IFoA SP2 common mistakes
IFoA SP2 · Understand the traps and how to avoid them.
Writing a calculation-only answer to a 'describe and explain' or 'discuss' question - producing numbers and a one-line conclusion but no prose on assumptions, method choice, or implications for the insurer.
Candidates carry over the habit of demonstrating arithmetic competence and treat the mark allocation as if accuracy of the final figure were the objective; under time pressure the prose feels like padding and gets cut first.
What to do instead
Before writing, count the command words in the question and allocate marks to them explicitly (e.g. 'state' 1 mark each, 'explain' 2–3 marks each, 'discuss' 4+ marks). Draft the prose points first as bullet headings, then attach the calculation as supporting evidence. In mocks, mark your own script against the command words and record how many marks you lost purely to missing commentary.
Presenting a single point estimate for a premium, reserve, or capital figure with no sensitivity, no stochastic range, and no statement of the key assumptions driving it.
Exam questions often say 'calculate' and candidates stop at the number; the professional judgement layer (which assumptions dominate, how the answer moves under stress) is not explicitly demanded by the verb, so it is overlooked.
What to do instead
Adopt a fixed post-calculation routine: after any numerical answer, add two or three sentences covering (i) the two or three assumptions with the largest leverage on the result, (ii) the direction of change if each is stressed, and (iii) any material interaction with the insurer's risk appetite or capital position. Practise this until it is automatic, and time-box it to about a fifth of the question's time.
Applying a formula mechanically - a net premium reserve, a profit signature, or an embedded value - without checking that the product's cash-flow pattern, benefit structure, and expense/investment basis actually satisfy the formula's underlying assumptions.
Revision rewards memorised formulae, and under exam pressure the familiar formula is the path of least resistance; candidates rarely pause to validate the model against the specific product described in the scenario.
What to do instead
For each major formula, write a one-line 'validity condition' card (e.g. reserves require the equivalence principle and a specified basis; profit signatures require a defined reserving basis and timing of cash flows). In the exam, state the validity condition explicitly in your answer before applying the formula - this both earns method marks and forces the check.
Ignoring the insurer's constraints and context given in the scenario - solvency capital, free surplus, reinsurance capacity, distribution channel, or regulatory regime - and answering as if the company were an unconstrained textbook entity.
Candidates focus on the actuarial mechanics they have drilled and treat the narrative preamble as scene-setting rather than as data that should shape the recommendation.
What to do instead
Read the scenario twice: first for the question, second solely to extract and list the constraints (capital, risk appetite, distribution, regulation, existing book, tax). Keep that list beside you and tick each constraint off in the recommendation. A recommendation that ignores a stated constraint typically forfeits the higher-order marks even when the maths is correct.
Treating the risk discount rate, the reserving basis, and the pricing basis as interchangeable, or justifying a rate purely by 'the cost of capital' with no reference to the risk profile of the specific cash flows being discounted.
These three bases are taught close together and share terminology, so candidates conflate them; the distinction between a best-estimate, a prudent, and a market-consistent basis is easy to recite but hard to apply under time pressure.
What to do instead
Build a comparison table with rows for purpose (pricing, reserving, capital, reporting), basis (best estimate, prudent, market-consistent), and permitted discount rate. In answers, always name the purpose first and derive the basis from it, and state explicitly whether the rate includes a risk margin or allowance for non-diversifiable risk.
Describing a product design or pricing proposal without linking it to the insurer's profitability metrics - internal rate of return, net present value, break-even year, and profit signature shape - and without saying how the design would be altered if those metrics failed a hurdle.
Product design questions invite descriptive answers about benefits and options, and candidates fill the space with feature lists because that is easier than committing to a financial judgement.
What to do instead
For every design or pricing question, end with a short paragraph structured as: metric, hurdle, result, action. State the hurdle the insurer would plausibly use, whether the design clears it, and the specific levers (premium, charges, surrender terms, reinsurance, asset mix) you would pull if it did not.
Discussing reinsurance, underwriting, or capital management in isolation, without tracing the second-order effects on reserves, capital requirements, policyholder outcomes, and the insurer's competitive position.
Examiners' questions are often phrased narrowly ('outline the types of reinsurance available'), and candidates answer exactly the narrow question, missing the marks attached to consequences.
What to do instead
Use a fixed consequences checklist for any risk-transfer or management-action question: effect on net cash flows, on reserves and capital, on the volatility of results, on policyholder terms and fair treatment, on competitive position, and on operational and counterparty risk. Apply the checklist even when the question does not prompt for it, and label each consequence so the marker can find it.
Writing long, unstructured prose with no headings, no numbered points, and buried conclusions, so the marker cannot locate the individual marking points within the time available.
Candidates equate length with depth and write continuously to feel productive; they also fear that short structured answers will look thin.
What to do instead
Adopt a house style: a one-sentence direct answer, then numbered points each beginning with a bolded keyword, then a short conclusion. Aim for one idea per point and one point per mark. In practice, mark your own scripts with a stopwatch to see how quickly a marker could find your credit-worthy statements.
Running out of time because early calculation questions are over-worked - rechecking arithmetic repeatedly and leaving the later, higher-mark discussion questions unattempted or answered in note form.
Calculations give a false sense of certainty, so candidates keep polishing them; discussion questions feel open-ended and are deferred until the end, when there is no time left.
What to do instead
Set a hard per-question time budget in proportion to marks and write it in the margin. Do a first pass answering every question at a basic level, then return to deepen the ones you know best. Never leave a discussion question blank: a structured five-point outline scores more than a perfect calculation plus an empty page.
Recommending an action - a hedging strategy, an asset-liability matching approach, a bonus or surrender-value policy - without stating the risks it creates, the monitoring that would detect failure, and the conditions under which the recommendation should be reversed.
Candidates frame the task as choosing the best option rather than as managing a decision through time, so implementation, governance, and exit conditions are never considered.
What to do instead
End every recommendation with three sentences: the principal residual risks it leaves, the metrics and frequency with which the insurer should monitor them, and the trigger that would cause the decision to be revisited. Treat this as a required component of the answer rather than an optional flourish.