Jamaica's Power Company Just Bought $106 Million in Storm Insurance. Here's What It Means for Your Light Bill.

Why a Power Company Buys Insurance Like a Government

Parametric insurance built its reputation in the Caribbean insuring governments. CCRIF SPC was created in 2007 specifically so that a member state's treasury would not have to wait months for a claims adjuster before rebuilding could begin, and Caribbean Insurance has covered how that same mechanism now reaches individual farmers and vendors through CCRIF's Livelihood Protection Policy. JPS's US$106.6 million plan extends the same logic one level down, to the physical grid that keeps the island's lights, refrigerators, and hospitals running. A utility's exposure is different from a government's treasury or a single farmer's income, but the underlying problem is identical: after a major hurricane, the difference between weeks and months of recovery often comes down to how quickly cash arrives, not how quickly the damage is assessed.

The Office of Utilities Regulation's approval describes the plan as a risk-mitigation framework built as a corollary to JPS's existing Electricity Disaster Fund, an internal reserve the utility already draws on after storms. Parametric cover does not replace that fund. It sits alongside it, adding capital that arrives automatically once a defined wind speed or rainfall threshold is crossed, rather than capital JPS has to apply for or borrow after the fact.

The Blackout That Made This Deal Happen

Hurricane Melissa made landfall in Jamaica on October 28, 2025, and the damage to the electrical grid was severe by any measure. Roughly 77% of JPS customers, more than half a million accounts, lost power. Field crews later recorded about 40,000 downed or damaged poles and roughly 1,400 miles of damaged line, a scale of destruction that turned restoration into a months-long campaign rather than a days-long repair job. JPS reported restoring power to 686,000 households and businesses in the weeks after the storm, but full recovery moved much more slowly than that headline number suggests: by mid-January 2026, roughly 51,000 customers were still in the dark, and by late January the JPS chief executive told Parliament's Public Administration and Appropriations Committee that 33,000 customers remained without power. The final several thousand accounts were not reconnected until May 2026, seven months after landfall.

That restoration effort was expensive in ways that went beyond materials and labour. In July 2026, IDB Invest lined up an US$80 million loan to support JPS's grid recovery, financing that arrived roughly nine months after the storm because the scale of the damage outran what the utility's own reserves and insurance at the time could cover quickly. The parametric plan approved that June was built to close exactly that gap the next time a major storm hits: instead of assembling financing after the damage is counted, JPS now has a contract that pays out automatically once the storm itself crosses an agreed threshold.

How the $106.6 Million Actually Works

The plan uses what insurers call a dual-trigger structure. The CCRIF layer, worth US$56.6 million, uses the same tropical cyclone parametric model that has paid Caribbean governments for nearly two decades, most recently a combined US$91.9 million to Jamaica's government following Melissa. The Descartes Underwriting layer, worth US$50 million, uses a separate model built on satellite imagery, radar, and AI-driven storm simulation. Running two independently modelled triggers side by side reduces basis risk, the danger that a storm causes real damage but somehow fails to register against a single insurer's specific formula. If either trigger is crossed, its layer pays out, regardless of what the other model concludes.

Both layers are genuinely parametric, meaning payment depends on a measured event, not a damage estimate. Neither CCRIF nor Descartes Underwriting sends someone to count downed poles before releasing funds. That is the entire point: cash that follows the storm within days is available to mobilise repair crews, order replacement poles and transformers, and begin restoration while a traditional insurance claim, or a loan application, would still be in its early stages.

The AI Models Deciding When You Get Power Back

Descartes Underwriting, the Paris-based insurtech writing half of JPS's coverage, builds its parametric triggers with a team of roughly 150 climatologists, data scientists, and engineers who combine decades of historical storm data with AI models trained on satellite imagery, radar feeds, and IoT sensor data from more than 80 technology partners. The company has also begun using generative AI, specifically diffusion models, to simulate thousands of realistic synthetic storm scenarios and estimate wind and rainfall risk at a specific location more precisely than historical records alone allow. CCRIF's own pricing infrastructure has moved in the same direction: its parametric risk pool grew 9% to US$1.57 billion in coverage limits ahead of the 2026 season, according to CEO Isaac Anthony, with new members including several regional electric utilities.

