FSA courses

CP 312

Model Development and Governance

2 discussions

Explanation

CP 312: Validation should test the model’s intended use and limitations

A model can be internally consistent while unsuitable for a decision. Validation needs boundary tests, independent comparisons and a clear statement of what the model cannot represent.

Original practice scenario

In an original projection, set mortality, lapse and interest to zero and compare with hand-calculated scheduled payments. Then test realistic stresses and explain any behaviour assumption. Record the model owner and monitoring trigger; a successful boundary test is necessary evidence, not complete validation.

Study check

Write the assumptions, the decision being supported and one limitation. Use the current course syllabus for the required terminology and scope; this introductory guide is not a complete course summary.

Reading sources

Official syllabus and resources

ActNet editorial guide · October 1, 2026 · Original illustrative examples.

0 repliesOct 1

Explanation

CP 312: Model change control separates code changes from assumption changes

A model review should identify what changed, why and the financial effect. Data, assumption and implementation changes need distinct checks, even when they appear in one release.

Original practice scenario

For an original long-term cash-flow model update, run old and new code on the same data and assumptions, then change one input category at a time. Reconcile differences and document interactions. A matching total alone does not prove every benefit calculation is correct.

Study check

Reading sources

Official syllabus and resources

ActNet editorial guide · October 1, 2026 · Original illustrative examples.

0 repliesOct 1