KIA Corporation (KRX:000270) became a part-owner of an American steel mill on Friday. Hyundai Steel broke ground in Ascension Parish, Louisiana. The $5.8bn electric-arc furnace plant will make 2.7 million tons of automotive steel a year from 2029. The joint venture is Hyundai-POSCO Louisiana Steel. Hyundai Steel owns 50%. POSCO owns 20%. Hyundai Motor owns 15% and Kia owns 15%.
Kia has not filed anything about this. I checked the DART record through Friday and there is no capital contribution disclosure, no board resolution and no related-party filing. Everything in this piece about the ownership and the money comes from the partners' press releases and from Korean press coverage of the ceremony. That is worth stating up front. A 15% stake in a $5.8bn project would ordinarily generate a filing. Its absence tells you the commitment hasn't yet been sized in a way that meets the disclosure threshold.
My view is that the stake is a sensible, cheap hedge on one input cost, and that it does nothing about the cost that actually hurt Kia last year.
The plant is an electric-arc furnace, which melts scrap rather than smelting ore. It will supply automotive-grade sheet to Hyundai Motor's Alabama plant and Kia's Georgia plant. Both are a few hundred miles away. The partners say commercial production starts in 2029. POSCO joined in August with a $582m investment for its 20%. It will handle North American sales and technical cooperation.
That POSCO figure is the only hard number on the funding. If $582m buys 20%, the equity in the venture is about $2.9bn. In that case, the other $2.9bn of the $5.8bn is debt or later contributions. On that basis Kia's 15% would be about $437m. Korean press has instead taken 15% of $5.8bn and reported Kia's share as about $870m. The truth is probably between the two and depends on how the debt is structured. Call it ₩600bn to ₩1.2tn at around ₩1,380 per dollar, which is an approximate rate.
Against Kia's balance sheet, either figure is small. Cash and equivalents were ₩16.08tn at June 30. The upper estimate is 7.5% of that. The lower is 3.7%. Kia paid ₩2.56tn in dividends in FY2025 alone. The mill is a rounding item on the cash flow statement.
Steel is the largest raw material in a car by weight. US tariffs on imported steel and on products made from it have been running at 50% since mid-2025. A car built in Georgia with Korean steel carries that cost on the steel content. A car built in Georgia with Louisiana steel doesn't.
From 2029, then, Kia's US-built vehicles get a domestic steel supply. The group owns that supply, priced without a tariff. It comes from a plant whose output is committed to the two automakers before it opens. That is real. It removes one line of tariff exposure permanently and it does so for a commitment Kia can afford without noticing.
It also makes Kia's US plant more American in the way US trade policy now measures it. Regional content rules under the North American trade agreement count where the steel came from. A mill in Louisiana raises the score.
The tariff that cost Kia money last year was not on steel. It was on cars.
Kia told investors in January that US tariffs cost it $2.3bn in 2025. An earlier piece here worked through the second-quarter arithmetic. The company said its cost of sales would have been 79.2% of revenue without tariffs. That compares with a reported 81.7%. It implies roughly ₩826bn of tariff cost in a single quarter. FY2025 operating income fell 28% to ₩9.08tn from ₩12.67tn in FY2024. Tariffs on imported vehicles were the largest reason.
Where that tariff stands today is not settled in the sources I've read. The July 2025 trade deal set the rate on Korean-built cars at 15%, down from 25%, effective from November 1. In January 2026 Washington said it would return to 25%. It said Korea's legislature hadn't passed the investment package that was the deal's other half. Korean press in early September describes negotiations as still in progress. A news item on this site last week reported Kia reassessing its Mexico strategy as the talks moved.
The steel mill has no bearing on any of that. Cars shipped from Korea will pay whatever rate applies to cars shipped from Korea. Kia's exposure there depends on what share of its US sales are built in Georgia versus imported. The company doesn't break that out in the filings on DART. The mill helps the Georgia share. It does nothing for the rest.
The company's market capitalisation was ₩49.4 trillion at Friday's close of ₩126,500. Book equity was ₩64.69tn at June 30. So the shares trade at 0.76 times book. An earlier piece here attributed that discount to governance rather than to the business. Trailing four-quarter net income is ₩7.05tn, so the multiple is about 7 times.
Second-quarter operating margin was 8.0% on record revenue of ₩33.04tn. The stock is 39% below its 52-week high of ₩206,000. The market has already priced a tariff regime that eats a quarter of operating income. A steel mill that starts in 2029 doesn't change that price. A resolution of the car tariff would.
The absence of a Kia filing could mean the stake is smaller than 15% of the headline. Or Kia's contribution could come later. Or it could flow through Hyundai Motor. Any of those would make the commitment even less material, which strengthens the argument here rather than weakening it.
The mill itself could slip. A 2029 start for a greenfield electric-arc plant is a tight schedule. US industrial construction has run late and over budget across the board. If the plant opens in 2030 or 2031, the tariff benefit arrives with it.
Steel prices could fall. A captive mill is insurance against tariffs and against price spikes. It is a liability if scrap and US sheet prices drop below what the plant can produce for. The automakers are committed to taking the output.
And I could be wrong about the car tariff staying unresolved. If a deal lands this autumn at 15%, Kia's second-half margin recovers sharply. The steel mill will then look like the smart end of a good year rather than the small end of a bad one.
The first thing is a Kia filing. A capital contribution to a US affiliate above the disclosure threshold would appear on DART as a board decision. Its size will settle the $437m-versus-$870m question and say whether the venture is equity-heavy or debt-heavy.
The second is the tariff rate on Korean-built cars. Kia's third-quarter results in late October will show cost of sales as a share of revenue. Below 80% means the 15% rate is being applied. Above 81% means it isn't, and the mill is still three years away from helping.
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.