Description of perils

Background

PERILS AG (“PERILS”) was incorporated on January 26, 2009, in Zurich, Switzerland, as a joint stock company. The company registration number is CH-020.3.033.447-4 and the principal office of the company is located in Zurich, Canton of Zurich, Switzerland.

Independence

PERILS operates as an independent company and is owned by ten shareholders, each with an equal interest in PERILS. The shareholders of PERILS are: Allianz, Axa, Generali, Groupama, Guy Carpenter, Insurance Australia Group, Munich Re, Partner Re, Swiss Re and Zurich Insurance.

The independent nature of PERILS is supported and maintained by its governance structure, and reflected in its organizational documents, its employment contracts with PERILS personnel, and PERILS’ license agreements with its subscribers (including its shareholders). PERILS’ independence is further supported and maintained by its purpose-built IT system, to which there is highly restricted access, that is used to carry out data processing, quality control and industry exposure and event loss estimation, all in in compliance with applicable antitrust and competition laws.

PERILS also fully owns Catastrophe Indices & Quantification (CatIQ) of Toronto, Ontario, Canada. More information on the background of CatIQ can be found on www.catiq.com.

Mission & Roles

PERILS’ principal aim is to contribute to the transparency and understanding of catastrophe insurance risk by granting access to catastrophe insurance data to all interested parties. The roles of PERILS are to prepare and to make available aggregated anonymous insurance data, to develop business-relevant catastrophe insurance industry exposure and loss estimates and to provide related services.

Our People (Human Resources)

PERILS has a staff of internationally experienced (re)insurance professionals whose educational backgrounds range from natural science to finance and business administration, with a combined industry experience that is greater than 100 years. Non-core activities such as office IT management, IT system development, accounting and human resources services are outsourced to third parties.

Transparency

PERILS strives for full transparency whilst always respecting the confidentiality agreements it has with its data providers, as well as competition and antitrust regulations.

Business Continuity Planning

The PERILS IT system and data are stored on secure servers located at high security hosting locations in Switzerland. The hosting provider is approved by the Swiss Federal Banking Commission (“SFBC”) and is certified by ISO 27001. It operates multiple locations at distances greater than 50 km. Data and server backups are created daily and are stored in different secure locations. A recovery system can be up and running within 24 hours.

Source of Data

PERILS receives data from insurance companies writing business in the territories covered by PERILS (with the exception of the territories of Indonesia, Philippines and Thailand). PERILS enters into a contractual agreement (the Data Provider Agreement) with each of the data providing insurance companies. The Data Provider Agreement defines, amongst other things, the data to be provided to PERILS, the reporting schedule, the data processing within PERILS, and the highly confidential nature and treatment of the data to be provided.

Each data providing insurance company, or their authorized intermediary, has an on-line account with PERILS, which has been created specifically for the purpose of delivering data to PERILS. Access is password protected with the site using SecureID technology. After receiving a request from PERILS to provide data, data providing insurance companies log into their user account and upload the requested data in a pre-defined format. This is followed by a standardized data format validation procedure allowing the data providing company to validate their uploaded data. After successful validation, the data-providing company signs off the data submission and delivers it to PERILS for further data processing and aggregation.

For Canada, PERILS receives the industry-level data directly from CatIQ (www.catiq.com).

For Indonesia, the Philippines and Thailand, PERILS takes a different approach to its other territories in that PERILS collaborates with leading industry stakeholders such as reinsurers, intermediaries and modelling firms as data sources.

Type of Collected Data

PERILS prepares and makes available, among other products and services, ultimate gross industry event loss data and industry exposure data (sums insured) regarding natural catastrophe perils.

Ultimate gross event loss data and exposure data are provided by the data providing insurance companies to PERILS on a per covered territory basis and on a per covered line of business basis. The geographical aggregation units used to report the data are, where feasible, CRESTA zones (for a definition of CRESTA zone, see www.cresta.org). The line of business aggregation units used to report the data are, where feasible, Residential Property, Commercial Property, Industrial Property, Agricultural Property (excl. crop, livestock and forestry), and Motor Hull.

Loss data provided to PERILS includes losses resulting from the designated main peril of an event as well as losses from ensuing perils and other additional loss costs as they are recorded in the data providing companies’ claims systems under the claims-affected policies for such an event.

Exposure data represent the data provider’s best estimate of the sums insured of the insured risks. Where the sums insured are not recorded in the original policies, PERILS relies on the conversion by the data-providing insurance companies of non-monetary exposure measures (e.g., sqm) into sums insured. Likewise, PERILS does not give any guidance for line of business classification and hence the data reported to PERILS are based entirely on the line of business classification of the data-providing insurance companies.

