Mongolian CPA common mistakes
Mongolian CPA · Understand the traps and how to avoid them.
Applying the general corporate income tax rate of 10% to every taxable entity and every income type without first classifying the taxpayer (small/medium enterprise versus large taxpayer) and the income category (main operations versus non-operating income such as interest, dividends, royalties, or capital gains).
Candidates memorize a single headline rate from the Law on Corporate Income Tax and then apply it mechanically. They skip the preliminary classification step because it feels like extra work under time pressure, and the graduated treatment for small and medium enterprises plus the separate withholding regimes on certain income types are exactly the parts that are least practiced.
What to do instead
Before computing any tax figure, write a two-line classification note on scratch paper: entity category, then income category. Compute the tax for each category separately with its own rate and threshold, then sum. Practice with a mixed-income scenario where the same company earns operating profit, interest, and a dividend so the different treatments must be kept apart in the working.
Treating a temporary difference as a permanent difference (or vice versa) when computing deferred tax, and then either recognizing a deferred tax asset/liability that should not exist or omitting one that should, for example on non-deductible fines, entertainment limits, or accelerated tax depreciation.
Both categories look like 'accounting profit differs from taxable profit' and candidates stop at the fact of the difference rather than asking whether it will reverse in future periods. Permanent differences are easier to spot in the tax reconciliation note, so candidates over-apply that label to items that genuinely reverse.
What to do instead
For every difference, ask one question: will the tax effect reverse in a future period? If yes, it is temporary and needs a deferred tax calculation; if never, it is permanent and only affects the current tax charge. Build a two-column table (temporary / permanent) and force each reconciling item into exactly one column before computing any deferred tax balance.
Recording revenue when cash is received or when an invoice is issued, rather than when the performance obligation is satisfied under accrual accounting, and failing to adjust for unearned revenue and accrued revenue at period end.
Candidates carry over cash-basis intuition from everyday bookkeeping and from small-business practice. Exam questions deliberately place the cash receipt and the delivery of goods or services in different periods, so the intuitive answer is wrong exactly when it matters.
What to do instead
Apply a fixed five-step check to every revenue question: identify the contract, identify the performance obligation, determine the transaction price, allocate it, and recognize when the obligation is satisfied. Then post the period-end adjusting entries for unearned and accrued revenue explicitly, even when the question does not ask for journal entries, because the numbers usually depend on them.
Answering audit procedure questions with substantive tests of detail when the risk described is a control risk or a fraud risk that calls for tests of controls, inquiry of management, or a modification of the audit opinion, and failing to link the identified risk to the specific assertion affected.
Candidates learn a catalogue of procedures and select by familiarity rather than by mapping risk to assertion to response. The scenario mentions a weak control or a management override, but the candidate answers with the standard 'confirm balances and vouch invoices' template.
What to do instead
Use a three-column working: risk identified, assertion affected (existence, completeness, valuation, rights, presentation), and responsive procedure. Only after the assertion is named should a procedure be chosen. For fraud and management override scenarios, default to procedures that do not rely on the client's controls, and check whether the question is asking about the audit opinion rather than a test.
Confusing the audit opinion decision with the auditor's reporting of key audit matters or with the going-concern paragraph, for example issuing a qualified opinion for a material but not pervasive misstatement when a disclaimer or an emphasis-of-matter paragraph is the correct treatment.
The materiality and pervasiveness matrix is learned once and then blurred. Candidates remember that 'serious problem means qualified opinion' and do not run the two-axis test, and they conflate the separate sections of the auditor's report because the report structure is memorized as a block rather than as independent decisions.
What to do instead
Draw the two-by-two matrix (material versus not material, pervasive versus not pervasive) on scratch paper for every opinion question and place the scenario in a cell before writing the answer. Separately, list the report sections in order and state which section changes for each scenario, so the opinion decision and the additional communication are kept as two distinct steps.
Selecting a cost allocation base or transfer price by the arithmetic that is easiest to compute rather than by the causal driver, and then failing to reconcile the effect on divisional profit, company-wide profit, and managerial behaviour.
Candidates focus on getting a number and treat allocation bases as interchangeable. The exam tests the consequence, not the arithmetic, so a correct calculation built on the wrong base still loses the marks, and the behavioural effect on the selling division's incentive is usually omitted entirely.
What to do instead
State the causal relationship in one sentence before choosing a base, then compute. For transfer pricing, always answer three questions: what is the range of acceptable prices from the minimum (variable cost plus opportunity cost) to the maximum (external market price), what does the chosen price do to each division's reported profit, and what behaviour does it encourage. Practice with a case where the selling division has spare capacity and one where it does not.
Computing a variance and stopping at the amount, without decomposing it into price and quantity components or explaining whether it is favourable because of a real operational gain or merely because of a budgeting error or a change in the activity level.
Variance analysis is learned as a set of formulas, so candidates produce numbers but not interpretation. Flexible-budget versus static-budget distinctions are skipped, and the candidate does not separate the effect of selling more units from the effect of spending more per unit.
What to do instead
Always restate the actual result against a flexible budget built at actual volume before comparing to the static budget. Then split each variance into its price and quantity parts and write one sentence on the likely cause and the manager who owns it. If the question asks for an explanation and the answer contains only numbers, the answer is incomplete.
Mixing up the discounting convention and the timing of cash flows in capital budgeting, for example discounting a full year for a cash flow that occurs at the start of the period, using the accounting rate of return as if it were the internal rate of return, or ignoring the effect of depreciation tax shields and working capital recovery on the terminal cash flow.
Candidates memorize the net present value formula but not the cash-flow timeline. Depreciation is a non-cash expense, so it is often dropped from the projection entirely, which removes the tax shield, and working capital is treated as a sunk cost rather than a recoverable investment.
What to do instead
Build a timeline table with a column per period before discounting anything, and place each cash flow in the period it actually occurs. Treat depreciation only through its tax effect, and always include the initial working capital outflow and its recovery at the end. Sanity-check the result by comparing net present value and internal rate of return conclusions, since they should agree for conventional projects.
Answering professional ethics scenarios by naming the threat (self-interest, self-review, advocacy, familiarity, intimidation) without applying the conceptual framework: evaluating the significance of the threat and then selecting a safeguard or concluding that the engagement must be declined.
The threat categories are memorable and easy to list, so candidates stop there. The exam marks the evaluation and the response, and it often rewards the answer that says no safeguard is sufficient, which candidates avoid because declining feels like a non-answer.
What to do instead
Structure every ethics answer in three explicit steps: identify the threat and its source, assess whether it is at an acceptable level and why, then state the response (safeguard, additional review, disclosure, or decline). Practice writing the assessment sentence, because that is the step most often missing, and be willing to conclude that the engagement should not be accepted.
Treating internal controls over financial reporting as sufficient IT controls, ignoring general IT controls such as access management, change management, and backup, and failing to consider how an automated control depends on the integrity of the underlying system and data.
Candidates from an accounting background focus on business-process controls and see the IT environment as a black box. The exam tests whether the candidate understands that an automated control can be applied consistently but is only as reliable as the general controls around the system that runs it.
What to do instead
For every automated control, ask what could make it fail silently: unauthorized configuration change, inappropriate access, or incomplete data. Then map each risk to a general IT control (change management, logical access, IT operations). When a question describes a system implementation or a data migration, check completeness and accuracy of the data before relying on any automated control downstream.