How PCS integrates Contributor, Verisk, and Public Data to define industry losses: White

Property Claim Services (PCS), the unit of Verisk that is a provider of industry loss estimates and loss data to the reinsurance and ILS industry, has provided us a detailed overview of the processes the organisation uses to designate and estimate insurance industry losses for perils such as U.S. tropical cyclones, severe thunderstorms, and winter storms.

As the use of PCS data has evolved from internal reserving to supporting risk-transfer and catastrophe model validation, the need for transparency regarding its underlying methodology has grown.

PCS data use is widespread in the insurance-linked securities (ILS) market, with its industry loss reporting utilised within the risk transfer triggers of instruments from catastrophe bonds to industry-loss warranty (ILW) arrangements.

In a recent interview with Artemis, Harry White, Head of Commercial Strategy, PCS, Verisk explained how the firm integrates contributor, Verisk, and public data to ensure that final loss bulletins accurately reflect the total market impact.

“The insurance industry that existed when PCS was founded in 1949 is unrecognisable from the industry we know today. It has transformed from regionally focused carriers with paper-based operations, to data-driven nationwide and global firms. Technology and analytics have reshaped underwriting, pricing, distribution, and consumer expectations, while competition has intensified, capital markets have become integrated, and risk management tools have grown in sophistication,” White explained.

“In this period, the use-cases for PCS data have seen an equal transformation. Whereas estimates were initially provided to help insurance companies verify event severity, set reserves, and ensure consistent event classifications, recent years have seen PCS data increasingly used as the basis for risk-transfer and for the validation and calibration of catastrophe models.”

White elucidated that as these new use cases expand, the need for stakeholders to grasp the mechanisms behind PCS’ loss estimates has similarly risen.

He further indicated that the amount of insured limit traded in the industry-loss warranty (ILW) market, which relies on PCS estimates, ultimately heightens the emphasis on PCS’ processes, as both buyers and sellers aim to comprehend the methodology.

The PCS at Verisk team have produced a methodology document that outlines the key processes involved in the firm’s designation and estimation process for United States Tropical Cyclones, Severe Thunderstorms, and Winter storms.

“Since inception, PCS’ process has been built around data collected from the insurance industry. Today, there are three key sources that provide this information: Contributor Data, Verisk Data, and Public Data, White explained.

Beginning with Contributor Data, this is gathered for individual companies from the approximately 150 Data Contributors PCS has established relationships with across the United States. This includes data from primary carriers, residual markets, captives, as well as excess and surplus lines writers,” White added.

Moving on to Verisk Data, White affirms that this is accessed through arrangements which provide the PCS team with detailed claims information from Verisk’s Property Estimating Solutions, along with claims data collected by Verisk Statistical Services.

And lastly, White explained that Public Data is sourced from communications issued by an insurer or regulator that provides sufficient granularity for use in the estimate process.

“Each of these sources is used not only to produce loss estimates, but for companies where data is available through two or more of these sources, they can be used to validate the information provided,” White told Artemis.

Moving towards event designation, this process follows a process of monitoring weather across the United States to help identify and track individual systems, which White refers to as Event Tracking.

“As potentially damaging systems are identified, PCS begins to communicate with any Data Contributors who are known to have exposure in the affected region to gather their initial observations on claims activity,” he explained.

“If the feedback indicates that the event could result in a significant number of claims, then PCS will initiate a formal request for information from all Data Contributors. This data is then collated and compared against the three criteria used to designate PCS Catastrophes in the United States.”

Adding: “These are whether the loss is expected to be greater than $25,000,000, whether the event is deemed likely to impact the majority of the insurance market, and if the event will likely result in a significant number of claims for the insurance industry.

“If it is deemed that all three criteria will be met, the event is designated as a PCS Catastrophe, which we refer to as Event Designation.”

Regarding Loss Estimation, this process builds upon the designation phase and utilises the same Contributor Outreach and Data Collection techniques. However, once the event has been designated, White emphasised that it is also feasible to obtain supplementary information from both Verisk’s Data and Public Data sources.

“Once all data is collected, the PCS team will look to apply PCS’ scaling methodology to account for any claims that are incurred but not reported to better reflect the final ultimate loss number for each individual company,” he added.

Crucially, White emphasises that since the data gathered by PCS represents only a portion of the losses within the insurance sector, it is essential to modify the figures to accurately represent the whole industry, a process he terms as Market Share Adjustment.

“For the admitted market, this is achieved by leveraging AM Best data to establish the market share of the companies where PCS has been able to gather loss information and scaling appropriately for each state and line of business, while for the non-admitted market the PCS team is able to leverage historical information provided by Verisk Statistical services to estimate the proportion of the loss attributed to admitted vs non-admitted markets in recent events.”

“Once all data has scaled to represent the full market for each state and line of business, PCS will combine the losses to produce the PCS estimate for the event and release the bulletin to the market,” White concluded.

The full methodology document is now available to PCS subscribers. Subscribers can access it through the PCS Portal here.