X

Log in

Forgot password?

Don't have an account with us?

Register Now

When forecasts meet fickle weather

Hurricane season stays quiet (for now), Verisk warns of rising catastrophe losses, and ZestyAI bakes mitigation into storm scoring. Plus: new quake, flood and underwriting models, and a look back at Katrina and Plainfield’s F5 tornado.

Exponential Risk is InsTech’s monthly newsletter dedicated to modelling, climate driven risk and near term forecasting insurance news.

When forecasts meet fickle weather


The Brief

Is this the calm before the named storm? With just over a month of “official” hurricane season left, it’s been unusually quiet in the North Atlantic this year, despite forecasts from NOAA and others of above-average activity in frequency and severity of storms due to warmer sea surface temperature.

Various factors are contributing to low hurricane activity, including dry air from the Sahara Desert, a more stable atmosphere, a persistent high-pressure area over the north and central Atlantic and cooler waters of Africa. Once again proving that despite longer-term trends of larger catastrophe losses, short-term forecasting beyond about a week is still highly uncertain. Still a few weeks to go though…

 

Latest from InsTech

 

 

Industry Updates

Liberty Mutual Re and Safehub launch ShakeNet Parametric

The earthquake solution uses dense seismic sensor networks and GEM’s OpenQuake engine to deliver locally grounded triggers.

ZestyAI adds mitigation-aware storm scoring

ZestyAI has introduced functionality that allows insurers to adjust severe convective storm risk scores based on verified property improvements such as roof upgrades and structural repairs.

JBA Risk Management updates India Crop Model

Insurers covering India’s crop schemes now have access to a more detailed model with improved validation and broader coverage.

MIS and Eagleview form catastrophe intelligence alliance

The partnership combines McKenzie Intelligence Services’ intelligence analysis with Eagleview’s aerial imagery to deliver post-event insights for North American catastrophes.

Aon launches Event Analytics platform

The solution integrates real-time loss estimates, mapping tools and reports into a single interface to support re/insurers’ catastrophe response before, during and after events.

Nephila to launch catastrophe-focused Syndicate 2359

Nephila Capital has secured in-principle approval for a new Lloyd’s syndicate dedicated solely to property catastrophe reinsurance, targeting $150 million USD in premium from January 2026.

Tokio Marine launches TMGX underwriting platform

Tokio Marine GX has begun operations with a $500 million USD capacity and a suite of products spanning renewable energy, finance and emerging green technologies.

 

Industry Insights

Verisk projects rising global catastrophe losses

The 2025 Global Modeled Catastrophe Losses report forecasts annual insured losses above $152 billion USD, driven by exposure growth, urban expansion, event frequency and climate change.

20 years since Hurricane Katrina

EigenRisk reflects on the storm’s $125 billion USD losses, its lasting impact on New Orleans, and parallels with Hurricane Betsy 60 years ago.

UK braces for storm risks after record summer heat

Following the UK’s hottest summer on record, Previsico warns that warmer seas, saturated air and compacted soils could amplify storm and flood impacts in the months ahead.

Cotality highlights widening wildfire insurance gaps

The report links underinsurance, increasing rebuild costs and carrier withdrawals to stalled recovery and rising mortgage stress, while warning that wildfire-induced conflagrations expose vulnerabilities even in areas classed as low hazard.

35 years since the Plainfield F5 tornado

Moody’s reflects on the 1990 Illinois event, noting today’s higher exposure could drive losses above $2 billion USD.

JBA examines exposure-driven uncertainty in flood losses

The applied global sensitivity analysis to Queensland flood risk finds exposure and vulnerability assumptions drive more variability in modelled losses than hazard intensity or climate scenario in this case study.

Stay up to date with the latest from InsTech