Samsung Electro-Mechanics has moved from an AI-component narrative to record quarterly revenue and operating profit, but the share price still requires investors to separate durable mix improvement from a peak-cycle extrapolation. At the November 6, 2026 KRX close of KRW 870,000, our 2027 Bear/Base/Bull range is KRW 527,699 / KRW 940,194 / KRW 1,391,735 per common share. Component and Package produced most of the company’s Q3 growth, and management attributes part of that strength to AI-server multilayer ceramic capacitors (MLCCs) and FC-BGA package substrates. The release does not quantify AI revenue separately or prove that the very high margins and capacity tightness embedded in the most optimistic broker targets will persist.
Prepared November 8, 2026 KST. This is our first Company Research baseline for Samsung Electro-Mechanics Co., Ltd. (KRX: 009150). Financial figures use consolidated K-IFRS and Korean won. The October 28 Q3 release is preliminary and was prepared before the external audit was completed. We distinguish that release from the latest filed statutory report and do not treat management’s market outlook as guaranteed revenue.
Three findings that matter
- Component and Package explain nearly all reported growth. Q3 revenue rose 30% year over year to KRW 3.763 trillion, while operating profit rose 165% to KRW 691 billion and the operating margin expanded to 18.4%. The two divisions explained about 97% of the sales increase; management links part of their growth to AI, but does not disclose AI revenue separately.
- The strongest businesses are also the most capacity-sensitive. Component sales rose 36% and Package sales 58%, but the next step depends on high-end MLCC pricing, FC-BGA yields, customer qualification and disciplined capacity additions.
- Our Base is materially below the analyst mean. Our KRW 940,194 Base is 61.0% below a KRW 2,411,111 eighteen-analyst reference average, triggering AHNIM’s 20% divergence flag. The difference is primarily margin durability and terminal multiple—not a disagreement about Q3.
How Samsung Electro-Mechanics earns money
Samsung Electro-Mechanics has three operating pillars. Component Solution sells MLCCs and related passive components used to stabilize voltage and filter electrical noise in smartphones, cars, networks, power modules and servers. Package Solution sells semiconductor substrates, including high-end flip-chip ball grid array (FC-BGA) products that connect processors and accelerators to the circuit board. Optical Solution sells camera modules for smartphones, electric vehicles, advanced driver-assistance systems and emerging sensing applications.
The economics are changing because AI servers require more electrical stability and more complex chip packaging than conventional servers. Management says the capacitance required per AI-server GPU or ASIC is doubling annually—its “AI MLCC Law”—and that AI customers continue to request more FC-BGA supply. That is a company market characterization, not an independently measured physical law. The investment question is whether the related product mix and pricing can remain favorable after capacity expands.
Q3 2026: the inflection is in margin, not just revenue
| Consolidated metric | Q3 2026 | Q2 2026 | Q3 2025 | YoY |
|---|---|---|---|---|
| Revenue | KRW 3,763.3B | KRW 3,457.2B | KRW 2,889.0B | +30% |
| Gross profit | KRW 1,067.9B | KRW 844.5B | KRW 600.5B | +78% |
| Operating profit | KRW 691.0B | KRW 440.4B | KRW 260.3B | +165% |
| Operating margin | 18.4% | 12.7% | 9.0% | +9.4 ppt |
| Net income attributable to owners | KRW 468.7B | KRW 315.7B | KRW 219.8B | +113% |
The key change is operating leverage. Revenue rose 9% sequentially, but operating profit rose 57%. Gross margin increased to 28.4% from 24.4% in Q2, while selling, general and administrative expenses fell 7% sequentially. That combination indicates both a richer product mix and cost absorption. It also means a future reversal in pricing or utilization could affect profit faster than revenue.
Non-operating expense was KRW 81.2 billion in Q3, compared with KRW 7.1 billion in Q2. Pre-tax profit therefore increased 41% sequentially, less than operating profit. The gap is a reminder that headline operating performance does not flow one-for-one into shareholder earnings.
Component and Package account for most growth
| Business | Q3 revenue | QoQ | YoY | Evidence to watch |
|---|---|---|---|---|
| Component Solution | KRW 1,883.0B | +14% | +36% | High-capacitance MLCC mix, pricing, long-term supply conversion and utilization |
| Package Solution | KRW 939.0B | +22% | +58% | AI accelerator/server CPU qualification, FC-BGA yields and added capacity |
| Optical Solution | KRW 941.3B | −9% | +3% | Flagship-phone timing, EV camera platforms and customer concentration |
Component represented 50.0% of Q3 sales and Package 25.0%, versus 47.7% and 22.3% in Q2. The combined mix shift is the clearest disclosed bridge from AI demand to corporate margin. Management cited ultra-high-capacitance and high-voltage MLCC penetration, long-term supply agreements with major technology customers, and new high-end FC-BGA products for server CPUs and AI accelerators.
The sequence still matters. A customer request is not booked revenue; a long-term agreement may specify supply access without guaranteeing every future unit; and installed capacity is not profitable output until yields and utilization are demonstrated. We would treat reported division sales, gross margin, operating cash flow and receivable collection as stronger evidence than capacity or “design-in” announcements alone.
