CMA Part 2 common mistakes
CMA Part 2 · Understand the traps and how to avoid them.
Computing a ratio correctly and then declaring the trend 'favorable' or 'unfavorable' without stating any comparison base - the candidate reports that, say, a receivables turnover of 6.1 'is good' rather than comparing it to the prior year, a competitor, or the industry and linking it to one underlying business driver.
Candidates memorize the ratio formulas from a list and rehearse the arithmetic, but the exam tests the analytical layer: direction of change, benchmark, and cause. Under time pressure the formula is the only thing that feels 'checkable', so the judgment step is skipped.
What to do instead
Write every ratio answer as a three-part sentence before looking at the options: (1) direction and size of change versus a named benchmark, (2) the business driver that would cause it (pricing, collection policy, mix, leverage), (3) the consequence for liquidity, solvency, or profitability. Do at least 20 practice items where the arithmetic is trivial and the whole task is the interpretation sentence.
Seeing ROE rise and attributing it to improved operating performance, when the DuPont decomposition shows the entire increase came from the equity multiplier (higher leverage) while margin and asset turnover were flat or declining.
Candidates compute the three or five DuPont components once as a memorized formula set but never re-run them side by side for two periods. A single blended ROE number is easier to compare than a stack of components, so the leverage effect is invisible.
What to do instead
Build a two-column DuPont table (prior year vs current year) every time a question mentions ROE, and force yourself to write which component moved most. If the equity multiplier is the mover, the answer is a financing decision, not an operating one. Also confirm whether the question's data supplies average or ending balances and use one basis consistently.
Computing WACC using the coupon rate on existing bonds as the cost of debt, using face-value or book-value weights, and/or forgetting to multiply the debt cost by (1 - tax rate).
The coupon rate is the most visually prominent interest number in a bond footnote, and book weights sit directly on the balance sheet, so both are grabbed first. The tax adjustment is a one-line concept that is easy to learn and easy to drop in a multi-step calculation.
What to do instead
Adopt a fixed four-line WACC template and follow it mechanically: (1) after-tax cost of debt = yield to maturity x (1 - t), never coupon; (2) cost of equity from the stated model; (3) weights at market value (use book only if the question explicitly says market values are unavailable); (4) sum. Write the yield-to-maturity step even when the coupon equals the yield, so the habit survives.
In a CAPM question, plugging the overall market return (Rm) into the formula instead of the market risk premium (Rm - Rf), and/or using a comparable company's levered beta to value a target with a different capital structure without relevering it.
The formula is remembered as a short string with a single 'return' term, so Rm and the premium collapse into one idea. Beta relevering is a two-step mechanic (unlever, then relever) that candidates learn conceptually but rarely drill, so the second step is dropped.
What to do instead
Write CAPM as Ke = Rf + beta x (Rm - Rf) and physically compute the parenthetical difference as its own line before multiplying. For pure-play or comparable betas, always show both steps: unlever the comparable beta at its own debt/equity and tax rate, then relever at the target's. If the question gives only one beta and one capital structure, state that no relevering is required.
In a special-order, make-or-buy, or drop-a-segment question, adding the allocated fixed overhead into the 'avoidable cost' total, or netting a sunk cost such as already-purchased raw material or prior R&D against the incremental cash flows.
Allocated overhead appears in the per-unit cost build-up that candidates are shown, so it reads as a real unit cost. Sunk costs feel relevant because they were actually incurred and the numbers are on the page. The relevance test is a discipline that has to be applied deliberately, not a calculation that announces itself.
What to do instead
Before computing anything, sort each cost line into two columns: 'differs between alternatives' and 'does not differ'. Only the first column enters the answer. Apply three named filters explicitly: sunk (exclude), allocated/committed fixed cost that continues regardless (exclude), opportunity cost or lost contribution from displaced activity (include). Write 'incremental' next to every figure in your working.
