CPA Rwanda common mistakes
CPA Rwanda · Understand the traps and how to avoid them.
Presenting a consolidated statement of financial position that omits the fair-value adjustments at acquisition and only aggregates the parent's and subsidiary's carrying amounts line by line.
Candidates memorise the mechanics of adding balances and eliminating intragroup items but treat fair-value uplifts and goodwill as a separate 'extra' rather than part of the working, often because they practised only on simple book-value acquisitions.
What to do instead
Build a standing acquisition working for every consolidation: fair value of consideration, fair value of net assets at acquisition (including identified intangibles and uplifts), goodwill as the residual, and post-acquisition movement. Reconcile group retained earnings to the parent's balance plus group share of post-acquisition reserves before writing the statement.
Computing variances on the wrong basis (for example, using actual quantity rather than standard quantity for the price variance, or reversing the sign convention) so favourable and adverse labels are swapped.
Under exam pressure candidates apply the formula from memory without anchoring it to a definition, and many textbooks phrase the price/rate variance differently, so the sign is lost.
What to do instead
Before calculating, write the definition in words: price variance = (standard price − actual price) × actual quantity; usage variance = (standard quantity for actual output − actual quantity) × standard price. State whether a positive result is favourable or adverse explicitly in the answer, and sanity-check one variance against intuition (paying more than standard should be adverse).
Describing audit procedures in vague terms such as 'check the invoices' or 'review the ledger' without specifying the assertion tested, the population, the sample selection, or the evidence that would confirm or contradict the figure.
Candidates learn the audit cycle at a conceptual level and never practise writing procedures that would satisfy a reviewer; they also confuse the direction of testing.
What to do instead
Write every procedure in a fixed four-part structure: assertion addressed, source document or record, action (inspect, recalculate, confirm, observe, re-perform), and the conclusion it supports. For occurrence, test from the ledger back to supporting documents; for completeness, test from source documents forward to the ledger. Practise writing procedures for a named balance until the structure is automatic.
Applying the wrong tax treatment to a transaction because the candidate has not identified whether the amount is inclusive or exclusive of VAT, or has misclassified the income or expense head under the applicable Rwandan tax law.
Candidates rely on general principles from other jurisdictions or on memory of rates rather than reading the facts, and they skip the preliminary step of classifying the item and establishing the tax base.
What to do instead
Adopt a two-step habit before any computation: (1) classify the item (income head, deductible or non-deductible expense, VATable or exempt supply, capital versus revenue); (2) establish whether the given amount is tax-inclusive and convert using the correct fraction. Show the classification explicitly in the answer so the marker can follow the reasoning, and cite the relevant provision where the syllabus requires it.
Answering a company-law or governance scenario by reciting the general duty ('directors must act in the best interests of the company') without applying it to the specific facts or identifying the remedy and the person who may enforce it.
Candidates treat law questions as recall tests and reproduce statute headings, having never practised issue-spotting on a fact pattern.
What to do instead
Use an IRAC structure for every scenario: identify the Issue, state the Rule with its source, Apply it to the named parties and facts, and Conclude with the consequence (validity of the transaction, liability, remedy, or who may sue). Always name the duty, the breach, and the enforcement route.
Calculating NPV, IRR or WACC with arithmetic that is correct but a discount rate or cash-flow profile that is internally inconsistent - for example, discounting nominal cash flows at a real rate, or including sunk costs and excluding working-capital recovery.
Candidates focus on getting the discounting right and treat the cash-flow forecast as given, so they never interrogate whether the inputs belong together.
What to do instead
Before discounting, list the cash flows and label each as relevant or irrelevant with a reason (sunk, committed, opportunity, incremental working capital, terminal value). Match nominal cash flows to a nominal rate and real to real. For WACC, confirm the cost of equity, cost of debt, tax shield and market-value weights all relate to the same entity and period.
Producing a PESTLE or SWOT analysis as a list of generic bullet points that could apply to any organisation, with no link to the case facts and no conclusion drawn for the decision the question asks about.
Candidates memorise frameworks and treat the model as the answer rather than a tool, often because they have not practised extracting relevant facts from a case.
What to do instead
For each framework point, cite the specific fact from the scenario, state the implication for the organisation, and rank by materiality. Finish with an explicit recommendation that follows from the analysis. If a point cannot be tied to a fact in the case, delete it.
Recommending controls or analytics outputs without specifying the data, the test, the threshold and the follow-up action - for example, 'use data analytics to detect fraud' with no description of what is tested against what.
Candidates understand the purpose of analytics but have not practised translating a risk into a concrete test, and they confuse the tool with the control.
What to do instead
Express every analytics recommendation as: data source, field or relationship tested, the test performed (duplicate check, gap detection, matching, trend, Benford), the exception threshold, and the action taken on exceptions. Link each test back to the specific risk or assertion it addresses.
Failing to identify a threat to independence or objectivity in an ethics scenario, or identifying it but proposing a safeguard that does not address it (for example, 'disclose the threat' when disclosure does not mitigate self-interest).
Candidates can name the categories of threat but have not practised classifying a fact pattern, and they default to disclosure as a universal safeguard.
What to do instead
Work through the five threat types systematically (self-interest, self-review, advocacy, familiarity, intimidation) and state which applies and why. Then choose a safeguard that actually removes or reduces that threat (rotate the engagement partner, remove the individual from the team, independent review, decline the engagement) and state the residual risk if no safeguard is sufficient.
Running out of time because the candidate writes long narrative answers to low-mark questions and then has insufficient time for the high-mark computational or discursive questions, leaving whole sections blank.
Candidates do not allocate time in proportion to marks and are reluctant to leave a familiar question part-finished, so they over-invest in the first section.
What to do instead
At the start, convert marks to minutes (roughly 1.8 minutes per mark) and write the budget beside each question. Answer in mark-earning points - one clear point per mark - and move on when the budget is spent. Attempt every question at least briefly, since a partial answer earns more than a blank.