Report Interpretation

Overview

The report examines HD Hyundai Heavy Industries’ announced engine and SMR investments, with the central conclusion that a much larger-than-expected expansion in land-based HiMSEN power-engine capacity creates a meaningful AI data-center opportunity for HHI and recurring maintenance upside for HD Hyundai Marine Solution.

Core views

HHI announced KRW1,072.2bn of investment on 10 September: KRW833.6bn, including land, for a new HiMSEN power-engine plant in Onsan, Ulsan, and KRW238.6bn for an SMR main-component facility in Ulsan. The power-engine plant is scheduled for construction from March 2027 to May 2028, with operations expected from June 2028. It will add 3.0GW of annual capacity for H54-class engines of about 20MW each. The SMR plant is intended to produce main components for two TerraPower Natrium reactor units annually, each around 345MWe, with completion targeted in 1H29. Nomura expects total HiMSEN capacity to rise from 3.0GW at end-2025 to 7.2GW by 2030E. Land-based capacity would reach 4.0GW, comprising 3.0GW at the new Onsan facility and 1.0GW at HD Hyundai Engine after its expansion, while marine capacity would be 3.2GW at the Ulsan main plant. Together with about 9GW of two-stroke main-engine capacity, the group’s Engine & Machinery capacity would reach roughly 16.2GW. The new site also allows HHI to internalize crankshafts and engine blocks, which management identifies as supply-chain bottlenecks. The expansion is positioned around rising demand for land-based power engines, particularly from AI data centers. HHI said its inquiry pipeline materially exceeds the previously cited 3-9GW range, though management cautioned that inquiries do not necessarily become orders. In 2026, HHI disclosed US orders from Aperion Energy for 660MW, worth USD0.42bn, and Corban Energy for 1GW, worth USD0.67bn. Management stated that the US represents around 80-90% of global AI data-center demand, while new inquiries are split roughly evenly between AI data centers and conventional land-based power. Discussions with major technology customers increasingly cover floating data centers and barge-mounted power plants, using the same 9MW and 20MW engines. Nomura argues that the market backdrop supports profitability because both marine and power engines are seller’s markets. Management indicated current engine operating margins are in the low 20% range and power-engine margins should be higher. Nomura assumes each 1GW of additional capacity produces KRW1tn of revenue and that power engines generate a 30% operating margin, around 5 percentage points above HiMSEN marine-engine margins. On this basis, it forecasts KRW3.8tn of additional revenue and KRW1.1tn of additional operating profit at full utilization. The new plant is expected to deliver its first engine in early 4Q28E, operate at about 30% utilization in 2028E, and approach full capacity around 2030E. The report also frames the expansion in competitive terms. Management estimates Wärtsilä holds around 60% of the power-engine market, while Wärtsilä and Everllence together are expanding to 4.7GW of capacity. With 4.0GW of land-based capacity, HHI would reach comparable scale and could gain a greater share of incremental demand. Nomura notes that Wärtsilä, Everllence and HHI are the only licensors with large-bore, approximately 20MW, four-stroke medium-speed engine lineups. HMS is presented as the principal downstream beneficiary because it has an exclusive contract to maintain HHI’s HiMSEN engines. HHI has previously delivered about 0.2GW of land-based engines annually; at full utilization, the new facility would enable HMS to service 3.8GW of additional engines each year from 2029F. Nomura estimates maintenance revenue from the new land-based capacity could reach up to KRW1.1tn in 2033F. The model assumes AI data-center engines require annual maintenance spending of about 8% because of frequent operation over a 24-year life. With a three-year maintenance cycle and roughly 70% of the budget allocated in the third year, the report expects material revenue contribution from 2032F. At full utilization, maintenance operating profit from the new engines in 2032F could reach Nomura’s current 2026F HMS operating-profit estimate of KRW426bn. For HMS, Nomura maintains Buy and raises its target price to KRW340,000 from KRW270,000, implying 48.1% upside from KRW229,500. The target is based on 37.2x 2027F EPS of KRW9,102, compared with the prior approach of applying 31.5x to 2028F EPS of KRW10,728 discounted for two years at an 11.4% cost of equity. The 20% premium reflects expected earnings upside from HHI’s four-stroke engine expansion. Nomura calls HMS its preferred Korea shipbuilding-sector stock, citing solid fundamentals, sustained growth, shareholder returns and the prospect that additional AI data-center engine contracts could re-rate the shares. For HHI, Nomura maintains Buy and raises the target price to KRW560,000 from KRW550,000, implying 23.5% upside from KRW453,500. The valuation uses a 4.44x target P/B multiple on 12-month-forward BVPS of KRW126,069. The target P/B includes a 10.9% premium for the US Navy opportunity and a 61.4% premium for the AI data-center generator opportunity. Nomura raises the AI data-center opportunity valuation from KRW10.9tn to KRW21.0tn after the 3.3GW expansion, which exceeded its prior 1.2GW capacity forecast. It expects a re-rating when actual AI data-center orders materialize.

