Assured Guaranty Ltd. (NYSE: AGO) closed the second quarter of 2026 with all three of its headline per-share valuation metrics at record highs, a quiet but consequential milestone for the only meaningful survivor of the U.S. financial-guaranty franchise. As of June 30, 2026, shareholders' equity attributable to AGL reached $126.18 per share, adjusted operating shareholders' equity reached $129.94, and adjusted book value (ABV), which adds back the present value of future premium earnings net of expected losses, reached $189.72. The equity finished the quarter at $76.79, leaving the public market valuing ABV at roughly 0.40x, the deepest structural discount the stock has traded at since the post-financial-crisis low, even as the underlying book quietly compounded. The quarter was also the first full print with the new Assured Life Re annuity reinsurance segment, acquired January 21, 2026, contributing $2 million of adjusted operating income, $484 million of pension risk transfer reserves, and $256 million of multi-year-guaranteed-annuity account balances.
The Q2 2026 print, released August 6, 2026, told a story the GAAP headline partially obscured. GAAP net income attributable to AGL was $39 million, or $0.88 per diluted share, down sharply from $103 million, or $2.08 per diluted share, in the second quarter of 2025. The Q2 2025 number, however, was flattered by $79 million of foreign-exchange remeasurement gains on long-dated premium receivables and insurance reserves. Stripping out the noise and the typical non-cash fair-value adjustments, adjusted operating income was $55 million, or $1.23 per share, up from $50 million, or $1.01 per share, a year earlier. The Financial Guaranty segment did the heavy lifting, with adjusted operating income of $85 million versus $76 million in the prior-year quarter, driven by $14 million more in net earned premiums and credit derivative revenues and a $22 million drop in loss expense. The shadow on the print was $19 million of mark-to-market losses on the CLO equity tranche portfolio, a recurring source of quarterly noise the market has been slow to look through.
The single load-bearing risk is the trajectory of insured-portfolio loss development in the second half of the year. Net expected loss to be paid, the present value of future loss-and-loss-adjustment-expense outflows the company expects to fund, climbed to $192 million at June 30, 2026, up from $186 million at June 30, 2025, with $48 million of net economic loss development booked in the second quarter of 2026 alone, almost entirely attributable to Brightline Trains Florida LLC, a privately financed high-speed rail project whose senior revenue bonds the company insures. The total Below Investment Grade (BIG) net par outstanding finished at $8.50 billion, down from $8.76 billion at year-end 2025, but the composition shifted: BIG 2 (the middle bucket, denoting credits that have shown deterioration but are not yet at risk of near-term default) jumped to $4.94 billion from $3.81 billion as U.S. public finance BIG 2 names migrated out of BIG 1. Capital return remains the equity's structural offset. The company returned $62 million to shareholders in the quarter and has $121 million of remaining buyback authorization as of August 5, 2026.
The falsifiable clock for the thesis is the third-quarter 2026 print, expected in early November, which the market reads for two signals. First, whether the $48 million of second-quarter Brightline-driven loss development extends or reverses, since the high-speed rail project remains in workout and the ultimate resolution timing drives the loss trajectory. Second, whether the $45 million of second-quarter buyback at an average $80.68 per share was a one-quarter pause or a structural slowdown: the trailing-twelve-month repurchase rate of $130 million at an average $83.70 has already been more than absorbed by the rise in ABV, and any further discount of the same magnitude implies the board is choosing to let the discount persist. A re-rating to the historical 0.55-0.65x ABV range would add roughly $40-50 to the share price even with no further book compounding; the equity sits on the wrong side of that math today.