Applied against the trailing ROE of 16.2%, a 1.60x P/B embeds an implied capitalization rate near 10%, which looks undemanding versus specialty commercial peers earning higher and higher multiples, but appears reasonable if the market is capitalizing the Title segment as a low-return housing cyclical rather than the durable trucking franchise. The gap between trailing and forward multiples is the more telling signal: the market is not paying up for growth, it is discounting mean reversion in earnings that the 2026Q3 combined ratio deterioration already appears to validate. One item that warrants scrutiny in the per-share valuation chain is that the trailing book value series implied by the ratio table has slipped from $25.79 in 2025Q3 to $23.98 in 2026Q3 even as total equity rose toward $5.9B, a divergence that may indicate share issuance or a data inconsistency rather than genuine book erosion.

The pattern across the ten-quarter window is instructive: quarters printing combined ratios in the low-to-mid 80s produced ROE near 5.4%, while quarters in the 93-94% range (2024Q2, 2024Q4) collapsed ROE to 1.5-1.7%, implying that roughly every ten points of combined ratio swing translates to about four points of quarterly ROE. That sensitivity is now working against the company, and the simultaneous rise in both loss and expense components suggests the deterioration is not purely a claims event but may reflect deteriorating risk-selection economics or unrecovered cost inflation alongside the loss-cost pressure flagged elsewhere in the analysis. The 2025Q4 datapoint, with an 88.0% loss ratio against a 0.7% expense ratio, appears internally inconsistent as a clean split and should be treated as a data anomaly or a large one-off charge whose reclassification across ratio lines distorts any simple ten-quarter average.

That decomposition matters because it means the majority of Old Republic's return on equity is plausibly sourced from investment income earned on float and invested assets rather than from underwriting profit, which is consistent with the sharp swings in interest coverage observed between 6.10 and 24.36 across the period. The 16.2% trailing ROE therefore appears less a measure of underwriting franchise strength than of a levered investment portfolio, and the falling coverage reading of 6.78 in 2026Q3 alongside weaker underwriting suggests both engines may be decelerating at once. Investors should recognize that a rising reinvestment yield could offset part of this, but the recent pattern of favorable development thinning means the cushion that historically amplified underwriting returns is, on the available evidence, no longer contributing as it did through 2025.

On underwriting leverage alone, the company appears materially under-levered relative to a 2.0-3.0x pricing-of-risk framework common to commercial-lines peers, which implies capacity to absorb the elevated claims outflows seen in 2026Q3 without capital strain. However, the elevated debt-to-equity reading of 0.30 in that same quarter, up from 0.27, and the reported $8.6B single-quarter swing in total liabilities between 2026Q2 and 2026Q3 warrant reconciliation before the low leverage figure is relied upon in a solvency judgment. The more relevant tension is not leverage but capital recycling: funding a $693.8M special distribution from investment sales against $281.4M of operating cash suggests the dividend is being sustained by balance-sheet liquidation rather than current earnings power, a pattern that becomes harder to maintain if underwriting margins keep compressing.

The relative positioning is coherent: Old Republic's 1.60x P/B sits between CNA at 1.13x and Markel at 1.24x on the low side and the specialty compounders at 2.3-3.0x on the high side, which is roughly where a business earning mid-teens ROE with a cyclical Title overlay should trade. The yield comparison is misleading as a valuation signal, since the quarterly series shows regular distributions near 0.7-0.9% punctuated by 5.8-7.0% special payments, so the recurring yield an investor actually earns is far closer to the peer group and the apparent yield arbitrage should not be extrapolated forward. Where Old Republic appears genuinely cheap relative to peers is on PEG at 0.69 versus W.R. Berkley's 0.56 and Markel's 0.44, though that comparison presumes a growth durability that the decelerating top line and rising loss ratios make difficult to underwrite with confidence.