Korean equity market and multi-sector growth outlook Report Interpretation
The report argues that AI infrastructure, export-market expansion and rising shareholder returns can sustain Korean earnings growth well beyond the 2025-26 upturn. It favors technology, defense, power equipment, shipbuilding, autos, beauty, bio and financials, while recognizing sector-specific competitive and execution risks.
Summary
The report argues that AI infrastructure, export-market expansion and rising shareholder returns can sustain Korean earnings growth well beyond the 2025-26 upturn. It favors technology, defense, power equipment, shipbuilding, autos, beauty, bio and financials, while recognizing sector-specific competitive and execution risks.
- Korea’s annual exports are expected to reach US$1.0tn in 2026, led by AI-related products and globally competitive non-tech exports.
- BofA expects global AI capex to exceed US$3tn in 2030 and memory TAM to approach US$2.0tn.
- Preferred sectors include memory, hardware technology, defense, shipbuilding, autos, electric power equipment, beauty, bio and financials.
- The report sees value-up policies, dividends and buybacks supporting shareholder returns alongside earnings growth.
Report Interpretation
Overview
This is a Korea equity-market outlook that links a long-duration AI investment cycle with the globalization of Korean manufacturing, consumer brands and services. BofA’s central conclusion is that these forces can extend Korea’s growth and earnings cycle through 2030, although opportunities and risks differ materially by sector and company.
Core views
BofA’s central thesis is that Korea is entering a super-cycle that can persist through 2030 because AI is reshaping demand for memory and technology infrastructure while globalization is broadening export opportunities for Korean products. The report expects annual exports to reach US$1.0tn in 2026, after US$0.7tn through August, and identifies advanced-packaging memory, air defense, data-center power and cooling, beauty and biotechnology as important contributors. It expects GDP growth of 3.6% in 2026 and 2.5% in 2027, helped by exports; semiconductors already represent 40% of 2026 exports. BofA also expects earnings momentum to remain durable into 2027-28, supported by low inventories, relatively low leverage or net-cash positions in parts of the market, and rising dividends and buybacks. Memory is the report’s core AI transmission channel. It argues that HBM, high-speed DRAM, NAND, eSSD and advanced packaging are becoming essential to AI training, inference, cloud systems, robotics, autonomous vehicles and smart factories rather than being tied principally to PCs and smartphones. BofA estimates global AI capex could exceed US$3tn in 2030, versus US$0.7tn and US$1.0tn in 2026 and 2027, respectively, and expects memory TAM to rise from US$0.2tn in 2025 and US$0.9tn in 2026 to about US$2.0tn in 2030. Assuming operating margins above 50% for many DRAM and NAND producers, it calculates potential memory-industry operating profit of US$1.0tn by 2030. The report also forecasts HBM revenue CAGR above 50%, from US$35bn in 2025 to the high-US$200bn range in 2030, benefiting Korean memory leaders and their equipment and materials supply chains. The AI buildout broadens beyond memory. BofA expects Korean hardware leaders including Samsung Electro-Mechanics, LG Innotek and LG Electronics to deliver more than 20% operating-profit CAGR in 2026-30 through MLCCs, FC-BGA substrates, copper-clad laminates and cooling equipment for data centers, robots and autos. It cites potential multibillion-dollar US Big Tech prepayments for FC-BGA supply in 2028-30, capacity expansion across Korea and Southeast Asia, and NVIDIA qualification supporting LG Electronics’ cooling capacity. In power equipment, the report sees the shift from AC to DC data-center architecture creating demand for power sidecars and solid-state transformers, with LS Electric targeting 800VDC products and expected revenue contribution from sidecars in 2028 and transformers in 2029. It also expects demand for 765kV transformers to improve mix and margins for suppliers serving expanding data-center power loads. The report identifies several additional export and infrastructure beneficiaries. Korean defense companies could benefit as procurement shifts toward cost-effective, multi-layered air defense and AI-enabled unmanned systems. Shipbuilding earnings are supported by higher-priced backlogs, but BofA views data-center engines as an additional, less cyclical earnings stream: HD Hyundai Heavy Industries secured about W1.6tn of US