Hanwha Systems Co., Ltd. (KRX:272210) had the quarter its backlog had been promising for two years. June-quarter revenue of ₩1,117.6bn was up about 45% on a year earlier. Operating profit of ₩103.7bn was more than triple the ₩33.5bn of the same quarter in 2025, and gross margin came in at 16.5% against 12.2% a year before. For a company that makes radars, fire-control electronics and naval combat systems, that is the shape of an export cycle finally landing in the accounts.
Book equity fell ₩471bn.
Not a typo, and not a dividend. Total equity was ₩5,042.8bn at the end of March and ₩4,571.7bn at the end of June, because total comprehensive income for the quarter was negative ₩473.8bn while net income was a positive ₩52.7bn. The gap — more than half a trillion won in three months — has nothing to do with radars.
That gap is the thing worth arguing about. Hanwha Systems reports two businesses that share a page: a defense electronics operation that is improving in a way you can follow line by line, and a large marked-to-market investment book whose swings are several times larger than anything the operating business produces. Most write-ups on the stock quote the first and price the second without saying so.
Start with what is actually working. Revenue has gone from ₩858.6bn in 2017 to ₩3,664bn in 2025, and the first half of 2026 alone did ₩1,924.7bn. Operating profit in the first six months of this year, ₩137.9bn, has already passed the full-year 2025 figure of ₩119.9bn.
The margin story is more interesting than the growth story. Gross margin was 10.7% for all of 2025 — the worst year in the dataset — and 16.5% in the June quarter. In between sat a genuinely bad fourth quarter of 2025, when ₩1,398bn of revenue produced ₩5.8bn of operating profit. Revenue for the full year rose 30.7% while operating profit fell 45%. That is what a defense company looks like when it takes volume at whatever margin the contract offers and absorbs cost overruns on newer programs.
The mix is what changed. Company disclosure has put defense export share above 30%, with Middle East shipments of the Cheongung-II multi-function radar and AESA radar work for the KF-21 fighter behind it. Export contracts carry different economics than domestic Korean defense procurement, which is priced on a cost-plus basis with the margin negotiated by the government's procurement agency rather than the market. That single structural fact explains most of the margin gap, and it's the thing a US reader would not think to check — a Korean defense contractor's domestic backlog and its export backlog are not the same product at all.
Total order backlog was around ₩11.3tn at the end of June, roughly ten times a quarter's revenue, off a record ₩12.2tn in March. Defense backlog was about ₩9.25tn in the first quarter. Multi-year visibility, in other words, against defense revenue running in the low-to-mid ₩2tn range annually.
Now the other company. Line up operating income against pretax income and the relationship falls apart.
In 2024, operating income was ₩219.3bn and pretax income was ₩581.4bn. In 2025, operating income was ₩119.9bn and pretax income was ₩115.0bn — but net income came in at ₩209.1bn, higher than pretax, because the tax line was a benefit of ₩85.7bn rather than a charge. The March 2026 quarter earned ₩34.3bn of operating profit and reported a pretax loss of ₩115.7bn. The December 2025 quarter earned ₩5.8bn and lost ₩99.8bn pretax.
Finance costs alone don't cover this. They ran ₩43.1bn in the June quarter against ₩26.2bn of finance income, which is meaningful but an order of magnitude short of the swings.
The comprehensive income line is where the size shows. Full-year 2025: net income ₩209.1bn, total comprehensive income ₩2,397bn. The company's reported earnings that year were less than a tenth of the change in its economic value as the accounts measure it. Then it reversed — the June 2026 quarter's negative ₩473.8bn.
Hanwha Systems holds investment positions, including a stake in a low-earth-orbit broadband venture, and sits inside a group that has been reorganizing its shipbuilding and aerospace assets. Whatever the specific holdings, the accounting consequence is plain: this share is partly a levered claim on a portfolio, and the portfolio is repriced quarterly.
Total assets were ₩5,810bn at the end of 2024, ₩10,327bn at the end of 2025 and ₩11,847bn in June. Non-current assets more than doubled in 2025, from ₩3,524bn to ₩7,399bn. Equity went from ₩2,456bn to ₩5,001bn.
Share capital did not move. It has been ₩944.6bn since 2021. So none of that equity growth came from selling stock; it came from retained earnings of about ₩119bn and from revaluation.
The funding came from liabilities. Total liabilities went from ₩3,355bn at end-2024 to ₩7,276bn in June. Current liabilities alone rose ₩1,840bn during the June quarter, to ₩5,354bn, alongside a ₩1,304bn financing inflow that took cash from ₩493bn to ₩1,591bn. Liabilities now exceed equity by about 1.6 to one, against roughly 1.4 to one at the end of 2024.
The cash flow statement says why the borrowing was needed. In 2025 the company generated ₩78.7bn of operating cash flow, spent ₩298.9bn on property and equipment, and paid ₩65.4bn of dividends. Financing brought in ₩1,237bn. In 2024 the pattern was the same, smaller: ₩132.6bn in, ₩239.8bn of capex, ₩52.4bn of dividends. Operating cash flow has not covered capex in either year, and the June quarter's ₩152.3bn against ₩74.0bn of capex is the first half-decent showing in a while.
The shares closed at ₩66,900 on September 2, down 5.2% on the day, for a market capitalization of ₩12.64tn — roughly $9.2bn at about 1,370 won to the dollar, which is a rough rate. The 52-week range runs from ₩45,350 to ₩162,700, so the stock sits about 59% below its high and 48% above its low. That range is itself a warning about how much of the price is portfolio and how much is operations.
Trailing four-quarter revenue of ₩4,130bn puts the shares near 3.1 times sales. Trailing operating profit of ₩166.2bn puts them at about 76 times, though that number is dragged down by the December quarter. Trailing net income is ₩78.2bn, which produces a P/E near 162 and tells you nothing at all.
Price to book looks reasonable at about 2.8 times. Be careful with it. When a large share of book value is financial assets carried at fair value, price-to-book compares one market price against another market price. It is not the same statistic it is for a manufacturer whose book is factories at cost.
Philly Shipyard is the near-term one. Hanwha took over a troubled US yard, and the company has said the full turnaround comes in 2027 rather than 2026, as the loss-making vessel types finish delivery and special-purpose work replaces them. A US Missile Defense Agency order for missile range instrumentation vessels is the anchor of that plan. Shipbuilding integration is where confident schedules go to slip, and the losses run through a segment that has nothing to do with the radar franchise investors think they own.
Leverage is the second. The company has been funding capex and dividends with borrowings during a period when rates are not free, and current liabilities of ₩5,354bn against current assets of ₩4,666bn means working capital is negative.
And the mark-to-market cuts both ways, which is the honest version of the argument above. The ₩2.4tn comprehensive income in 2025 was as real as the ₩473.8bn reversal in June. If the portfolio has been the reason the stock outran the operating business, it can be the reason it lags for a while too.
The third-quarter report is a clean test, because two things are separable in it. Whether gross margin holds near 16% on export mix, or falls back toward the low teens — that settles whether the June quarter was the new run rate or a delivery bunching. And whether backlog holds at ₩11.3tn after slipping from the March record.
Then the segment detail: the Philly Shipyard loss, quarter over quarter, against the 2027 turnaround claim.
Last, the simplest one. Whether net income and comprehensive income land anywhere near each other. Until they do, the earnings line on this company is a number you should read past rather than read.
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.