10 exam-style questions with answers and explanations, straight from our 1,030-question bank. Tap an answer to check yourself. When you're ready, take the scored version in the free practice test.
The ACPF exam has 70 questions and runs 2 hours.
These 10 free ACPF questions are organized by exam domain, so you can see how each part of the Advanced Certified Professional Forecaster blueprint is tested. Reveal the answer and explanation under each question.
A retailer forecasts weekly demand with Q = 520 - 20P + 150D - 10PD, where P is price in dollars and D is 1 for a display promotion and 0 otherwise. Next week's price is $9, and the forecast has been run with D = 1. Merchandising proposes adding another 150 units for the display. How should the forecaster interpret the modeled uplift?
Correct answer: D - 60 units of uplift, already included through the display and interaction terms.
Weekly appliance demand is modeled as y_t = b0 + b1P_t + n_t, with n_t = 0.65n_(t-1) + epsilon_t. Price P_t is known before forecasting. The fitted n_t series remains autocorrelated, but the innovation residuals epsilon_t show no material autocorrelation. Rolling-origin accuracy and interval coverage meet acceptance criteria. A reviewer objects that n_t is not white noise. Select the technically sound response.
Correct answer: B - Retain the model: dependence in n_t is modeled; the innovation residuals are the relevant check.
A nonseasonal demand series has an unstable mean in levels. One ordinary difference produces a series with stable mean and variance. A second difference increases variance, introduces a pronounced negative lag-1 autocorrelation, and worsens rolling-origin forecast error at the operating horizon. Beyond lag 1, the twice-differenced series has little autocorrelation. The next model revision should:
Correct answer: B - Return to one ordinary difference and identify suitable ARMA terms on that series.
A regression review finds that paid-search and social-media spending move almost together, with variance inflation factors of 18 and 21. Individual coefficients change sign across fitting windows, yet their combined contribution and held-out forecasts remain stable. Asked to allocate marketing budget using the separate coefficients, the forecaster should explain that:
Correct answer: D - Correlated spending can yield stable forecasts but unstable separate coefficients; those coefficients do not establish causal returns for budget allocation.
A consumer-goods company evaluates planner adjustments using forecasts frozen four weeks before demand occurs. Over two complete years, adjustments for verified promotions missing from the model improve MAE in every quarter. Routine adjustments made without new information worsen MAE in every quarter, although the final process still beats a seasonal-naive forecast. Which policy is supported by this review?
Correct answer: B - Keep the evidence-backed promotion adjustments and discontinue the routine adjustment step.
Minutes before a medical-oxygen distributor releases today's dispatch plan, a new interface is found to mix pallet and cylinder quantities under one unit label. Only one depot is affected, and physical inventory records remain correct. The approved continuity procedure permits the previous validated interface after quantities are reconciled. Hospitals need today's scheduled deliveries. What should the planning lead direct now?
Correct answer: C - Block the affected export, reconcile through the validated interface, and continue verified dispatches.
The S&OP team has three figures for next month: expected demand, 12,400 units; sales target, 14,000 units; and maximum feasible shipments, 11,200 units. No additional supply is available. To distinguish the market estimate from the supply commitment, the demand-forecast and shipment-plan columns should contain:
Correct answer: D - Demand forecast: 12,400 units; shipment plan: 11,200 units.
A K-means segmentation changes markedly when annual revenue is converted from dollars to thousands of dollars; zero-demand fractions are unchanged. Repeated initializations give consistent results within each scale. Revenue and intermittency are intended to have comparable influence. What adjustment addresses the currency-dependent grouping?
Correct answer: A - Standardize the features so their dispersions have comparable influence on distance.
On a supplier lane, 4% of shipments arrive late. An alert is triggered for 80% of late shipments and 10% of on-time shipments. These rates describe shipments managed without expediting. A planner says that an alerted shipment therefore has an 80% chance of arriving late. For a shipment selected only because it triggered the alert, what probability should replace that estimate?
Correct answer: A - 25%
A retailer's Monday forecasts use the previous month's consumer-confidence index. A gradient-boosted model beats the incumbent in chronological validation, with preprocessing fitted separately inside each training fold. However, the historical features contain today's revised index values; initial values were released only midway through the following month. Before accepting the reported improvement, what must the team change?
Correct answer: C - Reconstruct the index information available at each forecast date and repeat the validation.
The ACPF exam also covers these domains. Drill them in the full free practice test:
The full bank has 1,020 more ACPF questions with explanations.
Continue in the free practice test →
View plans