This cycle does have more contractual visibility than a purely speculative capacity build. Since May, the company has disclosed seven long-term supply agreements above Korea’s materiality threshold, with stated contract values totaling roughly KRW 4.3 trillion. They include about KRW 1.5 trillion of silicon-capacitor supply for 2027–2028, a KRW 1.072 trillion AI-server MLCC agreement and a KRW 360.5 billion memory-substrate agreement. The sum is not U.S.-style remaining performance obligations or complete backlog: ordinary orders and contracts below the disclosure threshold are outside it, and revenue still depends on delivery terms.
The company is also undertaking approximately KRW 6.78 trillion of announced FC-BGA expansion across Sejong and Vietnam. The Sejong program runs into 2028, with mass production planned after construction and qualification; our 2027 model does not assume full output from either new project. Management has emphasized customer funding support and medium-term volume commitments, but the timing, accounting and repayment conditions of that support require later filing verification. Capacity backed by customers lowers demand risk; it does not remove execution, yield or return-on-capital risk.
Competition is not only about nominal capacity. Japanese MLCC leaders and global package-substrate specialists compete on material formulation, defect density, thermal and electrical performance, customer qualification history and manufacturing yield. Samsung Electro-Mechanics’ advantage is the ability to supply both MLCCs and FC-BGA products into AI infrastructure, but the company does not disclose customer-level AI share or product-level margin. Customer qualification and high-volume yield—not the product label—determine whether that breadth becomes durable excess return.
Optical Solution remains useful diversification, but its 9% sequential decline shows product-cycle sensitivity. The company began mass production of a 200-megapixel folded-zoom module and increased automotive camera supply, while overseas mobile-customer demand fell. Automotive and humanoid sensing can become a second structural leg; today, the division is still partly tied to flagship-phone timing.
Cash generation improved, but reinvestment is heavy
| Q3 / September 30, 2026 metric | Value | AHNIM interpretation |
|---|---|---|
| Operating cash flow | KRW 1,145.2B | Strong single-quarter consolidated cash generation |
| Purchases of tangible/intangible assets | KRW 862.7B | Large reinvestment requirement |
| Simple free cash flow | KRW 282.5B | AHNIM calculation: CFO less asset purchases |
| Cash | KRW 4,113.7B | Up 24% from Q2 |
| Debt | KRW 3,426.3B | Up 14% from Q2 |
| Net cash | KRW 687.4B | AHNIM calculation |
| Inventory | KRW 2,411.2B | Down 8% QoQ, up 11% YoY |
| Accounts receivable | KRW 2,237.8B | Up 1% QoQ, 22% YoY |
Q3 cash flow supports the quality of the earnings beat. Operating cash flow exceeded operating profit, inventory fell sequentially, and simple free cash flow remained positive despite heavy equipment spending. The balance sheet is net-cash positive on the presentation’s debt definition.
There is still a financing trade-off. The cash-flow statement shows KRW 589.1 billion of net debt-financing inflow, while the balance-sheet debt amount rose KRW 431.9 billion from Q2 to Q3 and asset purchases absorbed KRW 862.7 billion. That is manageable with current cash generation, but the AI opportunity will not create per-share value automatically if every demand increase requires a comparable burst of capital spending. The test is incremental operating profit and cash earned per unit of capacity.
The latest statutory half-year filing also shows concentration and dilution points that the preliminary Q3 deck does not. Samsung Electronics and its subsidiaries represented about 26% of first-half revenue. At FY2025, Samsung Electro-Mechanics had 74.694 million issued common shares and 2.907 million preferred shares, including treasury stock; recent capital remained unchanged and 2026 H1 basic and diluted EPS were the same. Our review of company and DART disclosures through November 8 found no announced large convertible bond, warrant bond or equity issuance. The immediate financing issue is therefore capex, customer funding and debt—not a currently announced share issue—but Q3 statutory notes remain pending.
AHNIM valuation: a 2027 earnings framework
We use a 2027 price-to-earnings framework because Samsung Electro-Mechanics is profitable and the market is already valuing a visible AI earnings ramp. Formula: 2027 revenue × net-income-attributable-to-owners margin ÷ 77.601 million common-equivalent valuation shares × scenario P/E. The denominator is 74.694 million issued common shares plus 2.907 million preferred shares, reflecting the gross-issued economic participation convention used in current broker valuation work. It includes treasury shares and is an AHNIM valuation denominator—not the statutory diluted common-share count or a claim of outstanding employee options. The company has not provided the revenue or margin assumptions below. They are AHNIM estimates.
| AHNIM scenario | 2027 revenue | Owners’ net margin | Net income attributable to owners | P/E | Value/share |
|---|---|---|---|---|---|
| Bear | KRW 17.5T | 13% | KRW 2.275T | 18× | KRW 527,699 |
| Base | KRW 19.0T | 16% | KRW 3.040T | 24× | KRW 940,194 |
| Bull | KRW 20.0T | 18% | KRW 3.600T | 30× | KRW 1,391,735 |
Bear assumes AI products grow but product pricing normalizes as supply expands and Optics remains cyclical. Base assumes 2027 revenue of KRW 19.0 trillion—close to Samsung Securities’ KRW 19.943 trillion estimate—and a 16% owners’ net margin. That margin is 3.55 percentage points above Q3’s 12.45% and requires both further Component/Package operating leverage and partial normalization of Q3’s unusually large non-operating expense; it is not a direct comparison with Q3’s 18.4% operating margin. Bull assumes still stronger mix, tight high-end MLCC and FC-BGA supply, successful capacity ramps and a premium multiple. These margin steps are AHNIM assumptions that future filings must verify. Our multiple discipline produces a much lower target than aggressive segment sum-of-the-parts models.
