When developing loss severity distributions, risk professionals should organize loss data by

When developing loss severity distributions, risk professionals should organize loss data by

• A. Date of loss.
• B. Time of loss.
• C. Location of loss.
• D. Size of loss.

Answer: (C)

Organizing loss data by location helps reveal how geographic cost drivers shape how much losses cost when they occur. Different places have different construction costs, labor rates, regulatory environments, exposure mixes, and peril frequencies, all of which influence how severe a loss can be. By grouping losses by location, you create more homogeneous data sets, so the severity distribution you fit reflects the true cost structure in each area rather than a blend that washes out important differences. This leads to more accurate predictions for reserves, pricing, and capital needs tailored to each location. For example, claim severities in high-cost urban areas with expensive construction and stricter codes will tend to be higher than in lower-cost regions. If you mix these together, the resulting distribution can misrepresent tail risk and required reserves for any given location. Time or date might capture trends or seasonality, but it doesn’t address geographic variability in severity. Size of loss is the variable you’re modeling, not a way to organize the data, so it wouldn’t serve the goal of creating accurate, location-specific severity distributions.

Organizing loss data by location helps reveal how geographic cost drivers shape how much losses cost when they occur. Different places have different construction costs, labor rates, regulatory environments, exposure mixes, and peril frequencies, all of which influence how severe a loss can be. By grouping losses by location, you create more homogeneous data sets, so the severity distribution you fit reflects the true cost structure in each area rather than a blend that washes out important differences. This leads to more accurate predictions for reserves, pricing, and capital needs tailored to each location.

For example, claim severities in high-cost urban areas with expensive construction and stricter codes will tend to be higher than in lower-cost regions. If you mix these together, the resulting distribution can misrepresent tail risk and required reserves for any given location. Time or date might capture trends or seasonality, but it doesn’t address geographic variability in severity. Size of loss is the variable you’re modeling, not a way to organize the data, so it wouldn’t serve the goal of creating accurate, location-specific severity distributions.