Every Korean bank reported a record first half in July 2026, and after a while the press releases blur together. Shinhan Financial Group Co., Ltd. (KRX:055550, NYSE:SHG) earned ₩3,442.7bn attributable to its shareholders over the six months, up 13.3%, with ₩1,820.1bn of it in the second quarter alone. So did everyone. Records were general.
What wasn't general is where the money came from. Shinhan's net interest income grew 7.7% year on year and rose again sequentially in the quarter. At KB Financial Group, the only Korean holding company of comparable size, the same line grew 1.7% over the half and fell 5.7% quarter on quarter. Both reported records. Only one of them is still earning them from lending.
Net interest income was ₩6,158.5bn for the half, against ₩5,718.8bn a year earlier. The half-year report's three-month column lets you split it: ₩3,024.1bn in the first quarter, ₩3,134.4bn in the second, a 3.6% sequential gain.
The composition matters more than the total. Interest income rose 2.3% to ₩14,354.1bn while interest expense fell 1.5% to ₩8,195.6bn. Both sides moved the right way. That's not what a falling-rate environment usually delivers — the standard pattern is that asset yields drop and the bank survives on funding costs falling faster, which is exactly what KB's numbers show, with interest income down 1.5%. Shinhan grew the top line of its spread business while shrinking the cost.
Volume explains part of it. Loans measured at amortised cost went from ₩464,773.9bn at the end of December to ₩480,920.2bn at the end of June, up 3.5% in six months. Deposits grew faster, 6.2% to ₩475,494.1bn, which leaves the group with a loan-to-deposit ratio near 101% and more funding than lending demand — the same imbalance every Korean bank is carrying right now.
Volume doesn't explain all of it. A 3.5% loan book with a flat margin gives you roughly 3.5% more interest income, not 7.7% more net interest income. Something in the mix improved. The half-year summary statements don't segment the margin by subsidiary, so I can't tell you whether that's Shinhan Card repricing consumer credit, the Vietnam operation, or the bank itself. The segment note would say. I haven't read it.
Being right about the lending line doesn't make the rest of the income statement clean.
Net fee and commission income rose 47.9% to ₩2,129.8bn. Gains on financial instruments at fair value through profit or loss nearly doubled, from ₩1,366.6bn to ₩2,725.9bn. Foreign exchange trading gains rose 63.6% to ₩645.0bn. Those three lines together added roughly ₩2.5tn year on year, which is more than the entire ₩672.3bn increase in operating profit, and the reason it doesn't show up as such is that other lines swung hard the other way.
The biggest of those is insurance. Insurance finance charges ran to negative ₩2,387.4bn for the half against negative ₩417.4bn last year. That looks alarming and mostly isn't: in other comprehensive income, insurance contract net finance gains were positive ₩3,683.7bn against negative ₩792.8bn. IFRS 17 splits rate-driven movements in an insurance book across the income statement and OCI, and the two are moving against each other by design. Equity absorbed the net effect. The reported operating profit did not.
The insurance service result itself — the underwriting business, stripped of rate effects — fell 17.4%, from ₩519.9bn to ₩429.4bn. That's a real deterioration and it's small enough to be easily missed underneath the ₩2.4tn accounting swing above it.
Costs are the other soft spot. General and administrative expenses rose 10.9% to ₩3,216.1bn, faster than net interest income at 7.7%. A bank growing its cost base faster than its core revenue is running on the non-core lines to hold the ratio, which is precisely what the trading and fee numbers above were doing this half.
Credit provisions came in at ₩976.4bn against ₩1,057.2bn, down 7.6%. KB's fell 22.7% over the same period. Shinhan got much less help from provisioning, which cuts both ways: less of its earnings growth was borrowed from the credit cycle, and it has less room to release if the cycle turns.
The market capitalisation is ₩50.4tn on 469,450,239 shares at ₩107,400. First-half basic earnings per share were ₩7,075; annualise and the multiple is about 7.6x.
On book, equity attributable to the parent was ₩60,100.7bn at the end of June. Take out the ₩4,163.9bn of hybrid capital sitting inside that figure — Additional Tier 1 instruments, excluded from the CET1 ratio and not common equity in any regulatory sense — and common equity is roughly ₩55,936.9bn. That's about 0.90x book, against 1.04x on the same measure at KB.
So the cheaper of the two on book is the one whose lending is growing. That's an odd result and it deserves an explanation rather than a conclusion. Part of it is capital: Shinhan's CET1 ratio was 13.43% at the end of June against KB's 13.74%, and in a market where both groups have tied shareholder returns to capital above a threshold, 31 basis points of headroom is worth a valuation premium on its own. Part of it is goodwill, which is a longer argument than fits here.
The straightforward risk to any cheapness case is the chart. At ₩107,400 the shares are within 3% of a 52-week high of ₩110,000, off a low of ₩64,000. That's a 68% move. Whatever discount to fair value existed a year ago has largely been collected, and the remaining case rests on earnings continuing rather than on a re-rating that hasn't happened yet.
The bear case is simpler than the mix analysis suggests. Roughly ₩2.5tn of the year-on-year improvement came from fee income and trading — market-sensitive lines that fell hard in 2022 and will fall hard again. Net interest income growing 7.7% is genuinely better than flat, but it isn't enough on its own to hold group profit up if brokerage and trading revert. Shinhan is less exposed to that than KB, whose entire operating profit increase was matched by its fee line. It is not unexposed.
The third-quarter report, due in mid-November, and one number in it above all: net interest income against the ₩3,134.4bn printed in the second quarter.
Two prints in a row of sequential growth would establish that the first half wasn't a repricing artifact — that Shinhan has found something in its funding mix or its loan mix that its peer hasn't. A third quarter back below ₩3.05tn would mean the second quarter was the peak of the repricing benefit and Shinhan is on the same path as KB, just a quarter or two behind.
Secondary, but it settles the cost question: general and administrative expenses ran at ₩1,670.7bn in the second quarter. If that keeps growing near 11% while the revenue lines that carried this half normalise, the cost-to-income ratio does the damage before any credit event does.
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.