Chapter 9 Using the ODP Bootstrap Model: A Practitioner’s Guide - Shapland

Notations are similar to Venter

• See Table 9.1 for sample triangle layout

Model Parameters:

• Mean (9.1)

• Variance (9.3) and dispersion factor

  • Need residual: unscaled (9.8), scaled (9.9), standardized (9.10)

• Parameterize with GLM or simplified (ODP) (Note the condition required for ODP)

\(\star\) Simulation procedure

• Parameter variance (9.14)

• Process variance (9.15)

• Other variation of the bootstrap (e.g. incurred, BF, Cape Cod)

• Pros and Cons of GLM

• GLM vs ODP bootstrap

\(\star\) Practical issues:

1. \(\star\) Negative Incremental Values

   • During fitting: (9.16) and (9.17)

   • During simulation: (9.21)

2. Non‐Zero Sum of Residuals

3. Using an L-year Weighted Average

4. Missing Values

5. Outliers

6. \(\star\) Heteroscedasticity

   • Stratified sampling

   • Hetero adjustment to residuals

   • Non-constant scale parameter

7. Heteroecthesious Data

8. Exposure Adjustment

9. Tail Factors

10. Fitting a Distribution to ODP Bootstrap Residuals

5 diagnostics

1. \(\star\) Residual graphs

2. Normality test

3. Outliers

4. Parameter adjustment

5. \(\star\) Model results

Multiple models

Model testing