Hyundai AutoEver Corporation (KRX:307950) grew June-quarter revenue 20% year on year, to ₩1,250.7bn. Operating profit rose 11%, to ₩90.5bn. Across the first half, revenue was up 16.6% and operating profit up 3.4%, with the operating margin falling from 5.8% to 5.1%.
The share price has been doing something considerably more dramatic. The 52-week range runs from ₩149,300 to ₩938,000. The stock closed at ₩385,500 on September 2, down 5% on the day — 59% below its high, and still 158% above its low.
Somewhere in the last year the market decided this was a software-defined-vehicle company and then decided it wasn't. The income statement participated in neither decision. That's the thing worth writing about: not whether the SDV story is true, but why ten years of revenue growth have left the profitability of this business exactly where it started.
The numbers are almost unnervingly stable. Revenue: ₩1,336bn in 2016, ₩1,572bn in 2019, ₩2,754bn in 2022, ₩4,252bn in 2025. Better than a tripling in nine years, compounding around 14%.
Operating margin over the same years: 6.0%, 5.1%, 5.2%, 6.0%. The low was 4.6% in 2021. The high was 6.0%, hit in 2016 and again in 2024 and 2025.
Nine years. Three times the revenue. Same margin.
For a systems integrator that would be unremarkable — it's what SI economics look like, headcount in and billings out. For a company whose bull case is proprietary vehicle software, it's the whole argument. Software margins expand with scale because the code is written once. AutoEver's don't, which means the code is not where the money is being made, or not where the money is being kept.
Here's why, and it needs a Korean detail that most US comparisons miss.
AutoEver's customers are Hyundai Motor, Kia, Genesis and the rest of the group's affiliates. Its trade receivables were ₩1,066.8bn in June against ₩1,250.7bn of quarterly revenue — roughly 78 days, and almost all of it owed by companies with the same ultimate owner. What AutoEver charges for its work is a group transfer price, not a market price.
Korea regulates this directly, in a way that has no US equivalent. The Fair Trade Act polices intra-group dealing at the large conglomerates: transactions between affiliates must be disclosed, and the Korea Fair Trade Commission has authority to act where a group is judged to be steering excessive or favourably-priced business to an affiliate. The rules exist because captive subsidiaries were historically used to move value toward controlling families, and enforcement has produced real penalties.
The practical effect on a listed captive is a ceiling. A subsidiary that suddenly earned 20% operating margins on affiliate revenue would be advertising exactly the pattern the regulator looks for. So the group has both a commercial reason and a regulatory reason to keep the intercompany price near cost-plus-a-modest-return — which is a fair description of 6%.
That is not a scandal. It is the structure. But it means an investor buying AutoEver for margin expansion is betting against the incentives of the entity that sets the margin.
None of this is a claim that the engineering is thin. Mobilgene, AutoEver's AUTOSAR-compliant vehicle software platform, is deployed across mass-production Hyundai, Kia and Genesis vehicles. Both the Classic and Adaptive variants have obtained ASIL-D certification, the top automotive functional-safety rating. That is a genuinely difficult thing to achieve and very few companies anywhere have done it.
The problem is where the value lands. When Mobilgene makes a Hyundai vehicle better, Hyundai Motor sells the vehicle. AutoEver bills the development. In a group structure, improving your customer's product is not the same as improving your own P&L, and the two decades of margin data say the transmission mechanism doesn't exist.
The event that would change this is a customer outside the group. A non-affiliate automaker licensing Mobilgene would establish a market price for the software and, more importantly, would produce revenue whose margin the group has no reason to hold down. I can't find evidence of one, and the company's disclosure doesn't break out non-affiliate revenue in the standardized filings.
Where AutoEver does look genuinely good is conversion. Operating cash flow was ₩423.6bn in 2025 against ₩255.3bn of operating profit — 166% conversion. In 2024 it was ₩256.0bn against ₩224.4bn. Billing your parent has one clear advantage: the receivable is good.
The balance sheet is clean. Equity of ₩1,922.5bn, cash of ₩239.8bn, and liabilities that are mostly trade payables — current liabilities of ₩1,225.3bn against non-current of just ₩374.4bn. There is no leverage story here at all.
Capital intensity is rising, though. Property and equipment purchases went from ₩85.1bn in 2024 to ₩169.5bn in 2025, and the property and equipment balance from ₩184.1bn to ₩299.0bn in eighteen months. That is a company building data centre and development capacity, which is the right thing to do and also the thing that keeps the margin at 6%.
The dividend is the disappointment. ₩49.8bn paid in 2025 against ₩186.8bn of net income — a payout around 27% — works out under 0.5% on the current price. For a debt-free cash generator sitting inside Korea's value-up debate, that is a conspicuously small distribution.
Market capitalization is ₩10.57tn, about $7.7bn at roughly 1,370 won per dollar, an approximate rate. Trailing net income of ₩195.3bn puts the shares at 54 times. Trailing revenue of ₩4,563bn puts them at 2.3 times sales. Book value of ₩1,922.5bn puts them at 5.5 times.
Those are software multiples on systems-integration economics. A US IT services company growing 15% with a 6% operating margin and no third-party product revenue would not trade at 54 times earnings, and the gap is the SDV narrative.
The counter-argument deserves a hearing. Vehicle software content per car is rising fast across the industry, AutoEver's revenue is mechanically levered to it, and 15% compound growth for a decade is not a small thing. If Hyundai Motor Group's software spending doubles again, AutoEver's revenue roughly doubles too, at 6%. Growth at a fixed margin still compounds earnings.
It just doesn't justify a re-rating. It justifies the earnings growing at the revenue growth rate, which is what has happened.
Single-customer concentration, in the most literal form available — the customer is the parent group. Hyundai and Kia unit volumes, capex cycles and IT budgets set this company's revenue, and none of those decisions are made with AutoEver shareholders in mind.
Seasonality distorts every quarterly read. The March quarter earned a 2.3% operating margin this year and 3.2% last year, against 7.2% in June. Anyone annualizing a first quarter here will be wrong.
Regulatory attention to intra-group transactions can move in either direction. Tightening would pressure the revenue base; the same rules also currently justify the low margin, so an easing wouldn't automatically show up as profit either.
And the share price itself. A stock that ranged from ₩149,300 to ₩938,000 inside twelve months has a shareholder register with wildly different cost bases and a price that has demonstrated it does not need a fundamental reason to move by half.
One thing, mostly: a vehicle software contract with an automaker outside Hyundai Motor Group. That single disclosure would convert Mobilgene from an internal cost centre into a product with a price, and it is the only realistic path to a margin above the 6% band.
Failing that, watch the second-half operating margin against the 6.0% the company managed in 2024 and 2025. Falling below it while revenue grows 16% would confirm that scale isn't reaching the profit line at all.
And watch the dividend. A meaningful increase would say the group has decided this subsidiary's job is to return cash rather than absorb group spending, which is a smaller re-rating than the software one but a real one.
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.