The November 6 KRX close of KRW 870,000 multiplied by 74.694 million listed common shares gives a reproducible KRW 64.98 trillion common-equity market capitalization. Subtracting KRW 687.4 billion of presentation net cash gives KRW 64.30 trillion of common-equity market capitalization less net cash—not full enterprise value, because it excludes the preferred shares’ market value. Our Base implies 8.1% upside from the current common price; the current price is 64.9% above Bear and 37.5% below Bull.
Analyst targets imply much more than our Base
A Naver-derived tracker showed 18 analysts and a Buy consensus, with a KRW 2,411,111 average target as of November 5. Separately, seven post-Q3 targets that we could identify across October 29 reporting—Macquarie, JPMorgan, Nomura, Shinhan, IBK, Meritz and Samsung Securities—averaged KRW 2,314,286, with a KRW 2.30 million median and KRW 1.90 million to KRW 2.70 million range. We keep the two samples separate because their inclusion rules and dates differ.
Secondary reporting says Shinhan raised its target to KRW 2.70 million from KRW 2.20 million and IBK to KRW 1.90 million from KRW 1.75 million after Q3, while Meritz raised its target to KRW 2.30 million from KRW 2.20 million. The reports cited continuing MLCC pricing, higher AI-server substrate mix and 2027–2028 profit upgrades. Samsung Securities’ accessible model used high segment EBITDA multiples—18× for Component and 20× for Package—to reach KRW 2.0 million. Those are external estimates, not company guidance.
Our unrounded KRW 940,194 Base is 61.0% below the KRW 2,411,111 eighteen-analyst mean, triggering AHNIM’s 20% divergence flag. We would narrow the gap if reported 2027 revenue approaches KRW 20 trillion, net margin sustains well above 16%, new capacity earns attractive returns and high-end supply tightness persists without a multiple reset. We would not raise our target merely to match the consensus.
What would change the thesis?
- Stronger: Component and Package growth remains above corporate growth; gross and operating margins stay elevated after capacity additions; operating cash flow covers capex; receivables do not outgrow sales.
- Weaker: AI customers defer capacity, MLCC or FC-BGA pricing normalizes faster than volume grows, yields disappoint, or debt and capex rise without comparable cash returns.
- Next checks: Q4 division mix, high-end utilization and pricing, FC-BGA capacity timing, capex, inventory, receivables, operating cash flow and the conversion of supply agreements into recognized revenue.
Current conclusion: Samsung Electro-Mechanics has delivered the operating evidence that many AI-component stories still lack. Q3 revenue, margin and cash flow confirm a real mix shift. The valuation question is harder. Our Base allows substantial 2027 growth and margin improvement, yet remains far below the broker mean because we do not capitalize peak-cycle economics at the most aggressive segment multiples. Investors should watch cash returns from new capacity as closely as headline AI demand.
Related coverage
Our Samsung Electronics research covers the adjacent AI-memory and foundry economics. Our SK hynix baseline tracks HBM cash returns, while our NVIDIA supplier-capacity analysis explains why funding or capacity alone is not revenue. Browse Company Research for dated business and valuation baselines.
Sources and original documents
[1] Samsung Electro-Mechanics — Q3 2026 Earnings Result, October 28, 2026
[2] Samsung Electro-Mechanics — Q3 2026 Business Results press release, October 28, 2026
[3] Samsung Electro-Mechanics — earnings-release archive, reviewed November 8, 2026
[4] Samsung Electro-Mechanics — stock information, November 6, 2026 KRX close and listed shares
[5] Samsung Securities — Samsung Electro-Mechanics company update, October 29, 2026
[6] Seoul Economic Daily — secondary reporting on Shinhan and IBK post-Q3 targets, October 29, 2026
[7] ChosunBiz — secondary reporting on Meritz post-Q3 target, October 29, 2026
[8] Naver-derived analyst tracker — 18-analyst mean target snapshot, November 5, 2026
[10] Samsung Electro-Mechanics — FC-BGA investment announcement, 2026
[11] Samsung Electro-Mechanics — AI-server MLCC long-term supply agreement, September 1, 2026
[12] Samsung Electro-Mechanics — silicon-capacitor long-term supply agreement, May 20, 2026
We provide information and independent analysis, not individualized investment advice or guaranteed returns. Our editorial responsibility, AI-use, valuation and corrections policy applies.