KIA Corporation (KRX:000270) fell 4.03% on August 27 to close at ₩126,100. Across 390,412,998 shares that puts the whole company at ₩49.23tn, or roughly $33bn at about ₩1,475 to the dollar.
At June 30 the same company reported ₩64.69tn of total equity, ₩16.08tn of cash and equivalents, and ₩56.15tn of retained earnings. It earned ₩4.16tn of net income in six months.
So the market values one of the world's larger automakers at three-quarters of its book value, with a third of that value already sitting in the bank, and at roughly six times what it is currently earning. Either something is badly mispriced or the discount is doing a job. My view is that it's doing a job, and it's worth being specific about what.
Price to book is 0.761. Price to annualised first-half earnings is about 5.9 times; on FY2025's actual ₩7.55tn of net income it's 6.5 times, and on FY2024's ₩9.78tn it would have been 5.0.
The cash generation is the part that makes the multiple look strange rather than merely low. Cash from operating activities was ₩7.17tn in the first half, against ₩5.35tn in the same period of 2025. Purchases of property, plant and equipment were ₩1.85tn. Subtract one from the other and the business threw off roughly ₩5.32tn of free cash in six months, against a ₩49.23tn market capitalisation.
Against ₩106.83tn of total assets sit ₩42.14tn of total liabilities. This is not a leveraged company. It's an automaker with a fortress balance sheet trading below the value of the net assets on it.
Kia is not hoarding, which is the first objection people raise about cheap Korean industrials.
Dividends paid in the first half came to ₩2.64tn. Korean companies typically settle the prior year's annual dividend in the spring, so that figure is essentially the full-year distribution, and it compares with ₩2.56tn a year earlier, ₩2.19tn in FY2024 and ₩1.40tn in FY2023. On the current market cap ₩2.64tn is a yield of about 5.4%. The dividend has nearly doubled in three years.
There's a second, quieter return happening in the share count. Kia's share capital has been ₩2,139.3bn since at least FY2023 and has not moved. At the ₩5,000 par value that implies 427,863,400 shares issued at some point. The company's own July 9 filing reports 390,412,998 shares outstanding with no preferred class at all.
The gap is roughly 37.5m shares, about 8.8% of the issued total. Korean law lets a company cancel treasury stock out of distributable profit, which reduces the share count while leaving share capital untouched, and that mechanism is the only thing that explains the difference. The summary financials don't date the cancellations, so I can't say how much happened in which year. But a company that has retired something close to a tenth of itself while doubling its dividend is not one ignoring its owners.
Combined, the cash going out is substantial. What hasn't happened is any narrowing of the gap between price and book.
The first is earnings quality. FY2024's ₩12.67tn of operating income was a peak produced by pricing power that the whole industry enjoyed and has since lost. FY2025 came in at ₩9.08tn, down 28.3%, and a meaningful part of that decline is US automotive tariffs landing on a company that sells 81.8% of its volume outside Korea. A market that discounts earnings it doesn't believe will persist is behaving rationally, not stupidly.
The second is ownership. Hyundai Motor holds 137,318,251 Kia shares, or 35.17%. Add the chairman's 1.81% and the other affiliated holders and the largest-shareholder group reaches 36.99%. Kia in turn holds stakes in other group companies, and over this summer alone it filed several disclosures covering investments in and securities purchases from related parties.
That's the structure a minority shareholder is buying into. Capital allocation decisions at Kia are made with the group's interests in view, and the group's interests include supporting affiliates, funding shared platforms, and maintaining a cross-holding arrangement that Korean regulators have pressed the family to unwind for two decades without success. None of that is illegal or even unusual in Korea. It is a reason a rational buyer pays less per won of book value here than for a company where the board answers to the float.
The third is what the cash is for. The electric vehicle transition and the localisation of US production are both capital-hungry, and Kia's ₩16.08tn of cash is not idle treasure so much as a war chest against costs that have not been incurred yet. Capital spending ran ₩3.76tn in FY2025 and ₩1.85tn in the first half of 2026. A buyer who values the cash at face value is assuming it comes back to shareholders. Much of it will not.
Two arguments push back.
The first is comparative. Kia is not uniquely discounted. Legacy volume automakers around the world trade at low single-digit earnings multiples and below book, because the market is pricing an industry it believes is being disrupted. Judging Kia against a general market multiple rather than against Ford, Stellantis or Nissan overstates how anomalous it is. On that reading the stock is roughly where its peer group sits, and the interesting question is about the sector rather than the company.
The second is that Kia's operating performance has been better than the peer group's. A 7.95% operating margin in FY2025, in a year the company describes as tariff-damaged, is at or above what most volume manufacturers manage in good years. Electrified vehicles reached 35.3% of sales in the second quarter with volumes up 60%, which is a faster transition than most of the peer set has achieved without destroying margin. If you accept that Kia executes better than the companies it's priced alongside, the discount is a sector label applied to a company that has outgrown it.
Both arguments are fair. Neither of them is a catalyst, which is the whole problem with cheap stocks in structures like this one.
The share count. It's the cleanest signal available, because it can't be spun. If the next annual report shows fewer than 390,412,998 shares, management is converting the balance sheet into per-share value at a price well below book, which is the single most accretive thing it could do. If the count is unchanged and the cash pile has grown again, the discount is being earned.
The second marker is the disclosure of any move on the group's cross-shareholding structure. Hyundai Motor Group has floated holding-company conversions before and pulled them. A concrete proposal that clarified where Kia sits, and what happens to its stakes in affiliates, would remove the largest single reason a foreign investor applies a discount here. Nothing in the filings suggests one is imminent.
Neither of those is a forecast. They are the two things that would tell you whether the market is wrong or simply patient.
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.