- Overview of the ACPF Domain Structure
- Domains 1-5: Regression and Cause & Effect Models
- Domains 6-11: ARIMA and Box-Jenkins Modeling
- How the 11 Domains Map to Exam 4
- A Domain-by-Domain Study Timeline
- Registration Windows, Fees, and Retake Logistics
- Who Actually Uses These Domains on the Job
- Frequently Asked Questions
- ACPF's 11 domains split evenly between regression modeling (Domains 1-5) and ARIMA/Box-Jenkins modeling (Domains 6-11).
- These domains reflect IBF's Exam 4, Forecast Modeling (Advanced) - one of three exams required for ACPF.
- Each advanced exam has 70 multiple-choice questions, a 2-hour limit, and requires a 70% passing score.
- ACPF requires prior CPF certification plus passing Exams 4, 5, and 6 - this article focuses on the Exam 4 domain content.
Overview of the ACPF Domain Structure
The Advanced Certified Professional Forecaster (ACPF) credential, issued by the Institute of Business Forecasting & Planning (IBF), rests on three advanced exams: Exam 4 (Forecast Modeling, Advanced), Exam 5 (Technology and Worst Practices in Forecasting & Planning), and Exam 6 (The Role of Artificial Intelligence and Predictive Analytics in Demand Planning & Forecasting). The 11 domain lines covered in this guide come directly from the official syllabus topics tested on Exam 4, the quantitative modeling exam that most candidates find the most technically demanding of the three.
Unlike a weighted blueprint with percentage breakdowns per topic, IBF publishes these as unweighted syllabus categories - meaning every domain deserves genuine attention rather than a "skip the small ones" approach. If you haven't already reviewed the prerequisites for sitting this exam, start with our ACPF Requirements guide, since ACPF eligibility depends on holding CPF certification first.
Domains 1-5: Regression and Cause & Effect Models
The first half of the Exam 4 syllabus centers on multiple regression as a causal forecasting tool. Candidates need to move beyond "run the regression" and understand model-building as a disciplined, checkable process.
Domain 1: Regression/Cause & Effect Models
Establishes the conceptual foundation: how independent variables (price, promotions, macroeconomic indicators) drive a dependent variable like demand.
- Distinguishing correlation from causation in forecasting scenarios
Domain 2: Steps to Build a Regression Model
Covers the sequential process - variable selection, data preparation, estimation, and validation - that candidates must apply in scenario-based questions.
- Recognizing correct ordering of model-building steps in multi-choice prompts
Domain 3: Diagnostic Tools for Assessing Regression Models
Focuses on interpreting statistical outputs: R-squared, t-statistics, residual patterns, and multicollinearity indicators.
- Reading diagnostic output to judge whether a model is fit for use
Domain 4: Development and Application of Dummy Variables in Multiple Regression
Tests your ability to encode categorical effects - seasonality, holidays, promotional flags - correctly into a regression equation.
- Avoiding common dummy-variable coding errors that skew coefficients
Domain 5: Improving a Multiple Regression Model
Asks candidates to troubleshoot: adding/removing variables, addressing autocorrelation, or refining specification when diagnostics reveal problems.
- Selecting the appropriate corrective action given a described model flaw
Key Takeaway
Treat Domains 1-5 as one continuous workflow - build, diagnose, fix - rather than five isolated topics. Exam 4 questions often chain these steps together in a single scenario.
Domains 6-11: ARIMA and Box-Jenkins Modeling
The second half of the syllabus shifts to time-series modeling using the ARIMA (Autoregressive Integrated Moving Average) / Box-Jenkins methodology. This is where many candidates preparing via generic statistics courses fall short, because ARIMA reasoning differs meaningfully from regression logic.
Domain 6: ARIMA/Box-Jenkins Modeling
Introduces the theoretical basis of ARIMA - how a time series can be modeled purely from its own past values and errors, without external predictors.
- Understanding the AR, I, and MA components conceptually
Domain 7: Assumptions of ARIMA Modeling
Covers stationarity requirements, why differencing is used, and what happens when assumptions are violated.
- Identifying non-stationary series and appropriate transformations
Domain 8: ARIMA Models vs. Regression Models
A comparative domain - candidates must know when ARIMA outperforms regression (and vice versa) based on data characteristics and forecasting objectives.
- Choosing the right model family for a described business situation
Domain 9: Types of ARIMA Models
Distinguishes seasonal vs. non-seasonal ARIMA, and various order combinations (p, d, q) candidates should be able to recognize from described patterns.
- Matching a data pattern description to a likely ARIMA specification
Domain 10: Steps to Build an ARIMA Model
Mirrors Domain 2 but for time series: identification, estimation, diagnostic checking, and forecasting stages of the Box-Jenkins process.
- Sequencing the model-building stages correctly under exam time pressure
Domain 11: Improving ARIMA Models
Tests troubleshooting skills - reviewing residual autocorrelation (ACF/PACF), adjusting model order, and refining forecasts when initial specification underperforms.
