Goldman Sachs reiterates Buy on Samsung Electro-Mechanics as AI demand tightens MLCC and ABF supply

Overview

Based on U.S. non-deal-roadshow meetings, Goldman Sachs says Samsung Electro-Mechanics is benefiting from AI-driven demand in MLCCs and ABF substrates. The institution reiterates Buy, citing improving core fundamentals, tight supply, pricing potential, long-term customer agreements and higher-margin AI product mix.

Core views

Samsung Electro-Mechanics maintained its 3Q26 revenue and operating-profit guidance despite the KRW appreciating by around 13% quarter-to-date against the US dollar, which Goldman Sachs interprets as evidence that underlying fundamentals have improved. The company expects MLCC revenue to increase by high teens percentage quarter-on-quarter in 3Q26 with roughly 20% operating margin, while the substrate business is expected to deliver 20% quarter-on-quarter revenue growth and approximately 20% operating margin. Goldman Sachs attributes this resilience to strength across both MLCC and substrate operations. Tight MLCC supply is supporting pricing and customers' willingness to secure supply. SEMCO has raised prices twice this year for distributor-sold MLCCs, which represent around 10% of MLCC revenue, and is discussing potential price adjustments with direct customers. Customer concern over possible shortages during the next two to three years is driving advance volume commitments. SEMCO expects about 60% of next year's MLCC revenue to be covered by long-term agreements, generally for one year but sometimes longer; these agreements guarantee volume while allowing pricing to adjust with market conditions using reference pricing. Goldman Sachs identifies AI data centers as the key growth and margin driver for MLCC. The company expects the AI data-center MLCC market to grow 70%-80% this year, while its own AI MLCC revenue is expected to grow around 100% year-on-year and accelerate next year. AI-related MLCC revenue mix could approach 20% this year, exceed 30% next year and reach approximately 40% in two years. This mix shift matters because AI data-center MLCC operating margin is around 30%, compared with around 20% for the overall MLCC business; margins on certain high-end products are closer to 40%. Goldman Sachs therefore expects the AI mix increase to provide a significant tailwind to overall MLCC profitability. Silicon capacitors are presented as a longer-term incremental opportunity rather than a replacement for MLCCs. SEMCO characterizes them as complementary products: their smaller and thinner form factor can position them closer to chips and improve noise filtering. The company cited strong customer interest and a roughly US$1 billion silicon-capacitor contract with one large technology customer announced in May. Meaningful revenue is expected to begin next year, with a significant portion booked in 2028. Goldman Sachs highlights SEMCO's high-capacitance development capability and the Samsung group's foundry capability as advantages, as well as its potential to offer a turnkey solution for silicon-capacitor-embedded package substrates. AI data centers are also strengthening the ABF-substrate business. Data-center server and network applications now account for around 60% of ABF revenue and could rise to 70% in two years. Supply tightness has enabled price adjustments since 1Q26, and the shift toward AI data-center products has supported higher ABF margins. SEMCO plans new ABF capacity in Vietnam and Korea from 2028, expects capacity to double by 2030 versus last year, and expects ABF revenue to more than triple by 2030 versus this year. The report considers ABF tightness potentially more severe than MLCC tightness because ABF is customized and AI data-center products have larger body sizes and higher layer counts, making high yields more difficult to achieve. Customers are discussing long-term agreements aligned with their chip roadmaps, in some cases extending to 2032, and are making prepayments to secure capacity. Such agreements can also guarantee a specified annual utilization level. Goldman Sachs values SEMCO using a sum-of-the-parts approach driven by 2027E-2028E EV/EBITDA and retains a 12-month target price of W2,250,000.

Analysis framework

Goldman Sachs bases its conclusions on management discussions during a U.S. non-deal roadshow, linking maintained quarterly guidance to operating momentum, supply tightness to pricing and long-term agreements, and AI product mix to revenue and margin expansion. The target price uses a sum-of-the-parts valuation driven by 2027E-2028E EV/EBITDA.

Methodology notes

• Valuation methodsSOTP (Sum-of-the-Parts) Valuation

  Sum-of-the-parts valuation using 2027E-2028E EV/EBITDA.

  The report values the company by applying an enterprise-value-to-EBITDA approach across its businesses and combines the resulting values into a 12-month target price.

• Industry AnalysisSupply-demand framework

  Supply tightness, customer supply commitments and pricing adjustment analysis for MLCC and ABF.

  The report connects constrained supply and difficult ABF production yields with price increases, long-term agreements, prepayments and better utilization visibility.

• Industry AnalysisVolume-price decomposition

  AI-driven product mix, pricing premiums and margin differences across MLCC and ABF.

  Goldman Sachs explains expected growth through higher AI-related sales mix, premium pricing and higher margins on AI data-center components.

Asset mapping & comparison

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

• Samsung Electro-Mechanics (009150.KS)

  Primary covered company expected to benefit from AI data-center demand for MLCCs and ABF substrates.

  Strengths

  AI MLCC growth, higher-margin product mix, MLCC and ABF supply tightness, customer LTAs, silicon-capacitor capability and planned ABF capacity expansion.

  Comparison

  The report describes SEMCO as the second-largest global MLCC supplier.

  Risks

  A stronger-than-expected increase in industry MLCC supply and weaker-than-expected demand for AI servers, autos or smartphones.

Key data

• 3Q26 MLCC revenue growthHigh teens% qoqCompany expectation, with approximately 20% operating margin.
• 3Q26 substrate revenue growth20% qoqCompany expectation, with approximately 20% operating margin.
• MLCC revenue under LTAs next yearAround 60%Typical duration is one year, although some agreements are longer.
• AI data-center MLCC market growth70%-80% this yearSEMCO expects its own AI MLCC revenue to grow around 100% year-on-year.
• AI MLCC revenue mixNearly 20% this year, more than 30% next year, approximately 40% in two yearsExpected progression within SEMCO's MLCC revenue.
• MLCC operating marginAround 20% overall; around 30% for AI data-center MLCC; close to 40% for certain high-end productsHigher AI mix is expected to support overall MLCC margin.
• ABF data-center revenue mixAround 60%, potentially 70% in two yearsIncludes server and network applications.
• ABF expansion outlookCapacity to double by 2030 versus last year; revenue to more than triple by 2030 versus this yearNew capacity in Vietnam and Korea is planned from 2028.
• Target price and upsideW2,250,000 target price; W1,400,000 current price; 60.7% upsidePrice as of 11 Sep 2026 close.

Impact & implications

The report argues that AI demand creates a reinforcing growth path for SEMCO: greater content and premium pricing lift AI-related MLCC revenue, a higher AI mix improves margins, and supply constraints strengthen pricing and contractual demand visibility. Similar AI demand and tighter supply conditions are expected to support ABF pricing, margins, capacity utilization and long-term growth.

Risks

• A stronger-than-expected increase in industry MLCC supply could weaken the favorable supply-demand and pricing backdrop.
• Demand for AI servers, autos or smartphones could be weaker than expected.