Actuary Interview Questions
& Model Answers, 2026

It develops paid or incurred losses to ultimate using historical development factors, assuming future development follows the past and that the mix, claims handling and legal environment are stable. It fails on immature or sparse accident periods where a small early movement is leveraged into a large ultimate, and after a change in case-reserving practice, claim settlement speed or coverage — precisely when the answer matters most.

When the experience is too immature or volatile for development factors to be credible, because it blends an expected loss ratio with actual emergence and is far less leveraged by early noise. The judgement sits in the a priori loss ratio, which has to be defensible in its own right. Using it with an a priori simply set to produce the desired answer is the failure mode reviewers look for.

Bring premium to current rate level, develop and trend losses to the prospective period, load for expenses, reinsurance cost and profit, and compare the indicated rate with the current one. Then examine mix, large-loss treatment and any distortion from a single event. Rate indications presented without the trend and on-level assumptions exposed cannot be reviewed by anyone, which is why the working is the deliverable.

With evidence rather than judgement alone: the data that suggests the change, sensitivity of the result to it, whether it is a genuine shift or a random fluctuation over a short window, consistency with related assumptions, and documentation of the rationale. Changes made in the same direction as management's preference need the strongest support, and reviewers know exactly where to look.

Reconcile inputs back to source data, check totals and record counts, run the model on a period with a known answer, test sensitivity to key parameters, look at results by segment for anomalies rather than only in aggregate, and have someone independent replicate the critical calculation. A result that is not reconcilable to the ledger is not a result yet.

Lead with what changed and what it means for the financial statement, then the driver in plain terms — claims settling higher than expected, a legal development, a change in reporting speed — the range of uncertainty, and what you will watch next. Avoid method names as explanations. Actuaries who cannot do this get overruled by people with worse information.

Panels want a realistic plan: exams passed, the next sitting, study hours per week and how they fit around work. Exam progress is a hard hiring criterion in this field and vague answers are read as a stalled track.

Strong answers welcome the challenge, check the point rather than dismissing it, and separate a genuine error from a disagreement about assumptions. Actuaries whose first reaction is defensiveness are difficult to trust with a reserve opinion.

Interviewers look for immediate disclosure, an assessment of what was affected, and a control added afterwards. Concealing an error is career-ending in an actuarial function and everyone in the room knows it.

Good answers show curiosity about the operational reality behind the data — how cases are reserved, how business is being written — because that context is what makes an assumption defensible.

Anchor and position by exam progress. The BLS OEWS May 2025 national median for actuaries is $130,000 a year ($62.50/hr), with the top 10% above $215,100. Exam count, credential status and practice area are the strongest determinants inside that band, so state your exam position precisely and position accordingly.

Most employers run a study programme with paid study hours, exam fees and materials covered, and a defined salary increase per exam passed, sometimes with a bonus. Ask for the actual scale rather than a description, and ask what happens if you fail a sitting — policies on repeated attempts vary and matter.

Study hours and whether they are genuinely protected, exam fee and seminar coverage, the practice area and whether you can rotate, mentoring by credentialed actuaries, and remote arrangements. Protected study time is worth more than a small salary difference during the exam years.