- Interpreting residual diagnostics to justify a model revision
How the 11 Domains Map to Exam 4
Exam 4 (Forecast Modeling, Advanced) consists of 70 multiple-choice questions delivered under a 2-hour time limit, with a required passing score of at least 70%. Because IBF does not publish domain-level weighting, treat all 11 domains as equally testable rather than trying to guess which topics carry more points.
For a broader breakdown of scoring mechanics across all three advanced exams, see our ACPF Passing Score guide. If you're still deciding whether the exam's difficulty matches your current modeling background, our How Hard Is the ACPF Exam guide walks through what makes the regression and ARIMA sections challenging for different types of candidates.
| Attribute | Exam 4 Detail |
|---|---|
| Question count | 70 multiple-choice questions |
| Time limit | 2 hours |
| Passing score | 70% minimum |
| Domain count | 11 unweighted syllabus topics |
| Fee (member / non-member) | $425 / $475 per attempt |
Key Takeaway
Exam 4 is one-third of the ACPF requirement. You'll also need to pass Exam 5 and Exam 6 after completing CPF certification - plan your budget and study time across all three, not just this exam.
A Domain-by-Domain Study Timeline
Because the 11 domains split cleanly into a regression cluster and an ARIMA cluster, a straightforward way to structure preparation is to dedicate distinct weeks to each cluster before mixing them in review.
Regression Foundations (Domains 1-3)
- Study cause-and-effect model theory and the model-building sequence
- Practice reading regression diagnostic output (R-squared, t-stats, residuals)
Regression Refinement (Domains 4-5)
- Work through dummy variable coding exercises for seasonality and promotions
- Practice diagnosing and correcting flawed regression specifications
ARIMA Foundations (Domains 6-8)
- Learn stationarity, differencing, and the AR/I/MA components
- Compare ARIMA vs. regression scenarios until the distinction is automatic
ARIMA Application & Full Review (Domains 9-11)
- Practice identifying model types and building sequences from data descriptions
- Run full-length timed practice sets covering all 11 domains together
For a more detailed week-by-week plan with review checkpoints, see our ACPF Study Guide. If you want a condensed reference to keep nearby during final review, the ACPF Cheat Sheet summarizes the key facts from each domain in one page. You can also reinforce domain-specific question patterns using the timed drills on our ACPF practice test platform.
Registration Windows, Fees, and Retake Logistics
Domain knowledge is only half the equation - scheduling matters too. IBF's online exams run on the second and fourth Friday of each month, excluding U.S. holidays, and registration must be completed at least one week ahead of your chosen date; there is no registration during exam week itself. Candidates may take up to three exams in a single day, but everything must be completed within one calendar year from the date of registration.
Each advanced exam costs $425 for IBF members or $475 for non-members. Taking all three advanced exams (Exams 4, 5, and 6) on a first attempt totals $1,275 for members or $1,425 for non-members - a figure that excludes the prerequisite CPF certification, IBF membership, and any preparation materials. For the complete cost picture including recertification, review our ACPF Certification Cost breakdown.
For exact scheduling logic, blackout patterns, and deadline planning, our ACPF Exam Dates guide covers the registration calendar in more depth.
Who Actually Uses These Domains on the Job
The regression and ARIMA skills tested across these 11 domains map directly onto demand planning, statistical forecasting, and S&OP analyst roles where practitioners build and defend quantitative models rather than relying solely on judgmental overrides. Employers hiring for these positions often look for candidates who can explain why a regression model was rejected in favor of an ARIMA approach, or vice versa - which is exactly the kind of comparative reasoning Domain 8 tests.
If you're evaluating whether this level of technical certification translates into career or compensation impact, our ACPF Salary Guide and Is ACPF Worth It analysis dig into that question in detail. For a look at where ACPF-credentialed professionals typically work, see ACPF Jobs.
New to the credential entirely? Start with What Is ACPF? or our ACPF Certification overview before diving into domain-level study, and check ACPF Pass Rate data to calibrate your expectations going in.
Frequently Asked Questions
IBF publishes these as unweighted syllabus topics rather than a percentage-based blueprint, so candidates should prepare all 11 domains with equal seriousness rather than prioritizing a subset.
Yes. Domains 1-5 cover regression/cause-and-effect modeling and Domains 6-11 cover ARIMA/Box-Jenkins modeling - both halves appear on Exam 4, Forecast Modeling (Advanced).
ACPF requires prior CPF certification plus passing three advanced exams: Exam 4 (Forecast Modeling, Advanced), Exam 5 (Technology and Worst Practices), and Exam 6 (AI and Predictive Analytics in Demand Planning). This guide's 11 domains apply specifically to Exam 4.
Each advanced exam, including Exam 4, requires a minimum score of 70% across 70 multiple-choice questions completed within a 2-hour limit.
No. Online exams run only on the second and fourth Friday of each month (excluding U.S. holidays), and registration must close at least one week before the exam date, with no registration allowed during exam week.