US P&C’s repricing cycle needs more than 4.7% yields: reserve risk, business‑interruption gaps, and AI-era liability determine whether profits convert

Most investors track the US property & casualty (P&C) market through premium growth and “rate increase” headlines. But the cycle only turns into sustained earnings power when underwriting gains persist after (1) catastrophe volatility, (2) loss-reserve estimation error, and (3) policy/coverage expansion that raises the tail of future claims. The key question for 2026 is whether the market’s repricing finally outruns uncertainty—especially in business-interruption (BI) and liability lines that keep getting more complex.

Catastrophe losses are the “speed bump” that can make a softening or strengthening market look contradictory quarter to quarter. Even when rates are rising, weather-driven events can spike losses and delay the visible conversion of improved pricing into underwriting profitability.

Travelers’ disclosure shows the mechanism: when cat pressure drops sharply, underwriting profitability and earnings become easier to “see” even if the pricing environment is changing more gradually. Investors should therefore treat cat-loss timing as a first-order variable when judging whether the rate cycle is actually working.

Insurance earnings depend on both (a) what insurers expect to pay and (b) how close those estimates are to reality. Reserve adequacy is not a one-time number; it is a process with quantified sensitivities. When those sensitivities are large, even “good” pricing can fail to translate into shareholder returns if claim emergence differs from assumptions.

In AIG’s 2025 annual filing, the company lays out how changes in loss cost trends, tail factors, and development speed can move loss reserves by hundreds of millions of dollars. That matters for the pricing cycle thesis because repricing improves future premiums, but reserve-model error affects both the current accident-year P&L and the prior-year development that investors use to judge reserve quality.

Business interruption is not just another underwriting category—it is a contract-and-claims mechanism where the timing, causation, and documentation of loss can create lags and disputes. Two things can happen in a BI gap dynamic: (1) exposures expand faster than insurers’ ability to price them accurately, and (2) claim emergence reflects real-world events in ways that are hard to “translate” into premium rate adequacy.

From a structural perspective, BI losses can also become correlated with other coverage types (property and liability) and with broader operational risks. Investors should therefore expect BI “repricing” to be uneven: it can look successful on average premium measures but still fail in stressed events or specific customer segments.

Chubb’s filing, for example, references business interruption within certain product lines and also highlights how coverage can be impacted by evolving claim and coverage issues. The takeaway is not that every insurer is underpricing BI today; it is that BI is an area where documentation and claims mechanics can cause meaningful reserve/claims surprises.

The AI era changes P&C risk in ways that don’t behave like traditional property frequency/severity cycles. Liability exposures can broaden via cyber, privacy/data-use disputes, automated decision-making errors, and operational downtime tied to AI-dependent systems. This matters because “repricing” that focuses on near-term frequency may not be sufficient for liability growth with longer emergence lags.

Chubb’s filing explicitly notes evolving regulatory and cybersecurity obligations (including AI-related regulator guidance) and frames cyber risk as potentially catastrophic. That is the linkage investors should model: regulatory and governance pressure can raise both the likelihood and the cost of claims, while insurer underwriting must incorporate them into pricing and reserves.

A common misconception is that the “rate cycle” is mostly about improving underwriting margins. In reality, US P&C insurers also earn meaningful investment income; higher long yields can buoy net investment income and help cover underwriting strain. The risk for shareholders is that investment income can mask reserve or coverage issues until claims emerge.

AIG’s annual filing confirms the magnitude of net investment income as a recurring earnings component. But AIG’s same filing also documents how large reserve sensitivities can be. Put together, this creates the investor’s core framework: investment income can support current profitability, while reserve risk determines whether profits persist when the claims side surprises.