Specimen paper

Al Marfa Facilities Management in Abu Dhabi has revenue of AED 240m, and its auditors set performance materiality at AED 1.2m. Two items reach FP&A in the same month: a AED 900,000 revenue shortfall caused entirely by one client's mobilisation slipping from March into April, and a AED 60,000 overspend in a training cost centre whose annual budget is AED 350,000 and which funds a planned supervisor promotion. Which item should FP&A work on first, and why?

• A

  The revenue shortfall, because it is close to the audit materiality level.

  This imports audit materiality, which measures whether a misstatement could mislead a reader of the accounts, into a management judgement about where an analyst's hours will change an outcome. The two thresholds answer different questions and rarely agree. It would be right only if the shortfall were a possible misstatement in the reported figures rather than a timing difference that April already reverses.

• B

  The revenue shortfall, because the month closed short against phased budget.

  This treats a phasing difference as a performance problem. The month closed short only because the budget phasing put the mobilisation in March; no client was lost, no price moved, and the full year is unchanged. It would be right if the mobilisation had been cancelled or deferred indefinitely, so that the shortfall left the year rather than moving within it.

• C

  The training overspend, because it decides whether the promotion is funded.

  Correct: the smaller number is the one whose resolution changes a decision, and that is what makes it material to FP&A.

• D

  The training overspend, because it breaches the limit the centre was given.

  Right item, wrong reason, which is why it is the most attractive wrong answer. Framing the work as a breached spending limit is the control hat, policing authority, rather than the analysis hat. It would be the right reason if you sat in budgetary control; FP&A takes the item because a promotion decision hangs on it, not because a limit was crossed.

Sahel Facilities Services in Saudi Arabia is rebuilding the plan for its cleaning contracts on drivers. Four managers propose four different builds for the same revenue stream. Which one is genuinely driver-based?

• A

  Last year's revenue, uplifted by the contracted indexation of 4 per cent.

  This mistakes a contractual escalator for a driver. Indexation is a perfectly good input, but it belongs on the charge-out rate, not on the revenue total; applied to the total it means nothing in the model responds when sites are won, lost or rescoped. It would be a driver build only if site count and hours sat underneath it and the 4 per cent moved the rate alone.

• B

  Sites serviced, multiplied by hours per site and the charge-out rate.

  Correct: three observable quantities, each owned by somebody outside finance, multiplying up to money.

• C

  Market growth of 6 per cent, applied to the current contract portfolio.

  This substitutes a borrowed macro percentage for the company's own mechanics. Sahel's revenue comes from the contracts it holds and the hours it works, not from an average across a market it only partly serves, and no operations manager can tell you the 6 per cent is wrong. It would be defensible only if the portfolio tracked the market mechanically, such as a fixed share of an index-linked programme.

• D

  Gross margin of 22 per cent, applied to the revenue the board expects.

  This reverses cause and effect: revenue is derived from a target and a ratio rather than built from activity, so the plan asserts the answer it was meant to test. It also anchors margin as an input when margin is an output of price and cost. It would be the right calculation only if you were solving for the revenue implied by a profit target, which is a gap-closing sum, not a plan.

Marchford Homes approves a GBP 4.8m timber frame factory in November and orders it in January, with commissioning expected on 1 October. The operating plan assumes the additional output from 1 July, and the depreciation charge has been run from the order date. On what date should both the charge and the capacity be built?

• A

  The order date, which is when the company becomes committed to the spend.

  This confuses commitment with an asset in service. Placing the order creates an obligation to pay; it creates no frames and no capacity, and an asset not yet available for use carries no depreciation. The order date is the right date for a different schedule, the capital commitments register, and for nothing in the operating plan.

• B

  The approval date, which is when the board accepted the business case.

  This confuses a decision with an asset, and it is the earliest of the four dates, so it flatters the plan hardest by pulling both benefit and charge furthest forward. Approval starts a project, not a production line. It is the right date only for tracking the capital programme in the board minute.

• C

  The payment date, which is when cash leaves under supplier terms.

  This is the correct date for a different question. Supplier terms govern when cash leaves and belong in the cash and facility forecast, where they matter a great deal. They say nothing about when the factory can make frames or when the accounts begin to carry the charge, and using them here would put the operating plan on a treasury timetable.

• D

  The commissioning date, which governs both the charge and the capacity.

  Correct: depreciation starts when the asset is brought into use, and the extra output arrives on that same date.

