OLD SECOND BANCORP INC (OSBC) stock intrinsic value fair value estimate & owner earnings analysis

Old Second Bancorp is a community bank based in Aurora, Illinois, serving the western Chicago suburbs. It had about $5.4 billion of deposits at June 30, 2026, of which roughly $1.75 billion (about a third) pay no interest at all. That cheap funding is the heart of the business: it let net interest margin rise from 4.64% in 2023 to 4.96% in 2025. The loan book was $5.25 billion at the end of 2025: investor commercial real estate 23%, commercial and industrial 16%, owner-occupied commercial real estate 14%, leases 10%, multifamily 6.5%, construction 3%, and consumer loans acquired with Evergreen Bank Group in July 2025 (powersport vehicles $697 million or 13%, plus collector cars, manufactured homes and solar about 4%). Two acquisitions reshaped the company: West Suburban Bancorp (December 2021, which took the share count from about 30 million to about 45 million) and Evergreen (July 2025, to about 52 million). Earnings power. Per diluted share, earnings were $0.52 in 2016, $0.50 in 2017 (depressed by a one-time tax-law charge), $1.12 in 2018, $1.30 in 2019, $0.92 in 2020 (pandemic provisions), $0.65 in 2021 (merger costs), $1.49 in 2022, $2.02 in 2023, $1.87 in 2024 and $1.62 in 2025. We treat a bank's reported net income as its owner earnings, since stock compensation is already deducted and the cash-flow statement is dominated by loan and deposit movements. Because the company was transformed by the West Suburban deal, we measure from 2022 and use the middle of 2022 to 2024, $1.87 per share. We set 2025 aside because it carried about $28 million of one-time costs: $15.1 million of Evergreen acquisition expenses and a $13.2 million up-front provision on the acquired loans. The first half of 2026 supports that choice: net income was $53.8 million ($1.02 per share) against $0.91 in the first half of 2025, and the second quarter earned $0.54. That is above our level even though credit provisions rose to $17.0 million in the half (from $4.9 million a year earlier). Growth check. Per-share earnings fell 13.4% in 2025 ($1.62 versus $1.87), while our base case starts at $1.87 for year one, a 15% rebound. The dated reason is the one-time Evergreen costs disclosed in the 2025 annual report and the first-half 2026 results, which are 12% above the prior year. After that we credit no growth at all. Old Second's loans actually shrank from 2023 to 2024, deposits fell about 3% in the first half of 2026, and the growth of recent years came from large acquisitions and a rate cycle, neither of which is a steady, repeatable process we can count on. Growth from the Evergreen consumer platform, national expansion of that lending, margin expansion and share buybacks (a 1.9 million share authorization exists, but only 327,000 shares were bought in 2025) are all left out. Base case, $25.02 per share. Earnings hold at $1.87 for ten years, then fade 3% a year afterwards. Holding is credited because the low-cost deposit base has been stable through the 2022-2023 rate shock: non-interest-bearing deposits were 31% of deposits at the end of 2025 and 32% at June 30, 2026 even as total deposits fell. The fade after year ten reflects that nothing we can evidence guarantees a deposit advantage beyond what the record shows. As a bank, interest is an operating cost, so no debt is subtracted: the business is worth $25.02 and the lenders (depositors) are paid inside the earnings. Bear case, $13.04 per share. We include a full credit cycle modelled on the early-2010s: the company reported net losses as recently as 2012, and U.S. bank charge-offs peaked around 3% of loans nationally in that cycle, with Chicago-area commercial real estate lenders worse. About a third of Old Second's loans are investor commercial real estate, multifamily or construction, and the $697 million powersport book is new and untested through a recession, so we use net charge-offs rising to 4% of loans for two years (about $210 million a year, against $16.2 million in 2025), with lower interest income as rates fall. Earnings per share run about $0.52 in year one, a loss of $0.96 in each of years two and three, $0.50 in year four and $1.32 in year five, which is roughly break-even over five years and leaves capital intact with no stock issuance needed (shareholders' equity is $903 million). Afterward earnings settle at $1.40, 25% below our base level, reflecting heavier ongoing loan losses and tougher deposit competition, fading 3% a year after year ten. This bear value is slightly below tangible book value of about $14.43 per share (equity of $902.8 million less $129.2 million of goodwill and $21.3 million of intangibles, over about 52.1 million shares), which is consistent with a franchise earning little more than its cost of capital through a downturn. Bull case, $30.10 per share. The company's own levers go right: provisions return from $34 million a year annualized to a normal level and the second-quarter run rate of about $2.16 is held, which we round to $2.25 per share, flat for ten years then fading 3% a year. We do not credit a national roll-out of the consumer platform. What would make us wrong. The most important uncertainty is credit quality: nonperforming loans rose 74% in 2025 to $52.8 million (1.0% of loans), and 2025 charge-offs included several powersport loans, three commercial credits and one commercial real estate credit. The allowance for credit losses was raised to 1.4% of loans from 1.1%. A second is margin: 4.96% is high compared with the past, and the 2018 to 2019 record (return on assets high even with lower margins) is the evidence that earnings can survive narrower spreads, but with an older, smaller balance sheet. Prior estimates (bear $14.50, base $28.75, bull $51.25) were built on multi-year growth rates we can no longer support as secured, so they are not comparable. Margin-of-safety price, $11.00. We are fairly, but not fully, sure of the $1.87 level: the 2022 to 2024 record sits between $1.49 and $2.02, and two quarters of 2026 are consistent with it. We are far less sure of the consumer book and of the margin. The worst plausible event is a Chicago-area real estate recession combined with losses on the newly acquired consumer loans; the company's capital would absorb that without dilution, but it did record a net loss as recently as 2012, so an outcome of the bear kind is not remote. If we are wrong on margin and credit, the shares are worth roughly $13, against $752 million of tangible equity, about $14.43 a share. At $11.00 the price is about 5.9 times the $1.87 level (an earnings yield of 17%) and about three-quarters of tangible book, with a small dividend in the meantime. The price assumes a holding period of three to four years; from that price the base case implies about 20% a year, and the bear case implies a modest gain.