Analysis framework

Nomura starts with HHI’s announced capital investments and converts the added engine capacity into an operating timeline, delivery capacity, revenue and margin assumptions. It then assesses demand inquiries, customer applications, industry capacity and competitive positioning before tracing the expansion’s recurring maintenance effect on HMS. The report values HMS with a target P/E on forecast EPS and HHI with a target P/B on 12-month-forward BVPS, adding premiums for identified strategic opportunities.

Methodology notes

• Industry AnalysisSupply-demand framework

  Power-engine capacity expansion and demand from AI data centers and other land-based applications

  The report compares planned HHI capacity with its inquiry pipeline, disclosed orders and competitor capacity to explain why the added output may be absorbed and priced at favorable margins.

• Industry AnalysisUpstream-Midstream-Downstream Transmission

  Engine manufacturing expansion flowing into HMS maintenance revenue

  HHI’s delivery of additional engines is treated as expanding HMS’s serviceable installed base because HMS holds the exclusive maintenance contract.

• Valuation methodsP/E and PEG Valuation

  HMS target P/E valuation

  Nomura derives HMS’s KRW340,000 target price by applying a 37.2x target P/E multiple to 2027F EPS of KRW9,102.

• Valuation methodsPB valuation

  HHI target P/B valuation

  Nomura derives HHI’s KRW560,000 target price by applying a 4.44x target P/B multiple to 12-month-forward BVPS of KRW126,069, with opportunity premiums.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• HD Hyundai Marine Solution (443060 KS)

  Primary maintenance beneficiary of HHI’s increased HiMSEN engine deliveries through its exclusive maintenance contract.

  Strengths

  Exclusive HiMSEN maintenance position, expected recurring AI data-center engine service revenue, and solid projected fundamentals.

  Weaknesses

  Meaningful revenue contribution from the new facility is expected only from 2032F because of the maintenance-cycle timing.

  Comparison

  Nomura identifies HMS as its preferred stock in the Korea shipbuilding sector.

  Risks

  Potential headwinds in the newbuilding market.

• HD Hyundai Heavy Industries (329180 KS)

  Direct owner and operator of the expanded land-based HiMSEN engine capacity.

  Strengths

  Planned 4.0GW land-based capacity, large-bore engine capability, growing AI data-center inquiry pipeline and potential comparable scale to major global competitors.

  Weaknesses

  The new facility is not expected to reach full capacity until around 2030E.

  Comparison

  Management expects HHI’s 4.0GW land-based capacity to be comparable with the combined 4.7GW expansion planned by Wärtsilä and Everllence.

  Risks

  AI data-center investment could slow if interest rates rise.

Key data

• HHI announced investmentKRW1,072.2bnKRW833.6bn for the HiMSEN power-engine plant and KRW238.6bn for the SMR main-component plant.
• HiMSEN capacity7.2GW by 2030EUp from 3.0GW at end-2025; 4.0GW would be land-based power-engine capacity.
• Incremental engine revenueKRW3.8tnNomura estimate at full utilization, beginning from 2029F.
• Incremental engine operating profitKRW1.1tnNomura estimate at full utilization, using about 30% power-engine operating margin.
• HMS maintenance revenueUp to KRW1.1tn in 2033FEstimated contribution from HHI’s new land-based power-engine capacity at full utilization.
• HMS target priceKRW340,000Raised from KRW270,000; 48.1% implied upside from KRW229,500.
• HHI target priceKRW560,000Raised from KRW550,000; 23.5% implied upside from KRW453,500.

Impact & implications

Nomura views the expansion as evidence that HHI is responding to concrete AI data-center-related demand and can become a large-scale supplier of land-based power engines. It sees the largest recurring downstream benefit at HMS through its exclusive engine-maintenance relationship, while HHI gains direct revenue, profit and valuation upside from engine deliveries.

Risks

• Potential headwinds in the newbuilding market could impede HMS’s target-price achievement.
• A slowdown in AI data-center investment if interest rates rise is a downside risk for HHI’s AI data-center opportunity.

What to watch

• Progress of construction, commissioning and first delivery at the new 3.0GW Onsan power-engine plant, expected from June 2028 and early 4Q28E respectively.
• Whether AI data-center inquiries convert into firm orders, as management cautioned that not all inquiries become orders.
• Utilization ramp toward full capacity around 2030E and the resulting engine revenue and margin delivery.
• HMS maintenance-revenue realization from 2032F as the AI data-center engine service cycle matures.
• Further phases of the TerraPower Natrium SMR component program, which management said depend on demand.