data-center engine orders totaling 1,684MW within four months and announced W834bn of investment for a new 3GW HiMSEN engine plant. In power generation, BofA expects gas-turbine shortages, hydrogen-turbine commercialization and renewed nuclear and SMR investment to support Doosan Enerbility; it forecasts W7tn of annual new nuclear orders and W6tn of annual gas/combined-cycle EPC orders in 2026-28. In autos and batteries, BofA expects extended-range EVs to provide a transition option from 2027 where charging infrastructure remains inadequate, reducing range anxiety and battery capacity requirements. It also sees software-defined vehicles shifting differentiation toward software, data ecosystems and recurring-service revenue. For batteries, the report expects the US energy-storage market to grow about 20% annually, driven by renewable penetration and AI data centers; it sees Korean suppliers gaining through localized LFP output and non-Chinese supply-chain positioning. LG Energy Solution is presented as having greater grid-ESS and localized-LFP leverage, while Samsung SDI offers BBU/UPS exposure and earlier targeted solid-state commercialization. Consumer, healthcare, digital platforms and financials provide further globalization and structural-growth channels. Korean beauty exports grew 30% year to date in 2026, including 48% growth to North America and 69% to Europe; BofA highlights Latin America, where exports to Mexico and Brazil rose 80% and 83%, respectively. In healthcare, it expects peptide therapeutics to expand from US$94bn in 2025 to around US$200bn by 2031, with manufacturing capacity a critical bottleneck and Samsung Biologics’ PolyPeptide acquisition extending its peptide CDMO capabilities. NAVER’s planned cloud-AI expansion targets 100MW by end-2027, another 100MW by end-2028 and 1GW by 2031, while Korean telecom operators are expanding AI-data-center capacity. In financials, BofA expects capital-markets participation, ETFs, wealth management, digital products and shareholder returns to support brokers, banks and insurers. BofA frames valuation and capital returns as reinforcing the fundamental growth case. It notes low P/E multiples for selected semiconductor and auto companies, low debt or net cash in parts of coverage, and increasing payouts through dividends, buybacks and treasury-share cancellations. The report nevertheless distinguishes company opportunities rather than applying a uniform recommendation: competition from China, margin pressure, AI-capex cuts, policy shifts, order delays and execution risks can alter outcomes. At the macro level, it also flags Chinese competition in traditional exports, tariff and localization uncertainty for autos, and demographic headwinds that could reduce Korea’s potential growth to the 1% range from 2030 without AI-led productivity gains.
Analysis framework
BofA combines bottom-up analysis of roughly 100 Korean companies with sector demand, export, earnings, valuation and capital-return comparisons. It starts with the macro effects of AI and exports, then traces demand through memory, hardware, power, industrial, consumer and services value chains, using company capacity plans, orders, market-growth estimates and valuation frameworks to identify preferred exposures.
Methodology notes
AI infrastructure and export-demand analysis
The report links AI capex, data-center buildouts and global demand for Korean products to supply-chain capacity, pricing, orders, margins and earnings across sectors.
AI infrastructure supply-chain transmission
It traces demand from hyperscalers and AI systems through memory, packaging, substrates, cooling, power equipment, engines and related Korean suppliers.
Earnings-multiple valuation
The report uses P/E and PEG comparisons for several covered companies and compares target multiples with historical ranges and peers.
Sum-of-the-parts valuation
For selected diversified companies, the report values separate businesses or investments independently to capture different earnings drivers and holding-company discounts.
Asset mapping & comparison
Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).
- Samsung Electronics and SK HynixCore beneficiaries of the AI-memory super-cycle.
- Strengths
- Leadership in DRAM, HBM and NAND, supported by expanding AI-memory demand.
- Comparison
- The report notes low P/E multiples for selected memory companies despite strong earnings.
- Risks
- US hyperscaler capex cuts, excess memory supply, Chinese competition and market-share losses.
- HD Hyundai Heavy IndustriesPreferred exposure to data-center engines alongside shipbuilding recovery.
