toolbox
Home / Taxes, Insurance & Benefits / Product Liability Loss Probability Calculator

Product Liability Loss Probability Calculator

The Product Liability Loss Probability Calculator estimates the chance of at least one modeled product-liability claim from a specified number of units sold. It combines an incident rate per 10,000 units with the percentage of incidents assumed to become covered claims, giving a transparent volume-based scenario for product exposure.

This structure is useful when unit volume is a major driver of risk, but it is intentionally simple. Real claims may be correlated by batch, design defect, geography, customer behavior, or recall event, so independent-unit assumptions can understate clustered losses. Use internal quality and claims data when available rather than treating generic rates as universal.

Inputs

units
per 10k
%
Result
—
Probability of at least one modeled claim
Modeled claim probability per unit—
Expected modeled claims—
Probability of no modeled claim—
Modeled claims per 10,000 units—

1. Enter exposed units
Use the number of units sold or otherwise exposed during the period you are modeling.

2. Enter incident rate
State the assumed product incident frequency per 10,000 units.

3. Estimate claim conversion
Enter the percentage of those incidents assumed to become covered product-liability claims.

4. Review at-least-one probability
The main result converts the per-unit claim probability into a portfolio-level chance across all entered units.

5. Stress-test assumptions
Rerun the model with higher incident or conversion rates when uncertainty or batch concentration is material.

Formula: Claim probability per unit = (Incident rate per 10,000 ÷ 10,000) × Claim-conversion rate Probability of at least one claim = 1 − (1 − Claim probability per unit)^Units Expected modeled claims = Units × Claim probability per unit

Where:

  • Incident rate per 10,000 — modeled incidents for each 10,000 exposed units
  • Claim-conversion rate — share of incidents assumed to become covered claims, decimal
  • Units — number of independently modeled exposed units

Assumptions: Each unit is treated as an independent exposure with the same claim probability. This can be unrealistic when defects or incidents affect many units together.

What the result means

Independent-unit assumptions may not capture batch, systemic, or correlated product defects.

Review the actual policy, quote, endorsements, exclusions, limits, and applicable requirements before making an insurance decision.

Given:

  • Units sold: 50,000
  • Incident rate: 2.5 per 10,000 units
  • Incidents becoming covered claims: 35%

Calculation:
Per-unit claim probability = (2.5 ÷ 10,000) × 0.35 = 0.0000875. Expected claims = 50,000 × 0.0000875 = 4.375. Probability of at least one = 1 − (1 − 0.0000875)^50,000 ≈ 98.74%.

Result:
Modeled probability of at least one claim: about 98.7%.

The high portfolio probability comes from applying a very small per-unit probability across 50,000 exposures; it does not imply that any specific unit is likely to generate a claim.

Why can the portfolio probability be high when the per-unit rate is tiny?

Many independent exposures accumulate risk. A small per-unit probability can still make at least one event likely when the unit count is large.

What does “incident conversion” mean here?

It is your assumption about the percentage of product incidents that become covered liability claims. It is not a universal industry constant and should come from relevant experience or analysis.

Can I use shipments instead of units?

Only if a shipment is the exposure unit that matches your incident-rate data. Keep the numerator and denominator consistent.

What if one defect affects an entire batch?

The independent-unit model can understate clustered risk. For batch or systemic defects, use a scenario model that treats the batch as a common event rather than independent units.

Is expected claim count the same as probability of at least one claim?

No. Expected count is an average number of modeled claims, while the main probability only asks whether one or more occur. They answer different risk questions.