By Asad Khalil

The hardening insurance market has put significant pressure on the premium spend of large corporates. Captive formation enquiries over the last 18 months have significantly increased both in Europe and the US.

Actuaries help captives in a number of areas such as reserving, technical pricing, risk financing and capital requirements. At the initial stages of a feasibility study when a corporate is considering setting up a captive, the actuary will work with the captive manager and group insurance/risk manager to understand what the risk appetite of the proposed captive is.

They will then use this information together with historical claims to propose a number of (re)insurance programme structures which will allow the captive to self-insure a significant amount of risk. The purpose of this study is to help optimise the (re)insurance programme design of the proposed captive by minimising their Total Cost of Risk (TCoR).

For actuaries to build the models underlying the feasibility study they rely heavily on the data provided to them. This article will outline what data is needed and why it’s important.

Individual claims data

Individual claims data should contain basic information about each loss that has occurred. This will include the date of loss, reported date, currency, location/state, claim reference number, policy number, paid amount, outstanding amount, incurred amount, fees/expenses and excess/deductibles. This is by no means an exhaustive list but will form the core dataset.

When working with corporations with a dedicated group insurance manager who is considering setting up a captive, individual claims data is usually readily available in the desired format.

However the current market conditions have given rise to insurance brokers seeking to utilise captives as rates have been steadily increasing over the last 12 to 18 months. In these circumstances obtaining the individual claims data may be more challenging as the ownership of the data usually rests with the carrier and not the broker. While it is possible to obtain individual claims data, it is common to receive a loss summary on an aggregated level.

Individual claims data allow actuaries to fit severity and frequency distributions to accurately model the losses and therefore building a more realistic loss model. Summarised claims data limits the statistical techniques that actuaries can use to fit severity distributions and this reduced flexibility in some cases can lead to a less accurate loss model.

Triangulations

Triangulations for incurred claim amounts allow the actuaries to derive development patterns and to project claims to the amounts they will eventually settle to, this process makes allowance for Incurred But Not Reported claims (IBNR).

Claim numbers triangulations are also important, they help the actuaries to predict the number of claims the client is likely to have for each policy year.

In the event triangulations are not available then benchmarks development patterns will be used, if these patterns develop slower than the actual experience of the portfolio then the actuaries projection of the performance will be worse than it is in reality. This may result in a higher projected captive premium and a reduced return on investment for the captive.

Conversely benchmarks development patterns that develop faster than the actual experience of the client’s portfolio may result in a captive premium that may not be sufficient to cover retained losses.

Exposure data

Exposure data provides us an insight into the underlying risk. For example if we consider Workers’ Compensation or Employers’ Liability an appropriate exposure measures would be wage roll or headcount.

Year-on-year exposure data provides actuaries with an understanding into how the risk has changed over time and allows them to deduce trending factors which need to be applied to historical claims to accurately predict the losses for the forthcoming year.

Lack of exposure data would make it difficult to utilise historical data and would lead the actuaries to rely more heavily on data from recent years which are more reflective of the risk profile of the client’s portfolio at the present time.

Current (re)insurance programme design and premiums

Once the gross loss model is built, the actuaries can overlay the current and alternative (re)insurance structures. This allows the client to determine the most efficient structure for their business. This information is usually easy to obtain.

Summary

The current market dynamics have created a perfect storm of rate increases and reduced capacity from the insurance market. This provides corporates and insurance brokers with a great opportunity to retain a significant amount of risk through a captive structure and benefit from underwriting profits. The rewards are great and so are the risks. To maximise your chances of success having accurate and granular data that allows you to make informed decisions is essential.