HANMI FINANCIAL CORP (HAFC) stock intrinsic value fair value estimate & owner earnings analysis

Hanmi Financial Corporation owns Hanmi Bank, a Los Angeles community bank founded in 1982 to serve Korean-American and other immigrant-owned businesses. At December 31, 2025 it had 32 full-service branches and five loan production offices in California, Texas, Illinois, Virginia, New Jersey, New York, Colorado, Georgia and Washington. At June 30, 2026 it had about $8.0 billion of assets, $6.96 billion of deposits and $813 million of shareholders' equity, about $27.37 per diluted share, or about $27.00 after removing $11 million of goodwill. It earns most of its money on the spread between loan yields and deposit costs. Commercial real estate is the largest part of the loan book, at more than 300% of the bank's capital. Smaller earners are commercial and equipment-finance lending and the sale of SBA-guaranteed loans. How we measure earnings. For a bank, deposits are the raw material it lends out, so interest paid is an operating cost. We value owner earnings after all interest, including interest on the holding company's $110 million of 2031 subordinated notes, and we do not subtract the deposits or the notes again. Owner earnings are net income plus depreciation, minus spending on premises, with actual loan losses (net charge-offs) charged as a real cost. Only the part of the loan-loss provision that adds to or releases the reserve is treated as non-cash. On that basis owner earnings per share were $2.08 (2019), $1.86 (2020), $2.70 (2021), $3.38 (2022), $2.58 (2023), $2.10 (2024) and $2.57 (2025). We anchor on the 2019-2025 median of $2.57. That window covers a full cycle for this bank: a large problem loan in 2019, pandemic losses in 2020, the zero-rate peak in 2021-2022, the 2023-2024 squeeze when cash interest paid rose from $116.5 million to $200.5 million, and the 2025 recovery. Recent results run above that anchor. Diluted earnings were $0.75 and $0.79 a share in the first two quarters of 2026, against $1.08 for the whole first half of 2025. Cash interest paid fell to $85.7 million in the first half of 2026 from $92.6 million a year earlier. Owner earnings grew 22.4% in 2025. The base case assumes a smaller 11% step, to $2.85, next year. One offset is already known: the 2031 notes moved from a 3.75% fixed rate to three-month SOFR plus 3.10% on September 1, 2026. Growth is not free for a bank. Each new dollar of loans needs about ten cents of new equity, which comes out of earnings, so every case charges the capital its growth needs. The cash owners can take out is earnings minus that retained capital. That matches how the bank has behaved: it has paid out a little over half its earnings in dividends and buybacks and has funded loan growth without issuing shares, with diluted shares falling from 32.0 million in 2016 to about 29.7 million. Base case, $53.80 a share. Owner earnings are $2.85 next year, then grow 4.5% a year for ten years to about $4.24, with 43% retained to fund that growth. The 4.5% is below the bank's record: owner earnings per share grew about 6% a year from 2016 to 2025. After year ten we credit no further growth, only flat earnings, fully paid out. What sustains them is a 44-year-old deposit and small-business lending franchise, not any assumed expansion. The ten years of payouts are worth $14.39 and the years after $39.42. Bear case, $27.88 a share. A commercial real estate downturn starts immediately. We size it from Hanmi's own 2008-2010 record: in just the first nine months of 2009 and of 2010, net charge-offs were $65.3 million and $86.6 million on a loan book of about $3.0 billion, roughly 5% of loans, and the bank needed a $120 million share sale at $1.20 a share in 2010. Here we charge 5% of today's $6.56 billion loan book ($328 million) over three years and add $60 million to reserves. Pre-provision earnings of about $122.5 million a year absorb nearly all of it, leaving an after-tax loss of about $14.5 million, or $0.49 a share. Dividends and buybacks stop for three years. Year four replaces the lost capital, then earnings settle at $2.02 a share: the 2024 squeeze-year level less the higher cost of the reset notes. We credit no growth after that. This case is worth slightly more than tangible book value of about $27.00. The reason is that even in its weakest recent year, 2024, the franchise earned well above a bond return on its capital. The case keeps today's capital intact, so no new shares are needed. Bull case, $62.49 a share. The current run rate of about $3.10 a share holds, and the franchise grows 6% a year for ten years, in line with its record, retaining 53% of earnings to fund that growth. Earnings are flat after that. The bull case is only modestly above the base because returns of roughly 10-11% on equity leave a thin margin over a bond-like return, so faster growth uses up most of the capital it creates. What the business owes. In every case, deposits and the $110 million of notes are already paid for inside the earnings stream, so the business value and the equity value are the same: $27.88 bear, $53.80 base, $62.49 bull. What changed since our last estimate, two weeks ago. Nothing material in the business. The base and bull cases are unchanged. The bear case is lower, from $29.96 to $27.88, because we now size the credit downturn from the bank's own 2009-2010 loss record, about 5% of loans over several years, instead of 3% over two years. Margin-of-safety price: about $28. We are fairly confident in the earnings level. Seven years that include a pandemic and a severe rate squeeze ranged only from $1.86 to $3.38 a share, and credit costs have been modest since 2020. The worst plausible outcome is a repeat of 2008-2010, when losses ran roughly twice the 5% in our bear case. Losses of 10% of loans would cost about $8.28 a share of capital after tax and earnings, leaving a value of roughly $22, and less than that if regulators required a share sale at depressed prices. The structure helps: total risk-based capital was 15.06% at year-end 2025 against a 10% well-capitalized threshold. The risks are commercial real estate concentration (hotel and office loans included), equipment-finance losses like the $10.1 million charged off in 2025, and floating-rate costs on the subordinated notes. The things we are most likely to be wrong about are the depth of the next credit cycle and whether deposit costs keep falling. If both go against us we land near the bear case, about tangible book. During the holding period the bank pays $0.28 a quarter, about 4% a year at $28. If the base case plays out, from $28 with the shares reaching the base value in five years (about $62), the base case implies roughly 20% a year including payouts, and more if the value is recognized sooner. From a higher price, such as $33, it would imply only about 16.5% a year over five years. The margin-of-safety price is about $28, roughly the value of the bank's own capital: below that, a severe cycle without dilution implies only a modest loss. We could not tie a few small operating cash-flow items (income-tax and other asset timing) to specific events in every year. They are routine for a bank this size and do not change the result.