strengths
- One of only three global scale suppliers of large-frame heavy-duty gas turbines (with GE Vernova and Siemens Energy) at a moment of structurally scarce capacity and multi-year backlogs
- Diversified conglomerate — energy, defense, aero-structures, industrial machinery - smooths single-end-market cyclicality
- National-champion position in Japanese defense — GCAP fighter lead, missiles, naval, tied to a rising government budget
- Record backlog (JPY 13.2T) gives multi-year revenue visibility
- Deep nuclear franchise (PWR, SRZ-1200, SMR investment) leveraged to restart + carbon-free-baseload demand
Weaknesses
- Structurally lower margins (business profit ~8-9%) than best-in-class Western pure-plays; conglomerate discount
- History of large-program losses — the SpaceJet/MRJ regional-jet write-off and past nuclear/plant cost overruns show execution risk on new platforms
- Long-cycle project accounting exposes it to input-cost inflation and fixed-price contract risk during a capacity ramp
- Heavy reliance on capex-intensive expansion to capture the turbine cycle; risk of over-building if demand normalizes
- Revenue and margins sensitive to JPY/USD — reported figures and thin OTC ADR make it awkward for U.S. investors to own directly
Opportunities
- AI-datacenter electricity demand extending the gas-turbine super-cycle and CCUS/hydrogen-ready turbine retrofits
- SMR and next-gen nuclear as data-center and grid decarbonization scale
- GCAP and broader allied-defense export potential as Japan relaxes export constraints
- Decarbonization services — CO2 capture (MHI has a leading amine-based CCUS position), ammonia/hydrogen co-firing
- Aftermarket monetization of a growing installed turbine and aero fleet
Threats
- GE Vernova and Siemens Energy also doubling turbine capacity — a supply catch-up could erode pricing power by late-decade
- A stronger yen compresses reported revenue/margins and export competitiveness
- Program/execution risk on GCAP, SRZ-1200, and capacity build-out (cost overruns are the historical failure mode)
- Datacenter-power demand proving faster to build in renewables + storage or SMRs than gas, softening the GTCC thesis
- Geopolitical/supply-chain shocks to long-cycle projects; commodity input inflation
Moats, dependencies & bottlenecks
Moats
medium long-term Only MHI, GE Vernova, Siemens Energy can build utility-scale H/J-class turbines; certification, IP, and metallurgy are high barriers, but rivals are adding capacity too.
GCAP lead + Japanese-prime status; government-protected, decade-plus programs with switching costs measured in national security.
Moderate-Strong Turbine, plant, and aero fleets generate sticky, high-margin parts/service revenue for decades.
PWR/SRZ-1200 and leading amine CCUS give hard-to-replicate engineering IP; commercial payoff is policy-dependent.
Tier-1 composite wing/airframe work (e.g. 787) is designed-in and multi-program-long once qualified.
Dependencies
Microsoft, Amazon, Google) The GTCC order surge is levered to AI-power capex; a pause resets the growth thesis.
Budget trajectory and GCAP funding drive the defense engine; policy-dependent but currently rising.
Aero-structures revenue tracks OEM production; Boeing (BA) disruptions flow through to MHI.
Yen strength compresses reported revenue/margins and export competitiveness vs. Western peers.
High-temperature alloys and large forgings are supply-constrained; input inflation hits fixed-price projects.
Program success depends on trilateral UK/Italy/Japan execution and cost-sharing.
Advantages
- One of three global large-turbine suppliers into a structurally under-supplied market with pricing power
- Two uncorrelated growth engines (power + defense) plus nuclear optionality under one roof
- Record JPY 13.2T backlog and multi-year revenue visibility
- Improving margins (8.7% FY2025) and record FCF funding growth capex, de-leveraging, and dividends
- Leading CCUS and nuclear engineering IP positioning it for decarbonization + baseload demand
Weaknesses
- Structurally lower margins and a conglomerate discount vs. Western pure-plays
- Track record of large-program losses (SpaceJet, plant/nuclear overruns) - execution is the recurring risk
- Capex-heavy expansion into a possibly late-stage turbine cycle
- FX-sensitive reported results; FY2026 order-intake guide set below the FY2025 record actual
- Poor direct accessibility for U.S. investors (thin OTC ADR, yen translation)
Bottlenecks
- Large-frame gas-turbine manufacturing capacity — the binding constraint on near-term revenue; the doubling plan takes ~2 years to land
- Skilled heavy-engineering labor and specialized casting/forging supply
- Execution bandwidth across simultaneous mega-programs (turbine ramp + GCAP + nuclear) - historically MHI's failure point
- High-temperature alloy and critical-component supply chains
- Nuclear regulatory approvals gating SRZ-1200/SMR commercialization
Top signals & trends
Top signals
Multi-year revenue visibility led by GTCC, nuclear, defense; +JPY 3.0T YoY.
Records beat guidance; margin stepped to 8.7%.
Direct read on AI-datacenter power demand.
Formalizes the margin trajectory as annual guidance.
Supply catch-up could erode pricing power by ~2028-30.
Conservative as usual, but a hint order intake may be cresting.
Trends
High positive · Primary driver of the GTCC super-cycle and a nuclear/SMR tailwind.
High positive · Funds GCAP and the defense backlog for a decade.
Medium-High positive · SRZ-1200 and SMR investment positioned for carbon-free baseload.
Medium positive · MHI has a leading CCUS position and hydrogen-ready turbines.
Medium negative · Erodes long-run turbine scarcity premium.
Medium negative · Compresses reported results and export competitiveness.
Ecosystem & competitor graph
Suppliers feed the company; customers pull from it. Line thickness shows the strength of each tie (supply-chain dependency, customer earnings contribution). Hover to isolate a tie.
Specialty steel and heavy plate for turbines, plants, and defense.
Specialty-alloy / casting / forging suppliers High-temperature superalloys and large forgings for turbine hot sections - a key bottleneck input.
Controls, sensors, avionics for defense and industrial systems.
Asia) Buyers of GTCC plants and nuclear; the demand pulled by datacenter power.
AI-datacenter power buildout is the ultimate driver of turbine and nuclear orders, largely via utility intermediaries.
Prime customer for GCAP, missiles, naval; rising budget.
MHI supplies composite wing/airframe structure (e.g. 787) as a Tier-1 partner.
Aero-structures customer.
Primary global rival in large gas turbines, grid, and power; the cleanest U.S.-listed way to play the same GTCC/datacenter-power theme.
Third large-turbine supplier plus grid; also expanding capacity.
Aero-engine leader; overlaps in aero and (via legacy) power; benchmark for margins.
Aero engines, defense, and SMR - direct overlap in engines and small modular reactors.
GCAP partner AND defense competitor globally; both ally and rival.
Global fighter/defense prime; GCAP competes with next-gen U.S. programs for allied share.
Pratt & Whitney engines + missiles/defense electronics overlap.
Japanese heavy-industry peer across power, rail, industrial; GE Hitachi nuclear JV.
Japanese peer in aero-engines, turbochargers, industrial systems.
Japanese peer in gas turbines, aero, hydrogen, defense/naval.
Korean gas-turbine and nuclear builder; emerging challenger. Context only, not a buy/own call.
U.S. nuclear components / SMR overlap on the nuclear thesis.
Overlap in distributed power/engines and industrial machinery.