Invest1 distinct publisher2 min readUpdated
Interest income rose 8.3% to 32.2 trillion won in the first half, but a 5.7 trillion won swing in securities results turned that into a 6.4% profit decline.
The Investor · Invest desk

Compiled by The InvestorSomething wrong?How this is made
The securities line moved 5.7 trillion won year on year, from a 3.2 trillion won profit to a 2.5 trillion won loss [1]. The entire improvement in lending revenue was 2.5 trillion won [4]. The bond book was therefore worth roughly 2.3 times as much to the reported result as everything the loan book gained [2], and it pointed the other way.
The FSS attributes both movements to the same cause. Higher market rates lifted net interest margin to 1.56% from 1.52% [5], and higher market rates enlarged the losses on securities [6]. The asymmetry is in the speed. A bond portfolio remarks at once; a loan book reprices in basis points over quarters. Four basis points of margin on one side [5], a 4.5 trillion won swing in valuation results on the other [3].
It is worth separating volume from price inside that 8.3% interest income gain. Interest-earning assets grew 6.4% [5], while the margin improvement is worth about 2.6% in relative terms [4]. Most of the extra interest income came from lending more, not from lending at better spreads.
The four lines the FSS breaks out roughly account for the drop. Interest income up 2.5 trillion won [4], non-interest income down 2.3 trillion [6], selling and administrative expenses up 700 billion [9], loan-loss expenses up about 280 billion [5]: a net 780 billion won of pre-tax deterioration [6], against a 900 billion won fall in net profit [2].
A year ago, non-interest income equalled 17.5% of interest income. It is now 9.0% [9]. Whatever these banks earn outside the spread between deposits and loans has become a rounding difference next to what they earn from it.
Underneath, the credit line grinds. Delinquency reached 0.56% at the end of June, and each reading in the FSS series since 0.25% at the end of 2022 has been higher than the one before it [11]. Loan-loss expenses of 3.5 trillion won were already 8.6% up on last year [10]. The regulator said it plans "to continue encouraging banks to build up their loss-absorbing capacity, including through provisioning, so that they can maintain soundness even against unexpected shocks" [12].
The one group that grew was internet-only banks, up 15.1% to 400 billion won [c3a], which is about 3% of sector net profit [10]. Their gain does not move the sector's arithmetic. The marks on the bond portfolios do.
Follow any of these and your For You feed starts watching them — no settings page required.
Ranked by verification strength, evidence, and original report placement.
The Financial Supervisory Service released preliminary first-half 2026 operating results for domestic Korean banks on the 23rd.
Domestic banks posted first-half net profit of 13.8 trillion won ($9.9 billion), down 900 billion won or 6.4% from 14.7 trillion won a year earlier.
Net profit at commercial and regional banks fell 3.3% to 9.1 trillion won; nationwide commercial banks earned 8.1 trillion won, down 3.8%; regional banks earned 600 billion won, down 8.5%; specialized banks earned 4.6 trillion won, a 12.0% decline.
Internet-only banks earned 400 billion won in the first half, up 15.1% from a year earlier, the only bank group in the FSS breakdown to post growth.
First-half interest income rose to 32.2 trillion won, up 2.5 trillion won or 8.3% from a year earlier.
The interest income gain reflected a 6.4% increase in interest-earning assets such as loan receivables, along with a rise in net interest margin to 1.56% from 1.52%, up 0.04 percentage point, on higher market rates.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Primary regulator data, single relay
Every quantitative claim traces to one dataset — the FSS's preliminary first-half 2026 operating results — reported with specific levels, year-on-year deltas and a five-year delinquency series, and the disclosed drivers reconcile to within about 120 billion won of the reported 900 billion won profit decline. Strength is the primary-regulator provenance and internal arithmetic consistency; the limits are that only one publisher carries it, the figures are explicitly preliminary, and no named-bank or classification detail is available to test the securities-loss interpretation.
No adoption signal in scope
This is sector financial reporting, not a technology or product story: the supplied source contains no releases, deployments, benchmarks, pricing or usage disclosures that would constitute adoption evidence. Growth in internet-only bank profit is an earnings datapoint, not an adoption observation, and the source gives no user, volume or deployment detail behind it, so no adoption score is derivable.
Framing tracks the arithmetic, risk trend understated
The cluster's central framing — margins widened while profit fell because of the bond book — is arithmetically supported by the reported 5.7 trillion won securities swing against a 2.5 trillion won interest income gain, so there is no promotional overstatement. If anything the presentation is slightly flat relative to its own data: a delinquency rate that has more than doubled since end-2022 and an explicit regulator push for more provisioning are reported without emphasis, which argues for a marginally negative gap.
Regulator framing with a supervisory agenda
The data originates with a supervisor that is simultaneously arguing banks should provision more heavily and that it will tighten monitoring, giving it a mild interest in emphasising soundness pressure alongside falling profit; the reporting outlet relays that framing without an independent check or bank rebuttal. Offsetting this, the numbers are standardised regulatory aggregates with no commercial promotion attached and no vendor or issuer benefiting from the narrative, so distortion pressure is moderate rather than high.
Solid figures, single relay, preliminary basis
Confidence is held up by precise, internally consistent regulator figures and a multi-year risk series, and held down by three things: one publisher, explicitly preliminary results subject to revision, and the absence of the detail needed to interpret the key driver — whether the securities losses are realised or mark-to-market, and how the portfolios are classified. The direction of travel (core margin up, marks down, credit costs rising) is well supported; the durability of the securities hit is not testable from this material.
invest
Korea's brokerages added 402 staff in a quarter while mortgages headed to 5.78%1 distinct publisher
invest
Seoul's first EDCF-funded AI project is a $170m tied loan, and Tanzania holds the debt1 distinct publisher
invest
North Korea's 2015 shock cost 1.1% of output and a distribution that did not come back1 distinct publisher
invest
Bank of Korea says a $936bn swing in net external assets is arithmetic, not a solvency signal1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
en.sedaily.com
1 article · August 22, 2026