$10.91
$58.20
bear $10.91
base $39.06
bull $58.20
margin of safety $20.00
market price* $56.44
United Fire Group, based in Cedar Rapids, Iowa, writes commercial property and casualty insurance: business packages covering property, general liability, commercial auto and workers' compensation, plus contract and commercial surety bonds. It sells through about 850 independent agencies and 160 surety agencies, with smaller specialty and surplus-lines, managing-general-agent programs, assumed reinsurance and participation in Lloyd's of London syndicates. It finished leaving direct personal lines in 2025 and sold its life insurance company in 2018; every figure below is for the property-casualty business alone. How we measure earnings. An insurer's operating cash flow swings with the growth of loss reserves and unearned premiums, which is money held for policyholders, not profit. We start from operating cash flow and strip out those reserve and receivable movements, the trading-securities purchases and sales that run through operating cash, share-based pay and the cost of keeping offices and systems up to date. What remains closely tracks net income excluding investment gains and losses. On that basis owner earnings per share for 2016-2025 were $2.05, $1.25, $0.57, -$1.52, -$3.37, $1.76, $1.63, -$1.24, $2.52 and $4.73. Three losing years in a decade define this business: 2019 (weak underwriting that $54 million of investment gains only partly masked), 2020 (a $113 million net loss, including a $15 million goodwill write-off) and 2023 ($62 million of adverse development on prior-year claims). Normal earning power. Insurance pricing runs in cycles, so we use the whole decade. The median year kept about 3.8% of earned premium as owner earnings; on 2025's $1.29 billion of earned premium that is about $49 million, or $1.85 a share. Investment income adds to this. The bond portfolio now earns about 7.5% of premiums against a decade median of about 5.7%, and with an average duration of 4.3 years that higher yield is locked in for roughly four more years, worth about $0.73 a share a year. After that it depends on interest rates, which we cannot predict, so we fade it out. The 2025 result of $4.73 was flattered: catastrophe losses cost 3.2 points of premium, which the company says is below its ten-year average (2023's 6.2 and 2024's 5.4 points were also below average), and $5.2 million of older reserves were released. Normal catastrophe losses alone would have brought 2025 to roughly $3.50. Owner earnings grew 88% in 2025; the base case starts the coming year at $3.00, below that pace, because policies already written at 2025-26 rates will still earn through, then lets earnings glide back toward the cycle average as pricing softens. Base case, about $39 a share: $3.00 next year, about $2.60-2.70 in years two to four (cycle-average underwriting plus the locked-in investment income), easing to about $2.20 by year ten. The underwriting level grows only 2% a year, below the 3.6% a year at which premiums grew with insured values from 2016 to 2025. After year ten the business keeps earning that level without further growth: the independent-agency franchise has operated since 1946 and its renewal book is what persists. No credit is taken for new specialty lines, share gains or margin targets, because none of these is yet proven. Bear case, about $11 a share: a repeat of the company's own 2019-2020 losses in years two and three (-$1.50 and -$3.40 a share), a 2023-style reserve charge in year seven (-$1.25), and ordinary years of only $1.20 as commercial liability claims keep running ahead of reserves (adverse development in commercial other liability was disclosed in both 2024 and 2025, driven by litigation and social inflation) and the market softens. It assumes about $50 million of new borrowing after the losses, as happened in December 2020, costing about $0.11 a share a year in interest. The two bad years would use up about $5 a share of the roughly $38 of shareholders' equity per share. That is painful but survivable without issuing new shares, as in 2020, when equity fell about 9%. Bull case, about $58 a share: today's earning power with normal catastrophes, about $3.40, holds for three years, supported by commercial-auto reserves that have since released $34.5 million (2024) and $22.3 million (2025) and by the finished exit from personal lines. It then settles at $2.60 rather than $1.85, grows 3% a year, and keeps the investment-income gain for six years. Debt. United Fire owes about $146 million: $50 million of surplus notes issued in December 2020 and $100 million of senior notes ($70 million Series A due 2039, plus $30 million added in July 2025), costing about $11 million a year in interest. For an insurer, policyholder reserves are the raw material of the business and interest is already deducted from owner earnings, so we value earnings after interest and do not subtract the debt a second time. Leverage is charged in the bear case. In each case the business value per share shown is net of what the company owes. Debt is about 15% of equity and long-dated, so no maturity or covenant drives the outcome. What the shares appear to be worth: about $11 in the bear case, $39 in the base case and $58 in the bull case. Shareholders' equity was about $38 a share at June 30, 2026, backed mostly by a bond portfolio. The base case sits close to that book value, reflecting a business that has historically earned single-digit returns on its capital: book value per share was $36.88 at the end of 2025, slightly below the $39.06 at the end of 2017, after regular dividends and a large special dividend paid from the life-company sale. Margin-of-safety price. How sure are we of the earnings level? Not very. The decade ranges from -$3.37 to $4.73 a share, and the median sits near $1.85. The worst plausible event is a large reserve shortfall in commercial liability arriving together with a heavy catastrophe year. The structure would survive it: modest, long-dated debt, reinsurance, policies spread across all 50 states with no customer concentration, and assets that are mostly investment-grade bonds. But it could erase two or three years of earnings. If we are wrong about the cycle-average margin, the shares are worth nearer $11. If the investment-income gain fades sooner, take off $2-3. If the recent improvement proves durable, they are worth nearer $58. The price assumes a three-to-five-year holding period, and the company pays $0.80 a share a year in dividends ($0.20 a quarter in 2026, up from $0.16). A price of $20 is about half of the shareholders' equity per share and under eleven times cycle-average underwriting earnings before the investment-income gain, with a 4% dividend. From that price the base case implies 20% a year if results simply return to their cycle average, and equity would remain behind the shares if a 2019-2020 episode repeats. Above that price, the wide gap to the bear case would need the recent good years to last. The margin-of-safety price is $20. Our previous figures were prepared on a different basis and are not comparable.