Deductibles and loss limits are the only primary insurance condition which are reported to PERILS on a quantitative basis.

For Canada, PERILS receives the industry-level data directly from CatIQ (www.catiq.com).

Catastrophe Event Designation

PERILS, in its professional judgment and sole discretion, identifies and designates a natural catastrophe event start date and end date as well as a unique PERILS event identification number. In making these judgments, PERILS considers factors such as event definitions applied by the insurance and reinsurance markets, information received from the data providing insurance companies, as well as objective scientific data and event identification by governmental or academic institutions, .

PERILS also determines, in its sole discretion, whether various insured losses occurring close in time to one another are to be considered the result of a single event or multiple (and separate) events. In determining whether one or more events have occurred, PERILS may consider the prevailing event definition(s) applied by the insurance and reinsurance markets, information received from the data providing insurance companies, and factors such as the inability on the part of insurance companies to distinguish the damage caused by the various events. As such, the PERILS event identification may differ from the event identification made or used by other sources.

In order to preserve its flexibility to adjust to external circumstances and enhance the quality of its estimates, PERILS may also, at its sole discretion, change its methodology for catastrophe event designation at any time and modify the application of its methodology.

For Canada, PERILS follows the catastrophe event designation by CatIQ (www.catiq.com).

Catastrophe Peril Designation

PERILS, in its professional judgment and sole discretion, identifies and designates the main natural peril involved in a catastrophe event. In making these judgements, and as a general rule, PERILS applies the basic principle that the main designated peril is that which causes the majority of the losses. Sources used when making these judgements include the peril classifications applied by data providing insurance companies, the peril classification applied by reinsurance markets, as well as information from governmental or academic institutions, and national weather or seismological services,. As such, the PERILS catastrophe peril designation may differ from the peril designation used by other sources.

The covered main designated perils depend on the covered territories and include Earthquake, Fire, Flood, Hail / Convective Storm, Extratropical Windstorm, Tropical Windstorm, and Volcanic Eruption (see table below). Loss data in connection with a catastrophe event include losses resulting from the designated main peril as well as losses from ensuing perils and other loss costs as they are allocated and paid out by insurance companies in connection with the catastrophe event.

For Canada, PERILS follows the catastrophe peril designation by CatIQ (www.catiq.com).

Company Data Processing

Data provided to PERILS by insurance companies includes exposure data (sums insured) by CRESTA zone and by country, property premium data by country, and ultimate gross event loss data by CRESTA zone and by country. This company data is made anonymous upon receipt and is tested for quality and completeness using standardized data quality and completeness checks. If the quality and/or completeness of data provided is deemed unsatisfactory by PERILS, PERILS will reject the data submission and make a new request for data to the data providing company. If company data quality and completeness are found to be adequate, the data provided is accepted by PERILS and is added to the data which has already been accepted within the identical aggregation units in the PERILS database. The original raw company data is deleted at this stage, in compliance with applicable antitrust and competition laws. None of the shareholders of PERILS have access to the data that is reported to PERILS by the data providing companies, and within PERILS only two employees have access to such data prior to its deletion as described above.

Although PERILS performs a series of standardized data quality control tests in the course of company data processing, PERILS does not independently verify or audit the accuracy of the data provided as part of its methodology. As a result, there can be no assurance that the data provided to PERILS has been, is, or will be accurate or complete.

For Canada, PERILS receives the industry-level data directly from CatIQ (www.catiq.com).

Industry Exposure Estimation

For industry exposure estimation, with the exception of the territories of Indonesia, Philippines and Thailand, aggregated company data within the identical aggregation units is extrapolated to industry-level (i.e., market-level) using market property premium information. The latter is broken down into individual aggregation units using population data and other proxy data such as land use or census data. Aggregated company property premium data is broken down into individual aggregation units using average rates as derived from the provided sums insured and premium data. The relationship between aggregated company premium and market premium then gives the market coverage per aggregation unit. The latter is used to extrapolate the aggregated exposure data per aggregation unit to industry-level (i.e., market-level). If the in-force date of the exposure information as received by data providers precedes the date of the PERILS exposure estimates ("Date as of", e.g., 1 Jan 2020), the market exposure is adjusted on a pro-rata to time basis, based generally on the latest available nominal GDP change for each country under consideration.

If market coverage is insufficient to calculate a reliable industry exposure, professional judgment is applied by PERILS in its sole discretion to adjust the calculated industry exposure. In making these judgments, PERILS may consider various factors, such as information provided by national and local authorities or insurance industry sources, as well as census, land use and/or other statistical data, as it deems appropriate in the particular circumstances.

