Insurance

The science of estimating event losses

With the North Atlantic Hurricane season now underway and insurers getting primed for what might lie ahead, discover the approaches we take at Moody’s in response to live catastrophic events and how we look to support the insurance market. 

For our insurer carrier clients, when an event happens, our work looks to help facilitate the understanding of the following:

  • The effective allocation of claims adjusters to the most impacted areas and thereby managing loss adjustment expenses.
  • The setting of underwriting and policy binding moratoriums to avoid taking on more risk in the impacted areas.
  • The amount of earnings or capital at risk, to inform important conversations with reinsurance partners, executives, board members, and shareholders.

We know that any industry loss estimate has real-world ramifications. Our diverse client base, from (re)insurers to insurance-linked securities (ILS) providers, financial services, lenders, etc., use our data, modeling, and loss estimates to inform important business decisions, including:

  • Impact on the stock price of publicly traded carriers and reinsurers.
  • State of industry-loss warranties (ILW) contracts or ‘live cat’ trading.
  • ILS market and its impact on the capital that supports those contracts. 

We face the challenge of balancing the robustness and reliability of the data, models, and estimates with the desire to receive this information quickly. So, what do we consider in the loss estimate process? Let’s take the example of a storm in the North Atlantic.

 

Event response

As a storm officially gains tropical cyclone status, our Moody's RMS Event Response team goes to work, providing vital regular updates and data to clients pre-, during, and post-landfall. From regular event reports published on the Support Center, distributing shapefiles denoting the impacted areas to aid with understanding accumulations, through to issuing similar stochastic events (SSEs) from our Moody’s RMS North Atlantic Hurricane Version 25 model and U.S. Flood Version 1.3 model, all this helps provide a first look at potential loss severity. 

And every six hours, from the formation of the tropical cyclone until it dissipates, our HWind team issues live footprints to track the wind field and provide a ‘best forecast’ track from an ensemble of published tracks to help show where the forecast consensus lies.

With much of the Event Response data and HWind footprints available directly on Moody’s Intelligent Risk Platform, our clients can use their preferred applications, such as Risk Modeler or ExposureIQ, to seamlessly assess a storm’s potential impact on their portfolio.

 

Initial loss uncertainty

For hurricanes, pre-landfall and the hours/days immediately post-landfall can represent a period of forecast uncertainty, as it is only once an event has reached the peak of its hazard impacts that one can attempt to represent or reconstruct the hazard and begin to refine the loss impacts.

Pre-event loss forecasts will include hazard uncertainty (i.e., where the hurricane will make landfall, at what speed, tide level, etc.). We find this is best represented by using multiple events with differing hazard characteristics, including landfall location and angle, speed, and storm category.

At Moody’s, we offer several event options to represent variations in wind, storm surge, and inland flooding impacts, which we did last year for Hurricanes Helene and Milton using SSEs and HWind footprints.

Damage uncertainty (i.e., uncertainty in the amount of damage to a property) comes into play during and after the event has concluded. We may have a better understanding of the event’s hazard characteristics post-landfall, but even then, uncertainty persists in the impacts of the hazard on the built environment. 

To help capture this damage uncertainty, we sample the damage curve directly in our high-definition (HD) models, rather than relying on an arbitrary mean damage ratio. Users can directly quantify this uncertainty through sampling and how that propagates throughout the financial perspectives to understand the impact on lines of business, geographies, or overall company loss estimates.

 

Generating an industry loss estimate

We will generally wait until there is more certainty around the final hazard before attempting to represent the industry loss outcomes for an event. For example, follow-on events such as heavy rainfall in areas impacted by wildfires can trigger mudslides and floods, causing further damage and adding to the event losses.

Once an event has passed, we offer detailed event reconstructions within 1-2 weeks following an event, which contemplates the hazard features as best known at the time, to enable a more precise loss estimate. From satellite imagery, field visits from our experts, weather station reports, and a mix of public and proprietary sources, reconstructions inform the Moody’s RMS Industry Loss Estimate (ILE), which we publish to the press and our client base.

It’s worth explaining why it can take two weeks to issue an ILE, and why we rarely issue an interim or a simpler ILE, before an official estimate. This is because we take great care to be as accurate as possible, and produce a number we believe won’t greatly change, because we know how important these business decisions are and how much the insurance market relies on the stability of the Moody’s view of real-time risk.

We provide a range of loss outcomes, a lower and upper bound, but we also know there is a desire for a simple ‘best estimate’ loss figure to help with decision-making.  A point estimate is useful when performing a broader analysis of business impacts, such as claims personnel deployment, allocation of share of loss to business units, and impacts across treaties or policyholders. 

We always understand the need to issue event loss information as quickly as possible, but there is much to consider before we can establish an official industry loss estimate.

There is also much we can do to help each client understand their unique situation after an event, and we would be delighted to show how we can help during the current 2025 season.

Find out more about Event Response here.

Author: Julie Serakos, Managing Director - Product Management, Moody's

LEARN MORE

Moody's insurance solutions

Our differentiated solutions bring together technology, data and analytics and insights, helping insurers, reinsurers, and brokers address their most complex challenges and make better decisions with confidence – therefore helping to close the insurance gap and drive performance.