Trade Credit Rates Fall 9.2% as Insolvencies Extend a Fifth Straight Increase

Trade credit premium rates fell 9.2% year to date in 2026, the softest pricing Marsh has recorded in six years of tracking the line through its MiCredit platform (Marsh, May 2026), while Allianz Trade counts a fifth consecutive year of rising global insolvencies, running 24% above the pre-pandemic average (Allianz Trade, April 2026). For a book priced this cheap, that gap has to close somewhere.

Two Curves Moving in Opposite Directions

WTW's Insurance Marketplace Realities 2026 Spring Update frames trade credit as a line sitting at historic lows, projecting full-year rate movement of negative 5% to flat even after the 9.2% year-to-date decline already booked (WTW, May 2026). Marsh's MiCredit data puts a number on what that pricing is buying: coverage value per premium dollar reached $699 in early 2026, the highest ratio in six years of monitoring, while insurer risk acceptance held at 73.59%, down barely a point year over year. Carriers are still writing nearly every account that asks for a line.

Set against that is a default cycle that has not paused since 2022. Allianz Trade's April update puts global business insolvencies up roughly 5% in 2026, on top of a 6% rise in 2025, a fifth consecutive annual increase that leaves failures about 24% above the pre-pandemic average. Atradius, working from a different model, puts the global figure lower at 3% but converges with Allianz Trade on the number that matters most for a US book: insolvencies there are forecast up 8% in 2026, after 785 US companies filed for bankruptcy in 2025, the highest count since 2010 (Atradius, April 2026). Two forecasters disagree on the global magnitude and agree on the US direction, which is the segment that drives most domestic trade credit exposure.

The figure that should worry a pricing desk more than either insolvency forecast is the one measuring what is already sitting in the book. Marsh's weighted average risk grade across the portfolio it places moved from 4.45 to 5.20 on a 10-point scale between 2023 and February 2026, a full-point deterioration in buyer credit quality that ran alongside the rate decline rather than against it. On-leveled loss ratios computed off a book whose average risk grade shifted that much in two years will understate the trend embedded in exposure that is on risk today, not because the indication math is wrong but because the mix underneath it kept moving after the experience period closed.

Metric	Figure	Source
Trade credit rates, YTD 2026	-9.2%	Marsh, May 2026
Full-year 2026 rate forecast	-5% to flat	WTW, May 2026
Global insolvencies, 2026F	+5% (Allianz Trade) / +3% (Atradius)	Allianz Trade, Atradius, April 2026
US insolvencies, 2026F	+8%	Atradius, April 2026
Weighted average risk grade	4.45 (2023) to 5.20 (Feb 2026)	Marsh, May 2026
Insurer risk acceptance rate	73.59%	Marsh, May 2026
Coverage value per premium dollar	$699, six-year high	Marsh, May 2026
Top-3 carrier global premium share	~65% (2024)	Industry data, 2024

The Failures Are Showing Up Smaller

Pablo Brando, president of trade credit in the US at Intact Insurance Specialty Solutions, put the shift in blunt terms: "It's not just one event anymore, it's overlapping events" (Insurance Business, June 2026). Large corporate failures, he said, remain contained; the pressure has moved into companies with payment delays and insolvencies concentrated in the middle market, the tier of buyers with thinner public disclosure and less analyst or ratings coverage than the accounts a typical trade credit portfolio was underwritten and stress-tested against. Brando characterized the pattern as frequency over severity: "We haven't seen any major catastrophic (claims), so it's been more frequency and less severity."

That is the part of the standard credit-limit framework that breaks first. Buyer risk assessment for large, rated corporates draws on audited financials, syndicated bank data, and public credit ratings that update on a predictable cycle. Middle-market buyers file audited statements less often, if at all, and the payment-behavior signals Brando calls "the first sign of trouble" surface later and less reliably than they do for a rated multinational. A portfolio whose experience period was built on a large-corporate-heavy mix carries loss development factors calibrated to a segment where distress is visible months before default; the middle-market share of that same book behaves differently, with weaker early warning and less time between the first signal and the claim.

Marsh's frequency data adds a wrinkle that reads as good news until it is examined: claim counts on the book it tracks fell from 227 in 2023 to 136 in 2025, even as insolvencies climbed broadly, while loss severity on those fewer claims rose 9.4% year over year. Fewer, larger claims against a rising insolvency count is not what a stable book looks like.

