For 10-year Schedule P lines of business (LOBs) that we analyzed, we find that:
For personal LOBs, the two different calibration approaches produce similar RRFs, with the standard formula calibrations slightly lower;
For the larger commercial LOBs, the individual company RRFs are lower, often much lower, than the standard formula RRFs;
With any of the calibration approaches, RRFs nearly always decrease as LOB-size increases, measured either in terms of reserves or premium; and
Among the three types of individual company RRFs measurement approaches,
The Feldblum normal chain ladder (Normal CL) and normal Bornhuetter- erguson (Normal BF) simulation methods tend to produce the highest RRFs.
The Meyers correlated chain ladder, the Feldblum log-normal chain ladder method and Mack analytical method are similar and lower than the first two methods.
Our analysis did not allow us to conclude that the observed ordering of the five stochastic methods is statistically significant
The comparisons identify areas of similarity and areas of difference that warrant further research to more fully understand the implications of the different approaches in practice.
These results are based on a particular sample of companies and the robustness of these results needs testing with future research.
This is one of several papers being issued by the RBC DCWP.
Keywords: Risk-Based Capital, Capital Requirements, Analyzing/Quantifying Risks, Assess/Prioritizing Risks, Integrating Risks