None of this is visible to a JPS customer whose power flickers back on a few days after a storm rather than a few months. But a model most Jamaicans will never see is now a direct input into how fast that happens. The Caribbean AI Risk Management Council has argued that AI models pricing infrastructure risk deserve the same regulatory scrutiny regional bodies are beginning to apply to underwriting and claims algorithms in personal lines, and a US$50 million layer sitting on an AI-built trigger is exactly the kind of decision that argument is about. StarApple AI, founded by Adrian Dunkley as the Caribbean's first artificial intelligence company, has made a related point for years: understanding how a model prices or triggers a payout is no longer a specialist concern reserved for reinsurance desks in Bermuda, it is now relevant to a utility regulator in Kingston and, indirectly, to every ratepayer whose bill eventually funds the premium.

A Quieter Forecast Did Not Change JPS's Mind

JPS finalised this coverage in early June 2026, when NOAA's original outlook called for 8 to 14 named storms and a 55% probability of a below-average season. Both forecasts have since been revised down. NOAA's August update raised the probability of a below-average season to 75% and narrowed its forecast to 7 to 13 named storms, 2 to 6 hurricanes, and 0 to 2 major hurricanes. Colorado State University's final August outlook is more conservative still, projecting just 9 named storms, 4 hurricanes, and 1 major hurricane, against a typical Atlantic season of 14, 7, and 3. Peak hurricane season nonetheless runs from August 20 through early October, with the statistical peak falling on September 10, meaning the most dangerous stretch of the 2026 season was still ahead as this article published.

JPS has not cancelled or scaled back its coverage since either downgrade, and that is the more instructive data point than either forecast on its own. A utility whose grid took seven months to fully recover from one storm is not pricing its risk management around a single season's probability. Caribbean Insurance made a similar point after the July forecast cuts: a below-average outlook changes the odds of a claim, not the argument for holding the cover in the first place.

What It Means for Your Light Bill

The honest answer is that JPS's premium and every Jamaican's electricity bill are already connected, just not in a way that shows up as a separate line item. The Electricity Disaster Fund and JPS's operating cash flow, the two sources funding the roughly US$6.6 million annual premium, are both built from revenue the Office of Utilities Regulation allows JPS to collect through customer tariffs. That was true before this parametric plan existed and remains true now; the plan changes how that money is deployed after a storm, not whether ratepayers are ultimately backing it.

What should change for a household or small business is the expectation of what happens after the next major hurricane. A payout that arrives within days of a qualifying storm, rather than a financing package assembled over months, is the difference between mobilising repair crews immediately and waiting for money to arrive before work can properly scale. That does not guarantee a fast restoration; grid damage, road access, and crew availability still matter enormously, as Melissa demonstrated. It does remove one of the bottlenecks that turned a severe storm into a seven-month recovery the last time.

1. Do not expect a rate cut. This coverage is a hedge against a much larger and more disruptive cost, not a discount JPS is likely to pass through in normal years.
2. Do expect the Office of Utilities Regulation to reference this plan in future tariff reviews. A funded disaster risk-financing framework is a factor regulators weigh when assessing a utility's overall cost structure and resilience.
3. Keep your own coverage current regardless. JPS's grid insurance covers JPS's infrastructure. It does nothing for a homeowner's roof, appliances, or business interruption, all of which still depend on the policyholder's own cover being adequate, a point Caribbean Insurance has covered in detail since Melissa exposed how many Jamaican homes were underinsured against their own rebuilding costs.

Beyond JPS: A Template for Caribbean Businesses

JPS is not the first utility to buy this kind of cover, and it will not be the last. CCRIF's electric utility portfolio already includes ANGLEC in Anguilla, DOMLEC in Dominica, GRENLEC in Grenada, LUCELEC in Saint Lucia, and NEVLEC in Nevis, alongside JPS in Jamaica. What makes the JPS plan notable is the second, independently modelled AI-priced layer stacked on top of its CCRIF cover, a structure other regional utilities and large infrastructure operators are now positioned to study and copy.

The lesson for a hotel group, a port operator, a telecoms provider, or a manufacturer with hurricane-exposed assets is not that every business needs a nine-figure parametric plan. It is that parametric cover has moved well past the sovereign and microinsurance products most people associate with CCRIF, and it is now a mainstream option for any business whose recovery speed depends on cash arriving fast. The right move is a conversation with a broker about whether a parametric layer would close a genuine gap in existing property cover, not a purchase made on the strength of a single utility's headline.

Frequently Asked Questions