Global Rates Are Falling 6% but US Casualty Is Rising: The Trap for Indian Exporters

The Marsh Global Insurance Market Index for Q2 2026, published on 23 July 2026, records the eighth consecutive quarter of falling commercial insurance rates. Global composite rates fell 6 per cent on average. Property led the decline at 12 per cent down. Casualty went the other way and rose 2 per cent, driven largely by claims severity and litigation pressure in the United States.

The regional split sharpens the point. The India, Middle East and Africa region recorded the largest composite decrease of any region at 16 per cent, while the US composite fell only 2 per cent. Property in the India, Middle East and Africa region was down 19 per cent, against 10 per cent in the prior quarter, so the softening accelerated rather than levelled off.

For a purely domestic Indian manufacturer, that is a straightforward buyer's market. For a manufacturer that ships into the United States and buys a products liability tower there, the two halves of the programme now sit on different sides of the same index. The domestic property placement is repricing down toward the regional average. The US-exposed liability tower is priced off American severity trends and does not participate in the Indian softening at all.

Rate movement follows the loss experience of the pool a policy sits in, not the geography of the buyer. A fire policy on a plant in Pune is rated against Indian property loss data, Indian catastrophe modelling and the treaty capacity supporting Indian fire business. A products liability tower answering to US courts is rated against US jury verdicts, US defence spend and the reinsurance appetite for US casualty.

Marsh's own framing of Q2 2026 puts the decline down to abundant capacity, strong insurer profitability and favourable reinsurance conditions, with underwriting focus remaining on catastrophe exposure, casualty severity and systemic risks. Capacity is abundant in the aggregate and selective in application. Insurers are deploying the surplus into short-tail property, where they can reprice annually and see the loss within the policy year, and withholding it from long-tail US casualty, where a claim notified in 2026 may not settle until 2033.

• Property losses develop fast, so a soft year can be corrected the following year.
• US casualty losses develop over five to ten years, so an underpriced year cannot be corrected before it has already been written into reserves.
• Litigation-funded claims lengthen that tail further and widen the range of possible outcomes on any single claim.

Take an auto component maker in Chennai with three plants, exports to tier-one suppliers in Michigan and Ohio, and a renewal calendar that puts domestic property and the US products tower within the same quarter. Four things move at once.

1. Domestic property premium falls sharply. With the India, Middle East and Africa property number down 19 per cent, the property line releases budget.
2. The primary US products liability layer holds or rises modestly. Primary casualty pricing has been the more stable part of the tower.
3. The excess layers move most. Severity in US casualty is an excess-layer phenomenon, so the layers that were cheapest per unit of limit are where the percentage increases concentrate.
4. Defence costs consume more of the primary layer. Where defence sits within limits, higher US defence spend erodes the limit before any indemnity is paid.

The net effect is a programme that looks cheaper on a single-line view and roughly flat or worse on total cost of risk. Finance teams that benchmark against the published index will ask why the liability line did not move with it. That question is best answered before the renewal, with the two pools separated in the presentation.

Our note on product liability underwriting for auto component exporters sets out how underwriters assess the US-facing exposure itself.

The practical answer is to stop running one renewal and start running two, each with its own timetable, market set and negotiating argument.

Separate the submissions. Domestic property and US-exposed liability should go to market as distinct submissions with distinct data packs. Bundling them invites an insurer to fund the liability increase out of the property saving, which turns a genuine 19 per cent property gain into an internal subsidy the buyer never sees.

Give each line its own lead time.

• Domestic property: a shorter cycle is workable, since capacity is plentiful and quotes converge quickly.
• US products liability: start 120 days out. Excess layers need to be rebuilt participant by participant, and a single non-renewing carrier in the middle of the tower can take weeks to replace.

Redeploy the property saving deliberately. A property line falling 19 per cent frees real money. The choices are to bank it, to buy back deductible on property, or to fund additional US liability limit while capacity is still available. The third option is the one most often missed, because the saving and the shortfall sit in different budget lines and different renewal conversations.

Three features of US casualty deserve specific attention at renewal because they behave differently from anything in an Indian domestic liability placement.

