Fall 2016 Q20 b)

Comments

• Graham

  October 2023

  Yes, Sample Answer 2 touches on an important aspect of the frequency-severity technique. When using the frequency-severity method to calculate claims, the basic idea is to separate the two components:

  1. Frequency: How often a claim occurs.
  2. Severity: The average cost of those claims.

  The assumption highlighted in Sample Answer 2 is that there are no partial payments and all claims are paid at close. This is a simplifying assumption often made in basic frequency-severity models. This is briefly mentioned here:

  • https://www.battleacts5.ca/wiki5/Friedland11.FreqSev#Example_A:_Intro_to_FS_Methods

  Ok, now why is that an assumption?

  • Consistency: It ensures that when a claim is counted (i.e., in frequency), its full value is also captured in severity. If you have partial payments, then the claim might be counted in frequency, but only a fraction of its cost would be captured in severity. This can distort the calculations.
  • Ease of Calculation: If all claims are paid at close, it simplifies the calculation process. You don't have to account for claims that might be only partially paid in one period and then settled in a future period.
  • Predictability: It provides a more stable basis for predicting future claims. If claims are paid out in a piecemeal fashion, it might introduce volatility into the predictions.

  However, if an insurance line of business frequently deals with partial payments (like the one described), then the basic frequency-severity technique might not be the best fit without adjustments. This is because the underlying assumptions of the model don't align well with the real-world operations of that business.