In relation to the territories of Indonesia, the Philippines and Thailand, provided industry-level exposure data (refer to Source of Data) is critically evaluated by PERILS and individual weights are assigned to each industry exposure data set using expert judgement guided by the principles of accuracy, transparency and objectivity. The results of the evaluation are combined in a weighted-average approach to produce the best-estimate of the industry exposures (“Consensus IED”).

In order to preserve its flexibility to adjust to external circumstances and enhance the quality of its estimates PERILS may, in its sole discretion, change its general exposure estimation methodology at any time and modify the application of its methodology.

Industry Event Loss Estimation

For industry event loss estimation, aggregated company data within the identical aggregation units is extrapolated to industry-level (i.e., market-level) using market property premium information. For the first two loss declarations, aggregated company event loss data per country is used to extrapolate to industry-level. For all subsequent loss declarations, the property market premium is broken down into individual aggregation units using population data and other proxy data, such as land use or census data. Aggregated company property premium data is broken down into individual aggregation units using average rates as derived from provided sums insured and premium data. The relationship between aggregated company premium and market premium then gives the market coverage per aggregation unit. The latter is used to extrapolate the aggregated event loss data per aggregation unit to industry-level (i.e., market-level). If market coverage is insufficient to calculate a reliable Industry Event Loss Estimation, professional judgment is applied by PERILS, in its sole discretion, to adjust the calculated industry event loss data. In making these judgments, PERILS may consider various factors, such as meteorological and/or other scientific data, as well as information provided by national and local authorities or insurance industry sources, as it deems appropriate in the particular circumstances.

In order to preserve its flexibility to adjust to external circumstances and enhance the quality of its estimates PERILS may, in its sole discretion, change its general loss estimation methodology at any time and modify the application of its methodology.

Covered Territories, Business Lines and Natural Perils

PERILS currently collects exposure and loss data for the territories, lines of business and natural perils listed in the table below, with the exception of the territories of Indonesia, the Philippines and Thailand where PERILS only collects exposure data.

PERILS may collect data for other territories, lines of business and natural perils in the future.

Event Loss Reporting Threshold

PERILS will only report loss information if PERILS determines, in its sole discretion, that a particular event has resulted in total insured losses above a certain threshold for the covered perils and lines of business in the covered territories. For Extratropical Windstorms and Floods in Europe, this threshold is EUR 500m for pan-European events and EUR 300m for individual countries (currency exchange rate as at the time of the event). For Earthquakes in Italy and in Turkey, for Floods in Turkey, this threshold is EUR 300m (currency exchange rate as at the time of the event). For natural catastrophe events in Australia, this threshold is AUD 500m. For natural catastrophe events in Canada, this threshold is CAD 300m. For natural catastrophe events in New Zealand, this threshold is NZD 300m. For Floods and Tropical Cyclones in Japan, this threshold is JPY 100bn. For loss events below these reporting thresholds, PERILS will generally not report any loss information.

Reporting Schedule

PERILS expects to be able to make industry event loss data available in accordance with the following schedule:

1. First PERILS event loss declaration (loss information at country level only): expected no later than six weeks after the event end date.
2. Second PERILS event loss declaration (loss information at country level only): resurvey of industry event loss, expected no later than three months after the event end date.
3. Third PERILS event loss declaration (loss information at CRESTA Zone and line of business level): resurvey of industry event loss, expected no later than six months after the event end date.
4. Fourth PERILS event loss declaration (loss information at CRESTA Zone and line of business level): resurvey of industry event loss, expected no later than twelve months after the event end date.

After the twelve-month update, PERILS generally declares the event loss reporting period closed and gives its final estimate, unless there are substantial grounds to provide additional updates, which would occur at subsequent six-month intervals. Event loss reporting shall be closed, in any case, thirty-six (36) months after the event start date.

PERILS makes updated industry exposure data available once a year.

Reporting Currency

Industry loss data are made available in local currencies, Euros and US Dollars. Exchange rates used are daily mid-market rates by OANDA as at the event end date.

Industry exposure data are made available in national currencies, Euros and US Dollars. Exchange rates used are daily mid-market rates by OANDA as at the exposure in-force date.

Access to PERILS Data

PERILS delivers its industry exposure and loss estimates via the PERILS Portal and on a transaction-specific basis. Access to the PERILS Portal is via annual subscription in accordance with a PERILS database license. Transaction-specific delivery of PERILS data is in accordance with a transaction-specific license terms sheet.