Skip to content

IAI Actuarial Core Principles · Risk Modelling and Survival Analysis · Mortality projection

When projecting mortality with a CBD model, κ(t) = (κ1(t), κ2(t)) is commonly modelled as a two-dimensional random walk with drift. What is the main consequence for projected mortality?

With κ modelled as a two-dimensional random walk with drift, uncertainty about future mortality widens as the projection horizon grows, and the innovations in κ1 and κ2 may be correlated through a covariance matrix. Future values also depend on the latest fitted κ.

  1. AProjected rates are deterministic with no uncertainty
  2. BUncertainty about future rates grows with the projection horizon, and κ1 and κ2 changes can be correlatedCorrect
  3. CFuture κ values are independent of the current κ
  4. DThe two components must have the same drift
  5. Mortality at all ages is forced to converge to zero

Explanation

A random walk with drift has variance increasing linearly with time, so projection uncertainty widens. A multivariate version has a covariance matrix allowing correlated innovations between κ1 and κ2. Future values depend on the current value, so independence is false.

Did you get it right without looking?

One question tells you little. A timed set on Mortality projection shows your real accuracy, how long you take and where you lose marks.

More Mortality projection questions