- Strengths
- HiMSEN platform, US data-center orders and planned capacity expansion.
- Weaknesses
- Shipping remains cyclical.
- Comparison
- Data-center engines are presented as a differentiated growth driver beyond marine applications.
- Risks
- Higher rates, reduced shipping capex and weaker seaborne trade.
- Doosan EnerbilityPreferred power-sector exposure to gas turbines, nuclear and SMRs.
- Strengths
- Gas-turbine intellectual property, hydrogen-turbine plans and SMR supply-chain partnerships.
- Weaknesses
- Earnings timing depends on project execution.
- Comparison
- One of only five global gas-turbine OEMs, according to the report.
- Risks
- Order delays, anti-nuclear policy reversal and AI overinvestment.
- LS ElectricPreferred power-equipment exposure to DC data centers and transformer demand.
- Strengths
- DC product track record, data-center orders and capacity expansion.
- Comparison
- BofA forecasts 39% EBITDA CAGR for 2025-28E versus a 24% peer average.
- Risks
- Infrastructure-spending slowdown, delayed data-center orders and low-cost competition.
- LG Energy Solution and Samsung SDIPreferred battery exposures to US ESS growth and next-generation technologies.
- Strengths
- US ESS capacity expansion; LGES has localized LFP and grid-scale exposure, while SDI has BBU/UPS and solid-state optionality.
- Weaknesses
- Exposure to EV-demand conditions remains material.
- Comparison
- LGES is more leveraged to grid ESS; SDI is positioned for earlier solid-state commercialization.
- Risks
- Slower ESS ramp, US EV weakness, competition and policy changes.
Key data
- Korea annual exportsUS$1.0tn in 2026EExpected annual level; US$0.7tn was reached through August.
- Korea GDP growth3.6% / 2.5%BofA forecasts for 2026 / 2027.
- Semiconductor share of exports40%Share of total Korean exports in 2026 as of August.
- Global AI capexUS$3tn+ in 2030Compared with US$0.7tn / US$1.0tn in 2026 / 2027.
- Memory TAM~US$2.0tn in 2030Versus US$0.2tn in 2025 and US$0.9tn in 2026.
- HBM market growth50%+ CAGRFrom US$35bn in 2025 to the high-US$200bn range in 2030.
- US ESS market growth~20% CAGRUPS growth is 70% CAGR and grid growth is 11% CAGR in the report’s comparison.
- Korean beauty export growth+30% YoY in 2026 YTDNorth America +48%, Europe +69%, and other regions +43%.
Impact & implications
The report argues that Korea’s AI opportunity extends beyond memory into data-center hardware, power systems, industrial equipment and services, while global demand diversifies export growth across defense, ships, autos, beauty, bio and entertainment. Rising capital returns and value-up reforms are presented as a further support to shareholder value, but BofA stresses that stock outcomes depend on company-level execution, valuation and sector risks.
Risks
- AI infrastructure spending or US hyperscaler capex could weaken, reducing demand for memory, substrates, cooling and power equipment.
- Chinese competition could pressure Korean manufacturers, particularly traditional export industries, while tariffs and overseas localization create uncertainty for autos.
- Order delays, policy reversals and execution risks could affect nuclear, defense, power-equipment and industrial projects.
- Battery demand could be affected by slower EV growth, ESS competition and changes in US or European policy.
- Korea’s demographic decline could lower potential growth to the 1% range from 2030 without sufficient AI-driven productivity gains.
What to watch
- AI-capex commitments by US hyperscalers, memory pricing and HBM demand through 2030.
- Korean export growth, including semiconductor, defense, shipbuilding, power-equipment and beauty shipments.
- Long-term agreements, prepayments and capacity additions for FC-BGA, cooling, memory and data-center power infrastructure.
- US ESS orders, LFP localization, solid-state battery commercialization and EV-policy developments.
- Execution of data-center engine, gas-turbine, nuclear, SMR and telecom data-center capacity projects.
- Dividend, buyback and treasury-share-cancellation progress under Korea’s value-up program.