Treating a step-fixed cost as purely variable (or vice versa) in a CVP or break-even problem, and mixing a per-unit contribution margin with a total fixed cost without converting units, yielding a break-even point off by an order of magnitude.
The cost behaviour assumption is embedded in a narrative rather than stated as a formula, so candidates jump to the standard break-even formula and force whatever numbers appear into it. The unit/scale mismatch is arithmetic, but it originates in not pausing to classify cost behaviour first.
What to do instead
Classify every cost line in the stem as variable, fixed, or step-fixed before selecting any formula, and note the relevant range. Then write the target variable with units in the label (e.g., 'units', 'dollars of revenue') and check the magnitude of the result against the scale of the business described; an answer implying 40 units or 4 million units when the company sells thousands is a signal to recheck the classification.
In an FX or commodity exposure question, choosing a hedging instrument without first naming the exposure type, so a transaction exposure is hedged with a translation-style measure, or the candidate concludes a forward contract 'eliminates' the risk rather than converting it into a known amount with counterparty and opportunity risk remaining.
Hedging tools are memorable (forwards, futures, options, swaps) while the exposure taxonomy (transaction, translation, economic) is abstract. The exam often omits the label and describes the situation, so the taxonomy step has to be performed rather than read.
What to do instead
Force a one-line diagnosis before looking at choices: 'This is a [transaction/translation/economic] exposure because the cash flow/accounting translation/competitive position is affected.' Then select the instrument and state what residual risk remains. Separately drill the transfer-versus-retain distinction (insurance, hedging, and outsourcing transfer; retention, self-insurance, and deductibles retain) so the 'eliminates risk' trap is recognized on sight.
Mismatching the discount rate with the cash flows - discounting nominal (inflated) cash flows at a real rate or real cash flows at a nominal rate - and omitting tax on salvage value or failing to recover working capital in the terminal year.
Inflation adjustments and terminal-year items are learned as separate topics, so they are applied only when a question announces them. Salvage value is repeatedly treated as a pre-tax inflow because the sale price is the number presented, and working capital looks like a non-cash item that never returns.
What to do instead
Build every capital budgeting answer on a four-line timeline: operating cash flow, tax shield on depreciation, terminal flows (after-tax salvage plus working capital recovery), initial outlay. Add a rule at the top of your page: 'nominal flows with nominal rate, or real flows with real rate - never mixed.' Compute salvage tax as (proceeds - book value) x tax rate every time, including when book value is zero.
Ranking mutually exclusive projects by IRR, or by NPV without noticing that the projects differ in scale or life, and accepting a high-IRR small project over a lower-IRR project that adds more absolute value.
IRR is expressed as a single intuitive percentage that invites ranking, and it is the figure most often quoted in the stem. NPV is presented as a raw dollar amount that feels less comparable across a small and a large project, so the correct decision rule is overridden by the more legible number.
What to do instead
State the decision rule before computing: for mutually exclusive projects use NPV; use IRR only for accept/reject on independent projects, and never when cash flow timing differs. When the two conflict, compute the incremental IRR or simply note that NPV was already the value-maximizing signal. For unequal lives, compare equivalent annual annuity rather than raw NPV, and say which metric the question is asking for.
Answering an ethics item from personal judgment about who was 'right', or naming a generic control ('report it') instead of the specific IMA Statement standard violated and the prescribed sequence of escalation steps.
Ethics scenarios read like opinion questions and candidates respond in the register of a workplace advice column. The standards (competence, confidentiality, integrity, credibility) and the resolution order are memorized as a list but not mapped onto scenario facts, so the option that sounds most virtuous wins over the option that matches the framework.
What to do instead
For every ethics item, first name the standard implicated and the specific sub-clause behaviour (for example, credibility - communicating information fairly and objectively), then apply the escalation sequence in order: follow the organization's existing policy, discuss with the immediate supervisor (unless implicated), then go to the next level or an objective adviser, then consider legal counsel or the IMA ethics helpline. Eliminate any option that skips a step or that recommends resigning before the internal chain has been used.