Most of the attention on Samsung C&T Corporation (KRX:028260) goes to what it owns rather than what it does. That's understandable when marked-up affiliate stakes can add ₩56tn to book value in a half-year. It also means almost nobody has noticed that the operating business has quietly become twice as profitable as it used to be.
FY2019: operating income of ₩866.8bn on revenue of ₩30.76tn, a 2.82% margin. FY2025: operating income of ₩3.29tn on revenue of ₩40.74tn, 8.08%. Operating profit up 280% on revenue up 32%.
And the more interesting part is what happened in the last three of those years.
Revenue peaked at ₩43.16tn in FY2022. It was ₩41.90tn in FY2023, ₩42.10tn in FY2024 and ₩40.74tn in FY2025, a decline of 5.6% from the peak.
Operating income over the same stretch: ₩2.53tn, ₩2.87tn, ₩2.98tn, ₩3.29tn. Up 30%.
For a company whose largest division is engineering and construction, that's an unusual sequence. Contractors normally chase revenue, because backlog is the currency of the sector and idle capacity is expensive. A builder shrinking its top line for three straight years while expanding its margin has been declining work.
The margin structure confirms it. Gross margin was 12.36% in FY2019. It reached 13.99% in FY2022 and 18.87% in FY2025, and ran 18.46% in Q2 2026. Six and a half points of gross margin over six years is not a cyclical fluctuation in a business this size. It's a different mix of jobs.
The first half of 2026 has continued it. Revenue of ₩22.46tn against ₩19.76tn a year earlier, up 13.7%, with operating income of ₩1.75tn against ₩1.48tn, up 18.6%. Q2 set a quarterly revenue record of ₩12.00tn at an 8.60% operating margin, the highest in the data available here.
Korean contractors typically run operating margins in the low-to-mid single digits. Domestic apartment building, which used to be the profit engine, has been squeezed by land costs, financing and a soft housing market. Getting to eight percent means doing something other than apartments.
Samsung C&T has two structural advantages that most Korean builders don't. The first is captive work: the group builds semiconductor fabs, and fabs are among the most technically demanding and best-paid construction projects in the world. Building a fab for an affiliate is not an arm's-length competitive tender, and the margin reflects the specialisation rather than the relationship.
The second is that the company is not only a builder. Trading and investment, fashion, and resort and leisure sit alongside E&C, and the reported operating margin blends all four. The consolidated figure is not an E&C margin, and anyone using 8.08% to benchmark against a pure contractor is comparing different things. The half-year report's segment note breaks it out; the summary financials do not.
What the blended figure does tell you is that the whole enterprise has become substantially more profitable while getting slightly smaller, which is the harder trick.
The disclosed contract filings give a sense of the new work. The largest recent one is the Dukhan Solar Power Plant in Qatar: engineering, procurement, installation and commissioning of a 2,000MW solar facility for QatarEnergy.
The value is USD 1,047,382,760.80, which the filing converts to ₩1,540,071,611,480 at ₩1,470.40 to the dollar, the rate on the December 11, 2025 signing date. That's 3.66% of FY2024 revenue. The company received a letter of award in August 2025 and signed the main contract in December. An advance payment was made, and the rest bills against construction progress.
A billion-dollar solar EPC for a national oil company is a long way from where Korean contractors made their money a decade ago. It's also the kind of work that fits a company optimising for margin: technically specified, financed by a sovereign counterparty, and not competed against every mid-sized Korean builder.
On August 19 Samsung C&T amended the Dukhan filing. The reason given is a change in the contract period, and the completion date moved from February 28, 2030 to June 30, 2030.
Four months on a five-year job, disclosed less than a year after work began on September 1, 2025. That's early for a schedule revision.
I don't want to overread one amendment. Large EPC schedules move constantly and a four-month extension on a 2GW build could be permitting, grid connection, or the client's own sequencing. The filing gives no reason beyond "contract period change," and the contract value was not revised, which is the more important fact: the money is unchanged.
What makes it worth noting is the pattern around it. Samsung C&T filed five separate amendments to single sales and supply contracts between June 30 and August 19 of this year. I've read one of them closely. Whether the other four are schedule changes, scope changes or price revisions, I can't say from what I've reviewed, and characterising them all from a single example would be exactly the kind of inference this analysis should avoid.
Korean contractors have destroyed shareholder value in this business before, and the mechanism is worth remembering. Through 2010 to 2012 the sector won large Middle Eastern plant contracts at aggressive prices, and in 2013 the whole industry took writedowns as cost overruns landed on fixed-price terms. Several household names booked losses that erased years of profit.
The setup rhymes. Overseas contracts, a Gulf client, a five-year schedule, and a first extension inside year one. The Dukhan filing doesn't state whether the contract is lump-sum fixed price or reimbursable, so I can't say where cost risk sits. That's the single most important thing an investor would want to know and it isn't disclosed.
The second risk is concentration. If a meaningful share of the E&C margin comes from building fabs for affiliates, then the division's earnings are a derivative of Samsung Electronics' capital spending plans. Those plans are cyclical, and a memory downcycle would take fab construction with it, hitting Samsung C&T twice: once in the marks on its shareholdings, once in its order book.
The bull case is simply that six years is long enough to stop calling this a fluke. Margins have expanded through a housing downturn, a rate shock and a construction cost spike. A company that improved through all three is probably doing something structural.
The next contract amendment on Dukhan. A second schedule extension, or any revision to the ₩1.54tn contract value, would change the read from routine to concerning. Contract value revisions are separately disclosable in Korea, so a change would be visible.
The other marker is the E&C segment margin in the FY2026 annual report next March. The consolidated 8% blends four businesses. If the construction division alone is running above the mid single digits and holding it while the order book shifts overseas, this is a genuinely better company than it was in 2019. If the blend is being carried by trading or by the resort division, the story is smaller than the headline margin suggests.
kstock reads DART every morning and writes up what moved — the contract, the buyback, the number that does not add up. The daily post and a Saturday roundup, by email.