Excess layer pricing. In a stable market, each successive excess layer costs less per unit of limit than the one below it, on the reasoning that it is less likely to be reached. Rising severity attacks that reasoning directly. When large verdicts become more frequent, layers previously treated as remote start attracting claims, and the discount for height compresses. An Indian exporter renewing a tower should expect the largest percentage movements in the middle and upper layers, not the primary.

Defence costs. Whether defence sits inside or outside the limit is the single most consequential wording point in a US-facing liability insurance programme. Inside-the-limits defence in a severity environment means the limit is being spent on lawyers before any settlement. Our detailed treatment of defence costs in liability policies covers how the two structures behave through a claim.

Litigation funding. Third-party funding of US claims changes claimant behaviour: funded claimants can afford to refuse early settlements and run cases toward trial. For the insured, that means longer claim lifecycles, higher defence spend per claim and greater variance in outcomes. It also means a claim that would once have closed quietly within the primary layer can now reach excess layers.

The practical consequence for an Indian exporter is that limit adequacy, not price, is the question that matters most on the US portion of the programme.

Underwriters pricing US-facing casualty for an Indian manufacturer are working with less information than they would have on a domestic US risk. Filling that gap is the cheapest lever a buyer has.

• Jurisdictional revenue split. State-level US revenue, not a single country figure. Venue matters to severity, and a submission that cannot answer the question invites a conservative assumption.
• Product and end-use detail. Which components go into which end products, and whether any end use is safety critical. A component in a braking system prices differently from one in a trim assembly.
• Recall and warranty history. Five years of recall campaigns, warranty claims and field failures, with root cause and corrective action for each.
• Contractual position. Indemnity and hold-harmless terms agreed with US customers, and whether the Indian manufacturer has assumed liability beyond its own negligence. Assumed contractual liability is where uninsured exposure most often sits.
• Quality system evidence. Certification, batch traceability and the ability to identify affected production runs quickly.

Evidence on traceability does double duty. It supports pricing on the liability tower and it shortens the exposure window when a defect is found, because a manufacturer that can isolate three days of production is arguing about a far smaller claim than one that cannot.

Exporters selling beyond the United States face a related but different problem set, which we cover in the note on foreign jurisdiction claims for Indian exporters.

The Marsh index is a useful benchmark and a poor budget. Three habits keep it in its place.

First, benchmark line by line and region by region. The composite blends property, casualty, financial lines and cyber across regions with very different loss experience. A 16 per cent regional decrease tells an Indian buyer something real about domestic property and almost nothing about a US products tower.

Second, weight the benchmark by where the money sits. If 65 per cent of the programme spend is US-exposed liability, the relevant comparator is US casualty movement, not the India, Middle East and Africa composite.

Third, plan for the cycle to turn. Eight consecutive quarters of decline is a long run, and the drivers Marsh identifies (abundant capacity, strong profitability, favourable reinsurance terms) are cyclical. Soft years are when structural advantages get built: multi-year agreements where they are available, working relationships with lead markets, and documented risk-improvement evidence.

Cross-border programme design for Indian companies expanding overseas covers the structural side of that work in more detail.

For an Indian exporter with US exposure renewing between now and the end of the financial year:

1. Split the programme into two pools by loss geography, not by legal entity. Everything answering to US courts goes in one pool; everything answering to Indian courts goes in the other.
2. Model the total cost of risk across both, so a property saving and a liability increase are visible in the same number before the board sees either.
3. Start the US liability renewal 120 days out and map the current tower participant by participant, with each participant's renewal appetite tested early.
4. Test limit adequacy against severity, not history. A tower sized on the last decade of settlements is sized for a market that no longer exists.
5. Resolve the defence-costs question in writing before pricing is agreed, and price the inside-versus-outside difference explicitly.
6. Use the property saving as a decision, not a windfall. Bank it, buy deductible back, or buy limit, and record which one you chose and why.

Rates fell for the eighth straight quarter, and the one line that rose is the one an exporter cannot afford to under-buy.

The divergence in the Q2 2026 index is not a temporary anomaly in an otherwise soft market. It reflects a structural difference between short-tail property risk, where capacity is plentiful and repricing is quick, and long-tail US casualty, where severity compounds over years. Indian exporters should build their renewal process around that difference rather than around the headline.