American Financial Group arrived at its second quarter deep into a bet it has run for over a century: underwriting specialized commercial property-casualty risk selectively enough that the insurance itself, not the investment cycle, does the earning. The company is a Cincinnati-based holding that runs Great American Insurance Group across 36 specialty niches, and it is run by the Lindner family - Co-CEOs Carl and Craig Lindner - who have thrived for years by underwriting to a disciplined return bar, returning excess capital, and letting the market take the rest. This quarter was one of its strongest on record: pretax P&C operating income of $350 million, a record for a second quarter, up 28% from a year ago, on top of a combined ratio that improved 1.6 points to 91.5%. Net earnings rose to $2.99 per share from $2.07, and core net operating earnings to $2.82 from $2.14 - with annualized core operating return on equity of 19.2%, against 15.5% a year earlier.
The quarter deserves that billing, but two of its pillars are the kind that do not repeat at the same size. The first is reserve development: AFG released $55 million of prior-year reserves, worth 3.3 points on the combined ratio, versus 0.7 points a year ago - the reserve book gave back more than twice as much this quarter, and that is a favor the company cannot count on at the same rate forever. The second is the alternative-investment portfolio: P&C net investment income rose 23% because alternative investments returned 7.1% annualized, versus 1.2% a year earlier - a six-fold swing that added $42 million of the quarter's investment gain and that management itself frames as variable and reported on a lag. Strip both out and the quarter is still good - underwriting profit rose 26%, the core ex-alternative investment yield was flat, and the 91.5% combined ratio is a genuinely strong number - but it is a strong, not a spectacular, quarter.
What the next chapters fall on is concrete. AFG has agreed to sell Charleston Harbor Resort & Marina, a transaction it expects to close in the third quarter for a pretax core operating gain of roughly $125 million - a gain management says was not contemplated in its original business plan assumptions. The thesis, then, is not the record quarter so much as what comes after it: whether the underwriting engine sustains a low-90s combined ratio as reserve releases normalize, whether alternative-investment income holds anywhere near the recent pace, and whether the $125 million gain lands as guided.