Japan’s non-life insurers recorded 40,291 accepted claims within three days of the magnitude 7.1 earthquake that struck Kumamoto Prefecture on July 28, 2026, according to data released by the General Insurance Association of Japan (GIAJ) on August 18. The figure is a starting point, not a conclusion – and for brokers with clients exposed to Japan’s manufacturing and technology supply chains, the more consequential questions sit well beyond the residential count.
The GIAJ’s accepted claims definition is broader than formal indemnity filings. It captures requests for damage investigations of dwellings and household goods, as well as consultations and inquiries about earthquake insurance coverage or policyholder contracts. It is a measure of policyholder engagement in the event’s immediate aftermath, not of confirmed losses or payouts. Of the 40,291 accepted claims recorded as of July 31, 2026, Kumamoto Prefecture accounted for 37,620 – approximately 93% of the national total. Fukuoka Prefecture logged 922, Kagoshima Prefecture 396, Miyazaki Prefecture 244, Saga Prefecture 222, Nagasaki Prefecture 191, and Oita Prefecture 124, with 572 recorded elsewhere. Insurance money is typically paid around one month after an application through a screening process. The GIAJ plans to announce detailed data regarding the earthquake in mid-September, including the amount of insurance payments, according to The Japan Times.
The accepted claims volume must be read against Japan’s residential earthquake coverage landscape. Japan Earthquake Reinsurance Co. (JER) data published in Toa Re’s Japan’s Insurance Market 2025 shows the household earthquake insurance penetration rate against total dwellings stood at 35.1%, meaning fewer than four in 10 households are covered. The earthquake insurance attachment rate – the share of fire insurance policies that include an earthquake rider – stood at 69.7%.
The gap between those two figures has direct implications for the accepted claims count. Under Japan’s earthquake insurance framework, fire insurance does not indemnify against damage caused by fire resulting from an earthquake, and earthquake insurance must be attached to a fire policy as a prerequisite. A significant portion of consultations captured in the accepted claims figure may ultimately reveal policyholders holding fire insurance without an earthquake rider – leaving them with no residential earthquake cover at all, and reducing the proportion of accepted claims that will convert to paid indemnity.
For context, the 2016 Kumamoto sequence produced approximately JPY 391 billion in dwelling insurance payments alone, according to GIAJ. Verisk’s Catastrophe and Risk Solutions group estimated insured losses from the 2026 event at between JPY 220 billion (approximately US$1.4 billion) and JPY 340 billion (approximately US$2.1 billion), covering ground shaking and liquefaction damage before recoveries under Japan’s earthquake insurance program, according to its August 2 release. Verisk cited earthquake insurance take-up rate uncertainty as an explicit factor in its loss range.
The government reinsures earthquake insurance liabilities that exceed the amount underwritten by private insurance companies, with reinsurance premiums collected and managed separately in the Special Account for Earthquake Reinsurance, according to the Ministry of Finance. The programme's total payout limit for a single earthquake event stands at 12 trillion yen, combining the liabilities of private insurers and the government's reinsurance backstop. Net retentions at private non-life carriers will therefore represent a fraction of the gross residential figure – a point relevant for brokers assessing counterparty exposure or placing Japan property cat programs.
The more immediate issue for brokers advising manufacturing, automotive, or technology clients is whether contingent business interruption (CBI) policies will respond to what actually transpired in Kumamoto – and the real-world restart timelines make that question urgent. Renesas said it would only be able to resume operations at one of its Kumamoto plants – which sustained fallen ceiling panels, wall cracks, and some water leaks – in phases from August 5, according to Reuters reporting. Sony said it would begin resuming operations from the following Tuesday, expecting to be back at pre-earthquake operating levels by the middle of the month. Tokyo Electron said it was moving to restart production at its Kumamoto area factory the following week.
Those restart timelines are precisely what CBI indemnity periods are designed to cover – but only if the policy trigger has been met. Tokyo Electron Kyushu confirmed no major damage to buildings or facilities at either its Koshi or Ozu plants in Kumamoto Prefecture but suspended operations at both the following day to conduct safety inspections. Standard CBI wordings require physical damage at a dependent property to trigger coverage. Where no confirmed structural damage is recorded, a policy requiring physical damage at a named supplier location may not respond – even where production delays extend for weeks. CBI coverage also requires policyholders to identify and, in many cases, schedule dependent locations. A client that holds CBI cover but has not named Tokyo Electron Kyushu as a dependent location has no cover at all, regardless of actual loss.
Mizuho Bank strategist Masayuki Nakajima noted that while manufacturers have strengthened supply chain resilience through supplier diversification since prior Kumamoto and COVID-19 events, the concentration of chip and chip-related plants in the Kumamoto region – described by Nakajima as “Japan’s Silicon Valley” – means the potential ripple effect remains significant. “Semiconductors are used in a wide range of products, so the possible ripple effect could be huge,” he said, according to Reuters.
Kentaro Tada, CEO of Howden Re Japan, framed the exposure around restart timelines rather than structural damage scope. “Reported damage to Kumamoto’s semiconductor infrastructure appears limited so far, though restart timelines will determine the scale of business interruption losses. It’s too early for formal estimates, but we’d expect insured losses to remain well below the 2016 Kumamoto earthquake and closer to the 2024 Noto earthquake, unless the aftershock sequence reveals more extensive damage,” he said.
For any broker placing technology sector, supply chain, or Japan property cat risks, three questions apply now: Is the relevant supplier named as a dependent location in the policy? Does the trigger wording require confirmed structural damage, or does operational suspension suffice? And does the indemnity period reflect realistic semiconductor and industrial recovery timelines – which, as the Renesas and Sony experiences show, can span weeks rather than days? The GIAJ’s mid-September payment release will answer what insurers paid. These questions determine whether clients can claim at all.
