At 3pm on July 3, Samsung SDI Co., Ltd. (KRX:006400) told investors and reporters it would invest about ₩16tn in Ulsan through 2040. The stated purpose is to strengthen next-generation battery competitiveness. The detail line names three technologies: all-solid-state batteries, and LFP and sodium batteries for energy storage.
The number will get the headlines. The list is the interesting part, because two of those three chemistries are things this particular company spent a decade positioning itself against.
Samsung SDI built its identity at the expensive end of the battery market. High-nickel prismatic cells, premium European and American automotive customers, and a refusal to chase volume at Chinese prices. That strategy produced better margins than its Korean peers in good years, and it produced FY2025: an operating loss of ₩1.72tn on revenue of ₩13.27tn, down from ₩21.44tn in FY2023.
Lithium iron phosphate is the other end. It has lower energy density, uses no nickel or cobalt, costs much less, and is the chemistry with which CATL and BYD took the global volume market. Sodium-ion goes further still, replacing lithium entirely with a far cheaper and more abundant element, at the cost of yet more energy density. Neither is a technology a premium manufacturer reaches for when times are good.
To be fair, LFP was already on Samsung SDI's roadmap for storage, and the company has talked about it publicly for a couple of years. What's changed is the order of magnitude. A roadmap slide is not ₩16tn and a mass-production base at the company's main domestic plant. Committing that scale of capital to chemistries whose defining characteristic is low cost is a strategic decision, not a product line extension.
Sodium is the genuinely new item. Announcing mass production of sodium-ion cells is a statement that Samsung SDI expects the stationary storage market to become a cost war it needs to be able to fight, rather than a premium market it can serve from the top.
The obvious objection is that Ulsan is one of the more expensive places on earth to make a commodity product. Korean labour, Korean power, Korean land, competing against Chinese plants built for the purpose.
The answer is not cost, it's origin. US and European rules on battery sourcing and content increasingly determine which cells qualify for subsidies, credits and public procurement, and Chinese-made cells frequently do not. LFP made in Korea is expensive LFP. It is also LFP that a US utility or a European storage developer can buy without a policy problem.
That reframes the plan. Samsung SDI is not trying to beat CATL on price. It is trying to be the non-Chinese supplier of the chemistries customers actually want, at a price premium the rules make tolerable. Whether that premium survives a change in policy is the whole risk, and the filing does not address it.
The all-solid-state item sits at the opposite pole and is easier to understand. It is the technology every serious cell maker is racing toward for high-end electric vehicles, Samsung SDI has been among the more vocal about its pilot line, and putting mass production capital behind it is consistent with everything the company has said.
What the combination says is that Samsung SDI is abandoning the middle. The most advanced chemistry for vehicles, the cheapest chemistries for grid storage, and nothing much in between.
The ₩16tn deserves the same arithmetic every long-dated Korean investment number deserves.
The schedule in the disclosure runs from January 1, 2026 to December 31, 2040. Fifteen years. That averages ₩1.07tn a year.
Now the history. Purchases of property, plant and equipment totalled ₩26.77tn across the eleven fiscal years from FY2015 through FY2025, an average of ₩2.43tn a year. FY2024 alone was ₩6.27tn. So the plan's annual average is 44% of what this company has historically spent, and about a sixth of what it spent in its peak year.
The defence is that ₩16tn covers one site and three specific technologies rather than the whole capital programme. True, and it still doesn't rescue the number, because Ulsan and next-generation chemistry are precisely where management says the future is. A company funding its future at less than half its historical run rate is either being conservative in the disclosure or is not funding its future very hard.
There is also the possibility that spending is heavily front-loaded into 2027 through 2031, which would make the early years far larger than the average implies. The filing doesn't say. It gives a start date, an end date and a total.
Item four of the disclosure asks for the board resolution date. The answer is a dash.
That is not irregular. This is a fair-disclosure filing of guidance, provided by the IR group to domestic and overseas investors and media, and it carries the standard caveat that scale and schedule may change with market conditions. But it means no directors have committed to ₩16tn.
The venue is worth noting too. The filing names the event: a national briefing on the Yeongnam region's advanced-industry development vision. One day earlier, on July 2, Samsung Electro-Mechanics announced about ₩8tn for a package substrate hub in Sejong, at the equivalent briefing for the Chungcheong region.
Two Samsung affiliates, two consecutive days, two regional government events, two fifteen-year plans ending in 2040, neither with a board resolution attached. That is a pattern, and it should calibrate how much weight either number carries. These are real intentions announced in a policy setting, not capital committed.
The bull case rests on what the numbers have already started doing. Samsung SDI posted operating income of ₩203.8bn in the second quarter of 2026, its first positive quarter in six, with gross margin at 24.9% against 8.8% a year earlier. Revenue grew 18.5% year on year. Something in the mix is working, and the most plausible candidate is storage.
If US grid and data centre demand for non-Chinese cells is as durable as it currently looks, then a Korean LFP and sodium base with a policy moat around it is a genuinely good asset, and ₩16tn is a floor rather than a ceiling. Samsung SDI would be early among Korean makers in committing at that scale.
The bear case is that the company is buying into a commodity business at the top of a demand cycle, using capital it does not have in abundance, having just raised roughly ₩2tn of equity and cut capital spending by half. Cash was ₩1.51tn at June 30 against total liabilities of ₩20.79tn.
A new facility investment filing with a board resolution date and a specific won amount attached to Ulsan. Korean rules require one once directors approve a plant, and until it appears the ₩16tn is guidance with a 2040 end date and a disclaimer.
The nearer marker is the FY2026 capital expenditure figure, reported with full-year results in January. First-half spending was ₩1.13tn, down about 40% year on year. If the full year comes in near ₩2.3tn while a ₩16tn programme is supposedly under way, the plan's early years are not the heavy ones, and the fifteen-year average is closer to the truth than the headline.
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.