CPA Australia Taxation common mistakes
CPA Australia Taxation · Understand the traps and how to avoid them.
Treating the taxable income calculation as a single flat sum of assessable income minus all deductions, without separating the entity's ordinary income, statutory income, and the specific deduction provisions that apply to each item.
Candidates memorise the 'assessable income minus deductions' formula and then apply it mechanically to every fact pattern, forgetting that some amounts are non-assessable, some are exempt, and some deductions are quarantined to particular classes of income. The result is a plausible-looking number built on a wrong structure.
What to do instead
Before calculating, build a two-column working sheet: on the left list each receipt and tag it ordinary income (s 6-5), statutory income (s 6-10), exempt, or non-assessable non-exempt; on the right list each outgoing and tag it general deduction (s 8-1) or specific deduction, noting any positive-limb or negative-limb limitation. Only then total. Practise this tagging on five past-style scenarios until the structure becomes automatic.
Applying the wrong tax rate or threshold to a company, trust, or superannuation fund because the candidate assumed the individual marginal rate scale applies, or used the base entity rate without checking whether the entity qualifies as a small business entity.
Most study material is written around individuals, so candidates default to the individual scale. Company, trust, and fund rates, the base rate entity test, and the small business income tax offset each have their own conditions that are easy to skip when the question does not spell them out.
What to do instead
Memorise a one-page rate card: resident individual scale, non-resident scale, company base rate and standard rate with the base rate entity turnover and passive income tests, trustee rates, and fund rates. For every question, first write the entity type and residency at the top of your working, then select the rate from the card. If a rate depends on a test, write the test and your conclusion explicitly.
Calculating capital gains on the disposal of an asset by subtracting cost base from proceeds and applying the 50% discount without first checking whether the discount percentage, the CGT event, or the asset's CGT status is available.
The discount is the most memorable feature of the CGT rules, so candidates reach for it immediately. They then miss that non-residents, companies, and certain assets do not get the discount, that the discount applies to the gross gain not the net, and that some events produce a gain by a different formula.
What to do instead
Work a fixed sequence: identify the CGT event and its date; confirm the asset is not exempt or pre-CGT; compute the gross capital gain or loss using the correct event formula; apply losses in the required order; only then apply the discount if the entity and asset qualify. Write the discount eligibility test as a yes/no line in your working every time.
Ignoring the ordering rules for applying capital losses, revenue losses, and the CGT discount, and netting them in whatever order makes the arithmetic easiest.
The arithmetic produces the same-looking answer under several orderings, so candidates do not notice they have breached the statutory order. Under exam time pressure, the easy order wins over the correct order.
What to do instead
Memorise and always write out the order: current-year capital losses offset current-year capital gains before the discount; prior-year losses are applied after the discount to the discounted gain; revenue losses cannot offset capital gains. Practise three scenarios where the ordering changes the final tax payable so you can see why the rule matters.
Treating every payment to a worker as salary and wages and applying PAYG withholding, without testing whether the arrangement is an employee or contractor relationship or whether the payment is a fringe benefit.
Candidates anchor on the label the parties use in the contract rather than the substance of the relationship, and they conflate the PAYG withholding, superannuation guarantee, and fringe benefits regimes, which are tested separately.
What to do instead
For any worker payment, run a three-step test in writing: (1) is the person an employee at common law, considering control, integration, and the ability to delegate; (2) if yes, does PAYG withholding apply and is superannuation guarantee payable; (3) if a benefit is provided, is it excluded from the fringe benefits definition and, if not, which valuation rule applies. State each conclusion before computing.
Claiming input tax credits on acquisitions without checking whether the entity is registered, whether the acquisition is a creditable purpose, and whether a Div 69 or Div 72 adjustment is needed for private or non-deductible use.
The mechanical 'GST paid times the tax fraction' calculation is easy to perform, so candidates stop there. They forget that registration, creditable purpose, and the extent of creditable purpose are separate gates, and that adjustments can reverse credits already claimed.
What to do instead
Build a GST decision tree: registered or required to be registered; is the supply a taxable supply or input-taxed; is the acquisition for a creditable purpose and to what extent; is there a decreasing or increasing adjustment. For mixed-use items, compute the creditable purpose percentage explicitly and show the adjustment separately from the credit.
Preparing a return position without considering the anti-avoidance and integrity provisions, such as the general anti-avoidance rule, the alienation of personal services income rules, or the trust distribution integrity measures, when the facts show a structure rather than a straightforward transaction.
Candidates focus on the primary calculation and treat integrity provisions as an optional add-on. Exam questions often plant a structure precisely to test whether you spot the integrity issue, and the marks sit with that recognition rather than the arithmetic.
What to do instead
After computing the primary result, always run a two-minute integrity scan: does the arrangement have a dominant purpose of obtaining a tax benefit, is personal services income being alienated, is a trust distribution to a low-rate beneficiary being used, is there a non-arm's length dealing. Write one sentence per provision you considered, even if the conclusion is that it does not apply.
Misapplying the residency tests by relying on the number of days present in Australia alone, without working through the domicile test, the 183-day test, and the Commonwealth superannuation test in the statutory order.
The 183-day figure is widely quoted and easy to remember, so candidates treat it as the whole test. They then miss that a resident with a permanent place of abode in Australia can be a resident despite fewer days, and that the tests are alternatives rather than cumulative.
What to do instead
Memorise the four tests and the order in which they are applied, and for each fact pattern write a short paragraph per test with a yes/no and a reason. Pay particular attention to the meaning of 'permanent place of abode' and to the effect of dual residency and treaty tie-breaker rules. Do at least ten short residency fact patterns to build pattern recognition.
Leaving calculations unlabelled and unsupported, so that a correct final figure cannot earn method marks because the marker cannot see which provision or rate produced each line.
Candidates write answers as a single running total in the exam booklet, assuming the marker will follow the logic. In a taxation exam, marks are commonly allocated to the identification of the provision, the treatment decision, and the calculation, and an unlabelled number captures only the last of these.
What to do instead
Adopt a fixed layout for every calculation: a line reference, the amount, the provision or rule relied on, and a short reason. Show workings even where the final figure is uncertain, and state assumptions where the facts are ambiguous. Rehearse this layout under timed conditions so it becomes the default rather than an afterthought.
Running out of time on the long calculation questions and leaving the short technical and compliance questions unanswered, because the candidate allocated time in question order rather than by marks per minute.
Long calculation questions feel productive and absorb time disproportionately, while short definitional or compliance questions look trivial and get deferred. Candidates then discover the easy marks are gone and the long question was only partly completed anyway.
What to do instead
Before starting, scan the paper and write a time budget beside each question based on marks, allowing roughly one minute per mark plus reading time. Answer the short high-certainty questions first to bank marks, then tackle calculations with the remaining time. If a calculation stalls, write the method and the provisions you would apply, then move on, because method marks are usually available without the final number.