Ravenna Packaging Machinery pays its 40 salespeople on booked orders. Bookings beat plan by EUR 6m, the works shipped to plan, order book cover rose from four months to six, and commission at 3% cost an extra EUR 180,000 in a year that produced no additional revenue, gross profit or cash. What is the best structural response?

• A

  Move the sales team onto a scheme paid on recognised revenue.

  This cures the timing by transferring a risk the sales team cannot manage, since recognition depends on the works schedule and the shipment queue rather than on selling. It also destroys the forward demand signal, because nothing is recorded until delivery and the materials plan loses its lead time. It would be right in a short-cycle business where sales controls delivery, such as stock sold from the shelf.

• B

  Cap the bookings credited to each salesperson at the sales plan.

  This suppresses the reported number rather than the payment. Salespeople stop booking above plan or hold orders into next year, so the order book, the capacity plan and the forecast all lose the information the business most needs, and the cash still goes out on whatever is booked. It would be right only if the excess orders were fictitious, which is a data integrity problem, not a scheme design problem.

• C

  Accrue commission on booking, release it on recognition.

  Correct: it keeps the incentive and the forward signal while matching the cash cost to the revenue the order eventually delivers.

• D

  Reduce the commission rate so the extra bookings cost less.

  This prices the behaviour differently without changing it. The same EUR 6m of unwanted cover is still bought, still consumes cash, still lengthens the book, and the business simply pays a little less for the same distortion, while also cutting pay on the orders it does want. It would be right if the diagnosis were that the scheme is too expensive overall, which is not what the facts show.

Ardmore Dairy ships through a national wholesaler. This year sell-in was GBP 24m, sell-through restated at Ardmore's transfer price was GBP 21m, and opening channel stock was GBP 4m. What is closing channel stock, and what does that say about reported revenue?

• A

  GBP 1m: the wholesaler drew stock down as demand outran shipments.

  This reverses the identity, subtracting the movement instead of adding it, so GBP 4m less GBP 3m gives GBP 1m. It describes a destock, the opposite of what happened. It would be right if sell-through had exceeded sell-in, which is the case where the channel is shipping out of its own warehouse and your revenue understates demand.

• B

  GBP 3m: channel stock is the gap between sell-in and sell-through.

  This confuses the movement with the level. GBP 3m is the change in channel stock for the year; the closing balance also contains the GBP 4m the wholesaler already held, which has simply been dropped. It would be right only if the wholesaler had opened the year holding nothing at all.

• C

  GBP 7m: reported revenue holds GBP 3m the market has not absorbed.

  Correct: closing stock is GBP 4m plus GBP 24m less GBP 21m, and the GBP 3m build sits inside the revenue you have already reported.

• D

  GBP 7m: reported revenue understates demand by the GBP 3m stock build.

  The level is right and the direction is backwards, which makes this the most instructive wrong answer in the item. It reads a rising reservoir as suppressed demand, when a warehouse that is filling has taken more than the market bought. Understatement would follow only from a falling stock position, where sell-through exceeds sell-in.

Pennine Toolworks sells 20,000 cutting tools a year at GBP 500 each on a 40% gross margin. The sales director asks whether a 2% price cut or a 2% volume shortfall does more damage to gross profit. Which answer is right?

• A

  The volume shortfall, GBP 200,000 against GBP 80,000, since units carry cost.

  The two figures are correct and attached to the wrong effects. The stated reason is where the reversal happens: units carrying cost is precisely why a lost unit costs you only its margin and therefore does less damage, not more. This pairing would converge only in a business with no variable cost at all, where price and volume drop through identically.

• B

  They are equal at GBP 200,000 each, since both move revenue by the same 2%.

  This reads an equal revenue effect as an equal profit effect, which is the most common instinct in the room and skips the cost side entirely. It would be right only at a 100 per cent gross margin, the special case where no cost travels with the unit, such as pure licence revenue with no cost of sale.

• C

  The price cut, GBP 200,000 against GBP 80,000, since price carries no cost.

  Correct: the full price give-away lands in gross profit, while lost volume costs only the margin it carried.

• D

  The price cut, GBP 80,000 against GBP 32,000, since margin applies to both.

  This gets the ranking right and both numbers wrong by applying gross margin where no cost moves. GBP 80,000 is the price effect with a 40 per cent margin wrongly deducted, and GBP 32,000 applies the margin a second time to a volume figure that had already been margined. It would be right only if cutting price somehow reduced unit cost in proportion, which nothing does.