Tariffs and the Correlation Problem in Buyer Limits

Brando named the second driver directly: "The challenge today is the constant changes. Tariffs come on, tariffs come off, then new tariffs appear." Atradius's October 2025 outlook measured the mechanism behind that complaint: the effective US tariff rate on all imports had climbed above 18%, the highest level since the Smoot-Hawley Tariff Act of 1930 and up from under 2% in 2024 (Atradius, October 2025). The Supreme Court struck down the bulk of those tariffs in February 2026, ruling the administration's use of emergency economic powers unlawful, and the Yale Budget Lab's post-ruling tracking initially put the effective rate closer to 9% to 14% as replacement authorities phased in (Yale Budget Lab, February 2026). Atradius's April 2026 update did not walk the number back: it said the administration had pivoted to alternative legal authorities and a uniform import surcharge, and that the new effective tariff would land close to its previous level. The legal vehicle changed; the trade exposure did not.

For credit-limit setting, that distinction matters more than the headline percentage. A tariff shock does not select one buyer at a time. It compresses margins simultaneously across every importer, manufacturer, and distributor exposed to the same product category or supply chain, which is exactly the correlation that standard buyer-by-buyer limit assessment is built to ignore. The framework prices each credit line as though defaults across the book are close to independent, an assumption AM Best flagged in a different corner of the industry when it warned that affiliated-loan concentrations at PE-backed annuity writers could produce correlated rather than diversified losses. A trade credit portfolio concentrated in tariff-exposed sectors carries the same structural risk: a shock that loads onto many buyer relationships in the same quarter, not the isolated single-name default the aggregate excess-of-loss retention was sized against.

The Limit Interventions Behind the Frequency Drop

The reconciliation between falling claim counts and rising insolvencies sits in what insurers did to their books, not in the accounts themselves. One major insurer tracked in Marsh's data reported loss ratio deterioration exceeding 5 percentage points and responded with 50% more proactive limit interventions than the prior period. Cutting or withdrawing a credit limit on a deteriorating buyer keeps that buyer's eventual default off the claims register even though the underlying credit risk was real and had already been recognized by underwriting.

Reported frequency trends built on gross claim counts, without an exposure adjustment for limit reductions made mid-period, will read as improving even when the book is being actively de-risked underneath a flat or falling price. A pricing actuary relying on raw claims experience to justify holding or cutting rate at the next renewal is working from a trend that underwriting has already edited. The honest question is not whether frequency fell; it is how much exposure was pulled to make it fall, and whether that exposure reduction shows up anywhere in the rate indication.

Three Carriers, One Concentrated Renewal

Allianz Trade, Atradius, and Coface wrote roughly 65% of global trade credit premium between them in 2024, with Allianz Trade alone near a third of the market. The global trade credit insurance market is valued at approximately $19.87 billion in 2026 and is projected to grow at a 9.26% compound annual rate to $30.94 billion by 2031 (Mordor Intelligence, 2026). Concentration at this level means a correlated middle-market default wave does not diversify across dozens of balance sheets; it lands on the same three carriers and flows directly into the excess-of-loss reinsurance treaties that sit behind them. A tariff-driven cluster of defaults concentrated in a handful of import-exposed sectors would show up in the retrocession layer on roughly the same timeline it shows up in the direct book, a correlation a ceding actuary should weigh when the next treaty renewal is priced rather than assuming trade credit retro behaves like a diversified, low-correlation line.

Renewal Questions for the Pricing Desk

Three items belong on the agenda before the next renewal closes. First, ask the broker for buyer-level loss and limit data segmented by revenue band rather than the blended book average: a portfolio-level loss ratio that looks flat can still be masking a middle-market segment running well above the large-corporate segment it is averaged against. Second, ask whether the reported claim count reflects exposure net of limit interventions or gross counts against a book that was actively cut; the 50% jump in proactive interventions Marsh documented for one carrier suggests the answer varies by insurer, and the difference changes what the frequency trend is actually saying. Third, watch WTW's next Insurance Marketplace Realities update, typically published each October, for whether the -5%-to-flat full-year forecast held or whether the second half produced the firming that a fifth consecutive year of rising defaults would normally force. Allianz Trade and Atradius both publish their next quarterly insolvency outlooks in October as well, timed to land before most trade credit treaties renew on a calendar-year basis. A market pricing at a six-year low on coverage per premium dollar has very little room left to absorb a surprise in either report.