Two lines in Samsung Life Insurance Co., Ltd.'s (KRX:032830) accounts have been moving in different directions for four years, and between them they explain more about the underlying economics of this company than anything on the balance sheet.
Interest income: ₩7,221.8bn in FY2021, ₩7,748.6bn in FY2022, ₩8,437.0bn in FY2023, ₩8,461.0bn in FY2024, ₩8,406.6bn in FY2025.
Interest expense: ₩611.5bn, ₩1,119.2bn, ₩1,538.8bn, ₩1,788.4bn, ₩1,908.2bn.
The first line rose 16% and then stopped. The second more than tripled and is still going.
Subtract one from the other and the picture is stark. Net interest was ₩6,610.3bn in FY2021 and ₩6,498.4bn in FY2025. After four years of rising rates, a growing asset base and a bond portfolio repricing into higher yields, this company earns slightly less net interest than it did before any of that happened.
The quarterly detail says the squeeze is current, not historical. Interest income has sat between ₩2,078.7bn and ₩2,145.3bn for eight consecutive quarters, a band of about 3%. Interest expense in the second quarter of 2026 was ₩595.9bn, the highest in the series and up 18.8% from ₩501.6bn a year earlier.
So the most recent quarter shows flat income against the fastest expense growth in two years.
For a US reader used to bank accounts, this line needs translating, because it is not mostly borrowing cost.
Under IFRS 17, a life insurer's liabilities are the present value of promises to policyholders stretching decades ahead. Those liabilities are discounted, and every period the discount unwinds by one period's worth. That unwind is an expense, and it appears in the finance lines rather than in the insurance service result.
Which means interest expense at a company like this is largely the cost of time passing on obligations already incurred. It grows with the size of the liability book and with the discount rates applied to it, and it is not something management can decide to spend less on.
The uncomfortable implication is that the insurer's economics depend on the spread between what its assets yield and what its liabilities cost to carry, and that spread has been narrowing. Assets reprice slowly, because a life portfolio holds long bonds bought years ago. Liabilities reprice through the discount rate immediately.
I should note that the summary financials do not break the expense line into insurance finance expense and ordinary borrowing cost, and the notes to the half-year report are where that split sits. The magnitude and the growth pattern point at the former.
Cash from operating activities in the first half of 2026 was negative ₩1,265.7bn. In the first half of 2025 it was positive ₩2,512.9bn. A swing of nearly ₩3.8tn.
Before treating that as alarming, split the half. The first quarter alone consumed ₩3,087.7bn. The second quarter therefore generated roughly ₩1,822.0bn, which put the year-to-date figure back toward the surface.
So this is one very bad quarter followed by a normal one, not a deteriorating trend. Full-year operating cash flow was ₩5,256.3bn in FY2025 and ₩4,999.6bn in FY2024, and the pattern at insurers is lumpy because bond purchases and sales, policyholder flows and reinsurance settlements all land unevenly.
What the company did in the meantime is worth noting. Financing activities brought in ₩2,614.5bn across the half, roughly double the ₩1,306.5bn of a year earlier. Cash still ended the period at ₩4,796.9bn, up from ₩4,537.8bn at December 31.
Dividends paid reached ₩1,018.5bn in the first half of 2026. That is more than the ₩874.4bn paid across the whole of FY2025, and 16.6% above the ₩873.8bn paid in the first half of last year.
The trajectory has been steep and consistent: ₩491.2bn in FY2021, ₩592.7bn in FY2022, ₩597.4bn in FY2023, ₩723.1bn in FY2024, ₩874.4bn in FY2025. Up 78% in four years.
Against a market capitalisation of ₩61.40tn at the August 27 close of ₩307,000, the first-half payment alone is a yield of about 1.66%.
Paying a rising dividend in a half-year that consumed operating cash, funded by financing inflows, is a choice. At a company with ₩4.8tn of cash and ₩147tn of reported equity it is an easy one to make, and the signal it sends about management's confidence is probably worth more than the cash it costs. It is still worth noticing that the money came from outside the business in that particular six months.
The honest counter is the one that applies to everything about this company. Samsung Life's share price is a function of the Samsung Electronics stake and of Korean insurance legislation, not of a hundred basis points of net interest.
Investors are not buying a spread business. They are buying a claim on shares worth more than twice the market capitalisation, wrapped in an insurer, at 0.42 times reported book. Whether net interest is ₩6.5tn or ₩6.8tn barely registers against that.
The reason to look anyway is that the spread determines whether the insurance business can keep funding a rising dividend without help. If net interest keeps eroding while the insurance service result also falls, which it has, then the operating company stops being self-sustaining and the dividend starts depending on the balance sheet rather than on earnings. That changes the character of the investment even if it never moves the share price.
The interest expense line in the third quarter. It has risen every fiscal year since FY2021 and hit a record ₩595.9bn in the second quarter of 2026. Another step up, against flat interest income, would mean the spread compression is accelerating rather than stabilising.
The second thing is full-year operating cash flow, reported in February. FY2024 and FY2025 both cleared ₩5tn. A figure well below that, after a first half that consumed cash, would mean the ₩3.8tn swing was more than one difficult quarter.
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.