Grampian Turbines ships 600 units a year and holds a service book of 1,800 contracts at GBP 5,000 each, or GBP 9.0m. The attach rate on newly shipped units falls from 60% to 55%. A manager values the loss at 5 points on 60%, or 8.3% of GBP 9.0m, about GBP 750,000 a year. What is the right figure and why?

• A

  GBP 150,000: the rate applies only to new units, so 30 contracts are lost.

  Correct: five points on the 600 units shipped this year is 30 contracts, worth GBP 150,000 of annual value.

• B

  GBP 750,000: the rate touches the whole book, so 8.3% of it is lost.

  This is the manager's error reproduced in the stem, and it overstates the annual loss five times over. Attach governs the roughly 360 contracts added each year, a fifth of the book, and the other 1,440 were attached at rates set in earlier years. It would be right only if every contract were re-attached annually at the current rate, which would make the book indistinguishable from new business.

• C

  GBP 450,000: renewals carry the new rate too, so 90 contracts are lost.

  This applies the attach rate to the renewal population, conflating two rates with different causes: attach is decided at the point of sale, renewal is decided by service quality, price and competition. It would be right only in a model that deliberately ran a single combined rate across the book, in which case you could no longer tell a selling problem from a service problem.

• D

  GBP 300,000: the retiring cohort carries it too, so 60 contracts go.

  This doubles the correct 30 contracts by charging the new rate to the units leaving the book as well as those joining it. Retiring machines were shipped years ago and were attached, or not attached, under the rate prevailing then. It would be right only if the retiring cohort were itself being re-sold and re-attached this year.

Kesteven Homes, a UK housebuilder, completed 20 per cent more plots than planned on its northern sites. An analyst flexes the entire site cost budget up by 20 per cent, including site insurance, plant depreciation and the regional office recharge. Actual cost then lands 3 per cent below the flexed budget and is reported as a saving. What is wrong with the exercise?

• A

  Fixed costs were flexed with volume, so the site is credited with a saving it did not make

  Correct: insurance, depreciation and a recharge do not rise with completions, so flexing them by 20 per cent creates budget that never existed.

• B

  Variable costs were flexed at plan rates, so real material price movement stays hidden

  This calls the correct method an error. Flexing variable costs at plan rates is exactly how the flex isolates price and usage: the resulting variance is the price movement, so it is revealed rather than hidden. The statement would describe a genuine fault only if the flex had used actual rates, which absorbs the price effect into the budget and leaves nothing to explain.

• C

  The flex used completions rather than revenue, so the base moved by the wrong percentage

  This proposes the wrong driver. Site build cost is caused by activity, so completions are the correct base; flexing on revenue would import selling price and house type mix into a cost comparison and confuse two unrelated questions. Revenue would be the right base only for a cost genuinely driven by sales value, such as agency commission.

• D

  The flex used actual mix rather than plan mix, so efficiency and mix are combined

  This inverts the treatment. Flexing to actual mix is what removes mix from the comparison where cost per plot differs by house type, leaving efficiency behind as the residual. The fault described would arise from the opposite choice, holding plan mix while the actual mix moved, which leaves a mix effect sitting inside the efficiency number.

Kesteven Homes sold 128 plots in the half year against a plan of 120, and every plot was sold at its plan list price. The blended average selling price fell from GBP 250,000 planned to GBP 228,125 actual, and revenue finished GBP 0.8m below plan. The chief executive asks the sales director to explain the discounting. What should FP&A tell him?

• A

  No discount was given: more low-value units sold and fewer high-value ones

  Correct: with every list price held, the fall in the blended average can only be the change in what was sold.

• B

  A discount was given on the larger house types, which the blended figure hides

  This is contradicted directly by the stem, and it comes from the intuition that an average price can only fall if somebody cut a price. It preserves the chief executive's framing instead of correcting it. It would be right if realised prices differed from list, which is the test you run first: compare realised with list by house type before you attribute anything to mix.

• C

  The average fell because volume ran ahead of the plan it was set against

  This mistakes a volume effect for a rate effect. Selling more units at unchanged prices leaves a weighted average price exactly where it was; only the proportions between the types can move it. It would become right only if the extra units had been won by cutting price, which is a price effect appearing alongside volume, not volume itself.

• D

  The average fell because the list prices were set before the year began

  This raises a real but different question, whether the list is competitive against the market. Stale prices show up as pressure on realised price or on volume; they cannot create a gap between plan and actual when the plan was built on those same list prices. It would be right only if the plan had assumed a mid-year uplift that was never taken.

