ACCA Applied Skills · Performance Management

Analytical Techniques in Budgeting and Forecasting for ACCA PM

Analytical techniques in budgeting and forecasting are quantitative methods that turn past data into future estimates. In PM you use the high-low method, regression, correlation, time series, learning curves, expected values and index numbers. You pick the right method, apply its formula carefully, then state how reliable the forecast is.

What this chapter covers

This chapter gives you the tools to build numbers for a budget. A budget needs a cost estimate, a sales estimate and a view of how much to trust each one. Each technique answers a different question. The high-low method and regression split costs into fixed and variable parts. Correlation tells you how reliable that split is. Time series finds the trend and seasonal pattern in sales. Learning curves predict falling labour time. Expected values deal with uncertainty. Index numbers adjust figures for price changes.

The topics build on each other. High-low is the simple version of regression. Correlation tests regression. Time series also uses a trend line, which is the same idea as regression with time as the x variable. So learn them in order and the later ones feel familiar.

The chapter links to the rest of PM in many places. Cost behaviour feeds marginal costing, break-even and relevant costing. Forecasts feed budgeting and variance analysis. Learning curves feed pricing and make-or-buy decisions. Expected values feed decision-making under risk. The techniques can appear in Section A and Section B objective questions, and in Section C written questions where you must calculate and also comment.

This chapter is calculation-heavy, and objective test questions are marked all or nothing, so a single slip in a formula costs the full mark. Yet the methods are mechanical. If you learn each one and practise it, the marks are very reachable. The techniques also support Section C, where you may need to compute a forecast and then explain its limits. Students who only memorise formulas lose marks on that commentary. Putting in time here pays off across budgeting, costing and decision-making questions, so it is worth real effort.

Analytical techniques in budgeting and forecasting: topics in the order to study them

1High-Low Method and Cost BehaviourStart here. It is the simplest way to split fixed and variable cost, and it sets up the idea of a cost line used in every later topic.
2Linear Regression AnalysisIt does the same job as high-low but uses all the data points, so learn it straight after you understand the cost line y = a + bx.
3Correlation Coefficient and Coefficient of DeterminationThese measure how well the regression line fits, so you need regression first.
4Time Series Analysis and Seasonal VariationsIt uses a trend line and adds seasonal adjustments, so it builds on regression and averages.
5Learning Curve TheoryA separate technique with its own formula. Study it once the line-fitting topics are secure, so you can give it full attention.
6Expected Values and Probability in ForecastingIt covers uncertainty and is quick to learn. It also helps you judge the forecasts from earlier topics.
7Index Numbers and Forecasting TechniquesFinish with index numbers and a comparison of methods. Now you can choose the right technique and adjust figures for price changes.

How to prepare Analytical techniques in budgeting and forecasting

Treat this chapter as a toolkit. For each tool, know when to use it, how to compute it and what it cannot tell you.

Learn each formula from the formula sheet or by working through one example by hand. Know what each symbol means before you practise.
Do two or three short calculations per topic, and check every step. Common slips are using the wrong x and y, or the wrong number of periods.
Practise the Section A style: a single calculation with four answer options. Work out the answer first, then look at the options, so you are not led by them.
Practise Section B cases, where one scenario supports five questions. Often an early answer, such as the variable cost per unit, feeds later ones.
For Section C, practise setting out workings clearly and then writing two or three sentences on reliability and limitations, such as extrapolation beyond the data range.
In the last week, mix topics in one sitting so that you must choose the method yourself, then time yourself on past-style questions.

Common mistakes in Analytical techniques in budgeting and forecasting

Mixing up the fixed and variable parts of the cost line.
Fix: Write y = a + bx at the top of your working and label a and b with their meanings every time.
Using the highest and lowest cost instead of the highest and lowest activity in the high-low method.
Fix: Choose the two points by activity level, then take the costs that go with them.
Reading a high r or r² as proof that x causes y.
Fix: Say that it shows a strong linear relationship in the data, not cause. Comment on sample size and outliers too.
Forecasting well beyond the range of the data and presenting the result as reliable.
Fix: Always add a line on reliability: how far the forecast is from the data range, and whether the pattern is likely to continue.
Applying the learning curve to the wrong quantity, such as treating the cumulative average time as the time for the last unit.
Fix: Read the question for what the percentage applies to. Then compute total time for the cumulative units, and subtract to find the time for later units.
Giving a calculation in Section C with no comment.
Fix: After each calculation, add a short sentence on what the figure means for the budget and what could make it wrong.

Last-day revision: Analytical techniques in budgeting and forecasting

High-low: variable cost per unit = (cost at highest activity − cost at lowest activity) ÷ (highest activity − lowest activity). Then fixed cost = total cost − variable cost.
High-low uses only the highest and lowest activity levels, so it can be distorted by unusual points.
Regression line: y = a + bx, where b is the variable cost per unit (or slope) and a is the fixed cost (or intercept).
The correlation coefficient r runs from −1 to +1. Values near +1 or −1 mean a strong linear relationship. Near 0 means a weak one.
Coefficient of determination = r². It is the share of the variation in y explained by x.
Do not forecast far outside the range of your data. Reliability falls the further you extrapolate.
Time series: trend + seasonal variation (additive model), or trend × seasonal factor (multiplicative model). Seasonal variations should sum to zero in the additive model.
Learning curve: cumulative average time per unit falls by a fixed percentage each time cumulative output doubles. Check which version the question uses.
Expected value = Σ (probability × outcome). Probabilities must add up to 1.
Expected value is a long-run average. It may not match any single outcome.
Index number = (current value ÷ base value) × 100. Use it to convert figures to a common price level.
Objective test answers score all or nothing, so check units, signs and rounding before you submit.

Analytical techniques in budgeting and forecasting practice questions

Analytical techniques in budgeting and forecasting in other exams

The same ground in other exams, if you are preparing for more than one or want another angle on it.

Analytical techniques in budgeting and forecasting: frequently asked questions

Do I need to memorise the regression formulas for PM?

Check the formula sheet provided in the exam and learn what it gives you. You still need to know how to apply each formula and what each symbol means. Practise enough that you can use it quickly and without errors.

What is the difference between the high-low method and regression?

High-low uses only two data points, the highest and lowest activity levels. Regression uses all the points and fits the best straight line through them. Regression is usually more reliable, but high-low is quicker.

How do I know which forecasting technique to use?

Look at the data and the question. Use cost-activity methods for splitting costs, time series for patterns over time with seasons, learning curves for labour time falling with experience, and expected values when outcomes have probabilities. Then say why the method suits the situation.

How are these topics tested in the exam?

They can appear as single calculations in Section A, as linked questions in a Section B case, and in Section C where you calculate and then comment. Objective questions are marked all or nothing, so accuracy matters.