Loss development is the process of adjusting historical loss data to its ultimate level to account for the fact that claims take time to be reported, settled, and closed.
Causes of Development
Loss development is driven by three primary factors:
- Development of Known Claims: Changes in the estimated case reserves on open claims as more information becomes available.
- Reporting of New Claims: Claims that occurred during the experience period but are reported late (pure IBNR).
- Reopening of Closed Claims: Claims that were closed but are subsequently reopened for additional payments.
The Overlap Fallacy
[!IMPORTANT] There is no overlap between loss development and loss trends. They serve distinct purposes:
- Loss Development: Projects historical losses from their maturity at the valuation date to their ultimate levels (reflecting report lag, settlement lag, and case reserve changes).
- Loss Trending: Projects ultimate losses from the average accident date of the historical period to the average accident date of the future policy period (reflecting inflation and other prospective changes).
Age-to-Age Factor Selection
When selecting age-to-age factors (LDFs) from a triangle, actuaries evaluate:
- Progression down columns: Factors should progress smoothly (typically decreasing toward 1.0) across maturities.
- Stability within columns: Ratios within the same column should be stable; if not, identify if fluctuations are random or systematic.
- Data Credibility: If historical data is thin or volatile, blend it with industry benchmarks.
- Operational Changes: Adjust for shifts in insurer claims-handling practices or reserving guidelines.
- Prospective Applicability: Ensure the historical patterns are representative of the future policy period.
Tail Factors
- When to Use: Apply a tail factor when the age-to-age factor at the final available maturity in the triangle is still greater than 1.0.
- Methods to Calculate:
- Extrapolate by fitting a curve (e.g., exponential or power curve) to the selected age-to-age factors.
- Utilize industry benchmark tail factors.
General Assumptions of the Development Method
The chain ladder (development) method relies on several key assumptions:
- Consistency of Development: The relative development of future claims will be similar to the development observed in prior periods. This implies:
- Consistent claims processing and settlement speeds.
- Stable case reserve adequacy (CRA).
- Stable mix of claim types, policy limits, deductibles, and reinsurance limits.
- Predictive Value: Claims observed during an immature period contain useful information about ultimate claim levels.
Diagnostic Triangles for Reserving and Claims Changes
To detect changes in claims handling and reserving practices, actuaries analyze diagnostic ratios from development triangles.
Key Diagnostic Metrics
| Diagnostic Ratio | Metric Measured | Interpretation & Insights |
|---|---|---|
| Claim Settlement Rate (CSR) | Measures speed of closure. An increase indicates faster claims processing. This ratio is independent of claim severity. | |
| Closed-Without-Payment % | Measures the proportion of claims closed with zero payment. Changes indicate shifts in claims filtering or nuisance claim handling. | |
| Average Paid Claim Size () | Paid Severity / CSR Type | Indicates severity trends. If CSR increases for small claims first, average paid claim size will initially decrease. |
| Average Case Reserve () | Case Reserve Adequacy (CRA) | Measures the average strength of open reserves. An increase can indicate reserve strengthening (CRA) or that small claims are closing faster (leaving large claims). |
| CSR / CRA Mix | A decreasing trend down columns indicates a slowdown in claims settlement (lower CSR) or an increase in case reserve adequacy (higher CRA). | |
| YoY Paid Loss % Change | Growth Rate / Trend | Helps distinguish between claim speedup and actual loss deterioration. |
Reserving Changes: Mechanisms and Impacts
1. Speedup in Claim Settlement Rates (CSR)
- Mechanism: The claims department closes claims faster than in the past (typically starting with smaller, simpler claims).
- Impact on Triangles:
- Open counts decrease.
- Average paid claim size decreases (since many small claims are paid quickly).
- Average case reserve increases (since only large, complex claims remain open).
- Total case reserves decrease.
- Distortion: Standard paid LDFs will overstate ultimate losses because the historical factors project development based on a slower historical settlement pattern.
2. Increase in Case Reserve Adequacy (CRA)
- Mechanism: The claims department sets higher reserves on open claims earlier in the process (reserve strengthening).
- Impact on Triangles:
- Reported (incurred) losses increase in early maturities.
- Average case reserves increase.
- Paid-to-reported ratio decreases.
- Distortion: Standard reported LDFs will overstate ultimate losses because the historical factors project development assuming case reserves are still thin.
Strategic Diagnostic Analysis
Actuaries analyze diagnostic triangles in a structured order to isolate changes:
- Reported Claim Counts: Check for stability. If stable, reporting patterns are consistent, and changes are driven by settlement or reserving.
- Settlement Ratio (): Determine if the settlement speed (CSR) has changed.
- Average Paid and Average Case Ratios: Verify if the change is driven by CSR shifts (small vs. large claims) or reserving changes (CRA).
- Formulate Questions for Management:
- Have adjuster workloads or claim assignment policies changed?
- Has there been a priority change in handling small claims vs. large claims?
- Have there been changes in the authority levels required to settle claims?
Miscellaneous Considerations
Aggregation Basis (AY vs. RY vs. PY)
- Accident Year (AY): Standard for occurrence policies. Best for comparing with industry benchmarks and when events (e.g., weather) correlate with occurrence dates.
- Report Year (RY): Standard for claims-made policies. Best when changes in the legal/social environment correlate with the date the claim is filed.
- Policy Year (PY): Best when analysis needs to isolate the impact of policy changes (e.g., deductible or limit changes) based on policy effective dates.
Combining Lines of Business (LOBs)
- When to Combine: Combine LOBs if individual data volumes are too small to be credible.
- When NOT to Combine: Do not combine if the lines have:
- Different severities (e.g., long-tailed liability vs. short-tailed property).
- Different claim settlement rates.
- Different reserving philosophies.
- Product Mix Shift: If lines are combined, a shift in the mix of business will distort the combined development factors, leading to inaccurate projections.