1. The pricing and profitability of the overall segment has historically been very cyclical, starting from super-profits by early players followed by entry capital into the market driving the price down to unprofitable levels.
2. Insurers tend to focus selecting policies with lower risk exposure regardless of the market price which may be unprofitable in the long run.
3. Insurers may focus on writing high hazard risks for the high premium charged. Due to the typical low frequency and high severity claims for a typical Corporate portfolio, the high hazard risks can make super profits for a number of years but are susceptible to large losses which can result in a significant loss larger than all of the achieved profits over the period.
ADDING VALUE TO CORPORATE INSURANCE PORTFOLIO
1. Utilising natural peril models to assess both the exposure and the expected cost of these type of losses to charge an appropriate price for this exposure and manage aggregate exposure in high risk areas.
2. Develop a framework which provides a view of the appropriate price that should be charged for an individual policy. At the basic level, the framework should provide information on the total non-claims related costs such as administration expenses, reinsurance expenses, commission and profit margin.
3. Enhance the framework further by including claims costs based either on portfolio level data or the historical claims experience of the actual policy, with appropriate allowances for IBNR and IBNER claims, inflation and large losses.
4. Compare the impact of the different reinsurance arrangements on the expected profitability and the volatility of the profitability of the portfolio.
5. Analysis of portfolio level historical claims experience by industry and policies that performed poorly consistently.
LARGE LOSS MODELLING
1. Reasons causing a lack of data for estimates and pricing are as follow:-
2. The frequency of large losses is very low.
3. The large loss experience are heavily impacted by the skill of the underwriter or changes in underwriting standard.
4. Business mix changes causing certain types of historical large losses to not re-occur in the future as the insurer may no longer insure the particular type of policy that has caused the claim. Conversely, the insurer may be exposed to new risks which it was not exposed to historically.
ALLOCATING LARGE LOSSES BY SEGMENT
1. The standard methods for working losses such as GLMs and multi-way segmentation are not effective because the number of claims is small.
2. Same goes to portfolios or segments within a portfolio whose primary exposure is low frequency, high severity events.
3. Policies with sums insured below the large loss limit cannot incur a large loss.
4. A larger proportion of large loss costs should be allocated to industries that are known to have a higher risk and classified as high hazard industries.
MODELLING FOR COMMERCIAL PORTFOLIOS
1. Stochastic modelling for aggregate deductibles – Aggregate deductible is where the insurer will only cover the claims costs for the policy which exceed the aggregate deductible limit.Calculating a price without using stochastic modelling can be extremely difficult particularly if the policy covers a number of heterogeneous risks. On a side note, by applying different levels of aggregate deductable, a reduction in claims cost can be expected.
2. Simple Portfolio Segmentation – Although with limited homogenous data, significant value can often be gained from relatively simple analysis such as frequency, average claim size and loss ratio portfolio segmentation which can provide insights into the underlying drivers of the claims performance.
3. Credibility analysis – This can be used to refine technical premium at either a vehicle, postcode or industry level. In addition, credibility analysis can be used when analysing a portfolio with limited data to test the likelihood of emerging trends being persistent or purely fortuitous.
2. Simple Portfolio Segmentation – Although with limited homogenous data, significant value can often be gained from relatively simple analysis such as frequency, average claim size and loss ratio portfolio segmentation which can provide insights into the underlying drivers of the claims performance.
3. Credibility analysis – This can be used to refine technical premium at either a vehicle, postcode or industry level. In addition, credibility analysis can be used when analysing a portfolio with limited data to test the likelihood of emerging trends being persistent or purely fortuitous.
