Welcome to Cost Estimating!

Hello! Welcome to one of the most practical parts of your BA2 – Fundamentals of Management Accounting studies. In this chapter, we are going to learn how to play "cost detective."

Have you ever wondered how a business predicts its electricity bill or its factory costs for next month? They don't just guess! They use historical data to figure out how costs behave. By the end of these notes, you will be able to split a "messy" mixed cost into its fixed and variable parts using three different techniques. Don't worry if math isn't your favorite subject—we will break it down step-by-step!

The Golden Rule: The Cost Equation

Before we dive into the methods, we need to understand the "language" of cost estimation. Every cost can be represented by a simple linear equation:

\( y = a + bx \)

Think of this like a mobile phone bill from ten years ago:

  • \( y \) = Total Cost: The total amount you pay at the end of the month.
  • \( a \) = Fixed Cost: The "line rental" you pay even if you make zero calls. On a graph, this is where the line hits the vertical axis (the intercept).
  • \( b \) = Variable Cost per unit: The cost per minute of a call. This is the "slope" or gradient of the line.
  • \( x \) = Activity Level: The number of minutes you talked (the number of units produced).

Quick Review: Our goal in this chapter is always to find 'a' and 'b' so we can predict 'y' for any 'x'.

Method 1: The High-Low Method

The High-Low method is the simplest way to estimate costs. It assumes that the change in total cost between the highest and lowest activity levels is purely due to variable costs.

Step-by-Step Process

  1. Identify: Find the highest activity level and the lowest activity level from your data. (Warning: Always pick based on activity/units, not the cost amount!)
  2. Calculate Variable Cost (\( b \)): Use the formula:
    \( b = \frac{\text{Change in Cost}}{\text{Change in Activity}} \)
  3. Calculate Total Fixed Cost (\( a \)): Use either the high or low point:
    \( a = \text{Total Cost} - (\text{Variable Cost per unit} \times \text{Activity Level}) \)

Example

Units: 1,000 (Cost: \$5,000) | Units: 3,000 (Cost: \$9,000)

Step 1: Change in Cost = \( \$9,000 - \$5,000 = \$4,000 \). Change in Units = \( 3,000 - 1,000 = 2,000 \).
\nStep 2: \( b = \$4,000 / 2,000 = \$2 \) per unit.
\nStep 3: Fixed Cost \( a = \$9,000 - (\$2 \times 3,000) = \$3,000 \).

Common Pitfalls to Avoid

Don't get tricked! Sometimes the "Highest Cost" isn't the "Highest Activity." Always look at the units/volume first to select your two points.

Key Takeaway: High-Low is fast and easy, but it can be inaccurate because it ignores all the data points in the middle and can be distorted by "outliers" (unusual months).

Method 2: The Graphical Method (Scatter Graphs)

If you prefer seeing things visually, this is for you! We plot historical costs on a graph where the horizontal axis (x) is activity and the vertical axis (y) is cost.

How it works

1. Plot all your data points as dots on the graph.
2. Draw a "Line of Best Fit" by eye through the middle of the dots.
3. The point where your line crosses the vertical axis (y-axis) is your Fixed Cost (\( a \)).
4. The steepness of the line represents your Variable Cost per unit (\( b \)).

Analogy: The Starry Night

Imagine the data points are stars in the sky. You are drawing a straight path through the "center" of the galaxy. It won't hit every star, but it shows the general direction they are moving.

Did you know? This method is very subjective. Two different students could draw two different lines for the same data! This is why management accountants often prefer mathematical methods.

Key Takeaway: Scatter graphs are great for spotting outliers (weird data points that don't fit the pattern) but aren't precise enough for detailed budgeting.

Method 3: Linear Regression (Least Squares)

Don't let the name scare you! Linear Regression is just a fancy mathematical way of drawing the perfect line of best fit. It uses every single piece of data available to be as accurate as possible.

The Logic

Regression minimizes the distance between the actual data points and the line we calculate. In your CIMA BA2 exam, you usually won't have to calculate the long-form equations for \( a \) and \( b \) from scratch, but you must understand how to interpret the results.

Measuring Reliability

How do we know if our cost estimate is actually good? We use two statistical measures:

  1. Correlation Coefficient (\( r \)): Measures the strength and direction of the relationship between cost and activity.
    • \( +1 \): Perfect positive correlation (as units go up, cost goes up).
    • \( 0 \): No relationship at all (the data is random).
    • \( -1 \): Perfect negative correlation (as units go up, cost goes down—rare in costing!).
  2. Coefficient of Determination (\( r^2 \)): This is a very important term! It tells us how much of the change in cost is explained by the change in activity.
    • Example: If \( r^2 = 0.85 \), it means 85% of the change in cost is due to production volume. The other 15% is due to "noise" or other factors (like inflation).
Memory Aid: \( r \) vs \( r^2 \)

Think of \( r \) as the Relationship (is it there?).
Think of \( r^2 \) as the Reason (how much of the cost is explained by this reason?).

Key Takeaway: Regression is the most accurate method because it uses all data points, but it assumes the past will repeat itself exactly in the future.

Comparing the Methods: Which one to use?

Choosing a method is a balance between speed and accuracy.

  • High-Low: Quick, but uses only two data points. Good for a "back of the envelope" calculation.
  • Graphical: Good for a visual check and spotting errors, but too subjective.
  • Regression: Most accurate and uses all data, but more complex to calculate.

Quick Review Box

- Variable Cost (\( b \)): The cost that changes per unit.
- Fixed Cost (\( a \)): The cost that stays the same in total.
- Equation: \( y = a + bx \)
- \( r^2 \): Tells us how reliable our cost model is (aim for close to 1.0!).

Don't worry if these formulas seem heavy at first. The more you practice the High-Low calculations and interpreting \( r^2 \), the more natural it will feel. You've got this!