Shinhan Financial Group Co., Ltd. (KRX:055550) filed a ₩700bn buyback on July 23. The filing says something most buyback filings don't. After this one completes, the company will review further purchases this year under its 2026 corporate value-up plan. The bank is telling the exchange that it is behind on a promise. It says in a regulatory document that it intends to catch up.
The value-up plan is Korea's version of Japan's governance push. Companies publish targets for return on equity, payout and capital, and file them with the exchange. Some 756 listed companies have done so, covering 87.7% of the KOSPI's market value. Shinhan's plan promised four things. This piece checks each one against the numbers on file.
The plan's headline target was a 10% return on equity, up from 8.6% when it was written. First-half 2026 net income from the consolidated statements was ₩3,495.8bn. Annualise it against total equity of ₩62,519.8bn at June 30 and the return is 11.2%.
That figure is generous in one way and conservative in another. It uses total equity, which includes hybrid capital and minority interests. So the return on common equity is somewhat different. It also uses a half that included a record second quarter of ₩1,846.7bn. Korean press citing the results call reports Shinhan itself guiding to about 10.6% for the year. Either way, the 10% target is met.
The trend supports it. Annual net income ran ₩4,112.6bn and ₩4,755.5bn. It then ran ₩4,478.0bn and ₩4,558.2bn. The final figure was ₩5,084.5bn. This sequence runs from FY2021 to FY2025. The first half of 2026 alone is 69% of last year's total. Net interest income rose 3.6% quarter on quarter to ₩3,134.4bn. Fee income jumped to ₩1,188.9bn from ₩940.8bn in March.
The plan promised a total shareholder return of 50%. The July results call set the 2026 figure at ₩2.8tn plus an unspecified addition. Korean press puts the implied ratio at 52.5%.
Cash dividends are running at ₩740 a share per quarter, which an earlier piece here covered. Dividends paid in FY2025 were ₩1,293.8bn. The buyback side is where the money is moving. The July filing brings cumulative 2026 buybacks to ₩1.4tn, against ₩1.25tn for all of 2025.
The filing shows plenty of room. Distributable profit at the end of FY2025 was ₩4,125.1bn. After the ₩700bn already bought this year and dividends declared, the remaining limit for treasury stock purchases was ₩2,658.8bn. The company could buy back nearly four times the current programme without breaching the legal limit.
The plan promised to cut shares outstanding to 450 million by 2027. The count is 469,450,239 now. That leaves 19.45 million to retire, or 4.1% of the company. Those shares are worth about ₩2.1tn at Friday's close of ₩110,300.
The buyback filing gives the 2026 record. The company bought 10,840,573 shares between July 2025 and January 2026. It cancelled them on February 6. It bought another 5,204,122 between February and July and cancelled them on July 15. That is 16.04 million shares retired this year before the current programme started. The ₩700bn now underway will buy about 6.3 million more at Friday's price. That takes the 2026 total past 22 million.
At that pace the 450 million target is reachable by late 2027. It is not reachable by the end of 2026. So the target stands, and the filing's promise to review more purchases in October is the company saying it knows the arithmetic.
One structural note. A Korean law change effective March 6 requires companies to cancel treasury shares within a year of buying them. Shinhan was already cancelling everything it bought. The law now removes the option of holding shares in treasury as a governance tool. For a bank with no controlling shareholder, that was never much of a temptation anyway.
The plan set a CET1 management range of 13.0% to 13.4%. Common equity tier 1 is the regulatory capital ratio that decides how much a bank can return. Shinhan reported 13.43% at June 30, which sits just above the top of its own range.
That reads as a problem only if you want the bank to hold more capital. The range is a ceiling as much as a floor. Every ten basis points above 13.4% is capital the plan says the bank doesn't need and can return. On ₩62.5tn of equity, the arithmetic is loose. But the direction is clear. The bank is generating capital faster than it is handing it back. That is why the October review exists.
Here is the part that doesn't add up. The company's market capitalisation was ₩51.8 trillion at Friday's close. Book equity was ₩62.5tn. The shares trade at 0.83 times book, and at 7.4 times annualised first-half earnings.
The stock has done well. It closed Friday 71% above its 52-week low of ₩64,400. It touched a high of ₩114,500 on Thursday before slipping 3.7% on Friday. A US fund, Capital Group, crossed 5% in early September to become the largest foreign holder. The value-up plan has been noticed.
But this bank earns 11% on equity and pays out half of it. It shrinks its share count by 4% a year. Such a bank would trade above book in most markets. JPMorgan trades near two times. Shinhan trades at 0.83. The discount to KB, which an earlier piece here traced to goodwill, is a separate matter. The discount to book is the market saying it doesn't believe the return is durable, or that the capital returned won't keep coming.
I think the filings argue against the second doubt. The company has retired 16 million shares this year and filed for 6 million more. It has put in writing that it will review further purchases. The first doubt is the real one, and that is about Korean interest rates, not about Shinhan.
The Bank of Korea has been cutting. Net interest income grew 3.6% in the June quarter, but a bank's spread compresses when policy rates fall faster than deposit rates. An earlier piece here framed the third quarter as the test of whether Shinhan's interest income keeps growing. That test is still open.
Credit costs are the second risk. Provisions for credit losses were ₩456.3bn in the June quarter, down from ₩520.1bn in March. Korean household debt and the real estate that secures it are the exposure. A rise in provisions would eat the return on equity that the plan rests on.
The share count target could also slip without breaking any rule. The plan says 2027. A bank that reaches 455 million by then has kept the spirit and missed the letter, and the market will notice the difference.
And there is the withholding-tax question on the dividend for US holders of the ADR. It is covered separately here. It affects the after-tax yield but not the plan itself.
The buyback filing names its own checkpoint. After the ₩700bn completes, by October 22 at the latest, the board will decide whether to add more this year. A further programme of ₩500bn or above would put the 450 million share target within reach by mid-2027. No further programme would push it to the deadline.
The third-quarter results in late October will show CET1 again. If it is still above 13.4% after ₩700bn of buybacks, the bank has more room than its plan admits. In that case, the market's 0.83 times book is pricing a return that keeps arriving.
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.