Actuarial Notes

Limitation Period

Study guide: Exam 6C

A Limitation Period is the statutory deadline for starting a legal proceeding. Once it expires the claim is barred no matter how meritorious it is, so limitation law determines when an insurer's exposure to unreported claims finally ends — and a change to it moves reserves on business already written.

  • The modern Canadian model (Ontario's Limitations Act, 2002 is the template) is a basic limitation period of two years from discovery, subject to an ultimate limitation period of fifteen years from the act or omission regardless of discovery.
  • Discoverability is the doctrine that makes the two-year clock start when the claimant knew or ought reasonably to have known that the injury occurred, was caused by the defendant, and warranted a proceeding — not when the act occurred. It is what allows latent-injury and long-tail claims to surface decades later.
  • Tolling suspends the clock for minors, for persons under disability, and while the parties are engaged in a statutory dispute-resolution process. A claim by an infant can therefore be brought many years after the accident, which is why paediatric bodily injury claims have such long reporting tails.
  • Actuarial consequences: the limitation period defines the IBNR reporting tail. Shortening it truncates late reporting and releases reserve; lengthening it, or a decision expanding discoverability, extends the tail on claims already incurred. Sexual abuse claims, where several provinces removed the limitation period entirely, are the largest Canadian example — a legislative change that created liability on policies written decades earlier.
  • Policy conditions impose their own, shorter, notice requirements, but statutory limitation periods generally override contractual attempts to shorten them below the statutory floor.
  • Quebec's prescription periods under the Civil Code of Québec serve the same function with different rules — three years for personal injury, running from manifestation of the injury.

A Limitation Period Is Removed A province abolishes the limitation period for claims arising from sexual abuse, with retroactive effect. An insurer wrote general liability for institutional clients from 1970 to 2000, with policies typically carrying \$1 million occurrence limits.

What is the actuarial problem?

Answer The insurer now has exposure on policies it stopped writing decades ago, priced with no allowance for this, and the ordinary reserving toolkit is unusable.

Why standard methods fail:

  • No development triangle exists for a claim type that could not previously be brought. Historical patterns describe a legal regime that no longer applies.
  • Reporting is driven by social and legal factors — public inquiries, media coverage, other institutions' settlements — not by an exposure base. Frequency is essentially unforecastable from the insurer's own data.
  • Coverage questions are as uncertain as frequency: which policy year responds (occurrence versus manifestation), whether abuse is an "occurrence," whether intentional-act exclusions apply, and how multiple years stack.

What the actuary can actually do:

  1. Exposure-based estimation. Identify insured institutions of the type at risk, estimate claims per institution from other jurisdictions' emergence, and apply limits and attachment points policy year by policy year.
  2. Scenario ranges, not a point estimate. Present low/central/high with the drivers stated. A single number here conveys false precision.
  3. A large risk adjustment, reflecting genuine parameter and model uncertainty rather than process variance.
  4. FCT scenario testing — this is the archetype of a latent liability that can threaten financial condition, and it belongs in front of the board.
  5. Check the reinsurance. Recoveries on thirty-year-old treaties depend on wordings, aggregate limits and, critically, whether those reinsurers are still solvent and collectible.