Rheinsteg Software screens variances for commentary at GBP 100,000 and 5 per cent of budget. Hosting cost, on a monthly budget of GBP 700,000, has been adverse by GBP 22,000, then GBP 26,000, then GBP 31,000 in three consecutive months. Each month it falls outside the screen and no commentary is written. Which addition to the screen catches this pattern?

• A

  A persistence rule: same-signed variances across three periods above a floor

  Correct: it tests direction repeated over time rather than size in one period, which is the only dimension on which this pattern is visible.

• B

  A percentage rule applied monthly with the absolute floor removed from the test

  Removing the floor makes small budget lines hypersensitive, so a GBP 2,000 line moving GBP 200 trips a percentage test and the pack fills with trivia. It also fails on these facts: GBP 31,000 on GBP 700,000 is 4.4 per cent and still under the threshold. It would work only if every line in the pack were of similar size, which is never true.

• C

  An absolute rule set lower, so that anything above GBP 20,000 is commented on

  This is the tempting answer because it would catch this line, but it treats the symptom by lowering the bar everywhere. Every other account then generates single-month commentary at GBP 20,000, analyst time goes into writing it and executive attention is spent reading it. It would be right only if GBP 20,000 were genuinely the amount at which a decision changes across the whole pack.

• D

  A named exception rule for hosting, because infrastructure is board-sensitive

  This fixes one account and leaves the identical blind spot on every other line, while inviting a growing list of pet exceptions that no longer describe a rule at all. Board sensitivity is an argument for commentary, not for a screening threshold. It would be right only if hosting were uniquely and permanently the one line where small drift mattered.

Meridian Modular's executive pack shows operating profit GBP 1.2m ahead of budget at the half year. Over the same six months the drawn revolving facility rose from GBP 2m to GBP 9m, work in progress rose GBP 6m as three sites started, and debtor days moved from 45 to 62. Which conclusion is best supported?

• A

  Trading is strong, and the drawdown unwinds as the profit converts to cash.

  This forecasts profit and assumes cash will follow, which is how a growing project business runs out of money while reporting success. Work in progress on three live sites does not unwind; it builds until those sites complete and are handed over. It would be right if the drawdown had funded a one-off timing item with a known reversal date, which is not what a site start is.

• B

  Profit is overstated: work in progress holds costs that should be expensed.

  This reads a normal balance sheet build as an accounting fault. Work in progress on live sites is a properly recognised asset carrying costs with future revenue attached, and nothing in the stem points to a recognition problem. It would be right if the work in progress contained abortive or sunk costs with no revenue to come, in which case the issue is an impairment, not the cash position.

• C

  Profit ahead has been funded by working capital, and the pack carries no cash forecast.

  Correct: the balance sheet has financed the result, and the page that would show whether that is sustainable is missing.

• D

  Debtor days are the issue, and collections should return to the 45 day target.

  This picks one real component and mistakes it for the whole. Seventeen days of debtors is a genuine drag, but the larger movement is the GBP 6m of work in progress, and restoring collections alone would not close a GBP 7m facility drawdown. It would be right if the collections position had deteriorated while the rest of the balance sheet stood still.

A partner at Vistula Analytics judges the sales director's forecast to be EUR 1.5m optimistic. She sends her analysis to the chief financial officer, who raises it at the executive meeting the next morning, where the sales director sees the numbers for the first time. What was the error?

• A

  The analysis should have waited until the forecast was final

  This converts influence into criticism. A challenge that lands after the number is locked cannot change it, so the work becomes a post-mortem and the partner earns a reputation for being right too late to matter. It would be the correct sequencing only if the purpose were to learn from a closed cycle rather than to improve the forecast now in front of the business.

• B

  The partner should have offered no view on another's plan

  This is the reporting-service position that the partnering role exists to leave behind. A partner who may not hold a view on a plan is producing numbers rather than improving decisions, and the EUR 1.5m would simply be discovered later by the outturn. It would be right only if the analysis had no evidential basis, in which case the fault lies in the work, not in holding a view.

• C

  The escalation should have gone to the audit committee, not the CFO

  This misroutes a judgement disagreement into a governance channel. An audit committee exists for matters of financial integrity, control failure and reporting, not for a difference of opinion about how optimistic a forecast is, which is exactly what a CFO is there to resolve. It would be right only if there were evidence of deliberate misstatement rather than optimism.

• D

  The owner should have seen the analysis before anyone else did

  Correct: the sequence, not the analysis or the escalation, is what destroyed the partner's position.