2.3 The High-Low Method and Cost Prediction

In business operations, many overhead expenditures—such as plant maintenance, utility consumption, equipment servicing, and fleet transport—are semi-variable. To prepare operational budgets, determine contribution margins, and forecast total production expenditures at anticipated activity volumes, management accountants must decompose mixed costs into their distinct fixed and variable elements.

The high-low method is the standard mathematical technique taught and applied for separating semi-variable costs when historical volume and cost data are available.

---

Purpose and Principles of Cost Estimation

The fundamental premise of the high-low method rests on the linear cost equation:

$$y = a + bx$$

Where:

• $y = \text{Total mixed cost}$
• $a = \text{Total fixed cost}$
• $b = \text{Variable cost per unit of activity}$
• $x = \text{Level of activity}$

Because total fixed costs ($a$) are assumed to remain constant between different activity levels within the relevant range, any change in total cost between two operating periods is assumed to be caused entirely by the variable cost associated with the difference in activity volume.

---

The Step-by-Step High-Low Methodology

Applying the high-low method involves a disciplined, multi-step calculation procedure:

Step 1: Identify Extreme Activity Levels

Review the historical operational records across all available observation periods. Select the period with the highest activity level ($x_{\text{high}}$) and the period with the lowest activity level ($x_{\text{low}}$), along with their corresponding total costs ($y_{\text{high}}$ and $y_{\text{low}}$).

Step 2: Calculate Variable Cost Per Unit ($b$)

Compute the difference in cost ($\Delta y$) and the difference in activity ($\Delta x$) between the high and low points. The variable cost per unit ($b$) represents the slope of the cost line:

$$b = \frac{\text{Cost at highest activity} - \text{Cost at lowest activity}}{\text{Highest activity} - \text{Lowest activity}} = \frac{y_{\text{high}} - y_{\text{low}}}{x_{\text{high}} - x_{\text{low}}} = \frac{\Delta y}{\Delta x}$$

Step 3: Calculate Total Fixed Costs ($a$)

Substitute the calculated variable cost per unit ($b$) into either the highest or lowest activity data point and rearrange the equation to solve for $a$:

$$a = y_{\text{high}} - (b \times x_{\text{high}})$$ $$\text{or}$$ $$a = y_{\text{low}} - (b \times x_{\text{low}})$$

Both substitutions must yield the identical fixed cost value, providing an immediate arithmetic reconciliation check.

Step 4: Formulate the Forecasting Equation

Express the final cost function as $y = a + bx$. Management can now predict total cost for any desired future activity level ($x_{\text{target}}$) falling within the relevant range.

---

Comprehensive Standard Worked Example

Starlight Manufacturing records the following machine hours and total plant maintenance expenditures over a six-month period:

Calculation Walkthrough:

1. Identify High and Low Activity Points:

   • Highest Activity (May): $x_{\text{high}} = 7,500 \text{ hours}$; $y_{\text{high}} = \text{\textdollar}55,500$
   • Lowest Activity (March): $x_{\text{low}} = 3,500 \text{ hours}$; $y_{\text{low}} = \text{\textdollar}35,500$

2. Calculate Activity and Cost Differences:

   • $\Delta x = 7,500 - 3,500 = 4,000 \text{ machine hours}$
   • $\Delta y = \text{\textdollar}55,500 - \text{\textdollar}35,500 = \text{\textdollar}20,000$

3. Compute Variable Cost Per Machine Hour ($b$):

$$b = \frac{\text{\textdollar}20,000}{4,000 \text{ hours}} = \text{\textdollar}5.00 \text{ per machine hour}$$

1. Compute Total Fixed Cost ($a$):
   • Using the high point:

$$a = \text{\textdollar}55,500 - (\text{\textdollar}5.00 \times 7,500) = \text{\textdollar}55,500 - \text{\textdollar}37,500 = \text{\textdollar}18,000$$

• Validating at the low point:

$$a = \text{\textdollar}35,500 - (\text{\textdollar}5.00 \times 3,500) = \text{\textdollar}35,500 - \text{\textdollar}17,500 = \text{\textdollar}18,000$$

1. Predict Cost at Future Activity: The resulting cost formula is: $\text{Total Cost} = \text{\textdollar}18,000 + (\text{\textdollar}5.00 \times \text{Machine Hours})$. If Starlight plans to operate 8,000 machine hours in July (within the relevant range):

$$\text{Forecasted Maintenance Cost} = \text{\textdollar}18,000 + (\text{\textdollar}5.00 \times 8,000) = \text{\textdollar}18,000 + \text{\textdollar}40,000 = \text{\textdollar}58,000$$

---

Advanced High-Low: Stepped Fixed Cost Adjustments

In practical operations, fixed costs do not always remain flat across wide swings in volume. If an increase in activity crosses a capacity threshold, a stepped fixed cost will occur. When fixed costs differ between the highest and lowest activity points, applying the standard high-low formula without adjustment will distort the variable cost rate.

The Adjustment Technique

If fixed costs increase by a known stepped amount at higher activity volumes:

1. Either deduct the stepped fixed cost increase from the cost at the highest activity point before computing the cost difference;
2. Or add the stepped fixed cost increase to the lowest activity cost.

This adjustment isolates the change in cost attributable purely to variable operational activity.

Step-Cost Worked Numerical Demonstration

Highland Engineering records the following quarterly output and total production overheads:

• Lowest Activity (Quarter 1): 4,000 units produced; Total Cost = $47,000
• Highest Activity (Quarter 4): 9,000 units produced; Total Cost = $92,000

Condition: Plant records indicate that because production exceeded 6,500 units in Quarter 4, an additional supervisor was employed, stepping up total fixed overheads by $5,000.

Step-by-Step Solution:

1. Adjust the High Activity Cost for the Step:

$$\text{Adjusted High Cost} = \text{\textdollar}92,000 - \text{\textdollar}5,000 = \text{\textdollar}87,000$$

(This normalizes the high cost to the same fixed-cost baseline as the low activity period).

1. Calculate Variable Cost Per Unit ($b$):

$$b = \frac{\text{Adjusted High Cost} - \text{Low Cost}}{\text{High Activity} - \text{Low Activity}} = \frac{\text{\textdollar}87,000 - \text{\textdollar}47,000}{9,000 - 4,000} = \frac{\text{\textdollar}40,000}{5,000 \text{ units}} = \text{\textdollar}8.00 \text{ per unit}$$

1. Calculate Baseline Fixed Cost ($a_{\text{base}}$):
   • Using the low activity point (Quarter 1, where the step has not occurred):

$$a_{\text{base}} = \text{\textdollar}47,000 - (\text{\textdollar}8.00 \times 4,000) = \text{\textdollar}47,000 - \text{\textdollar}32,000 = \text{\textdollar}15,000$$

1. Define the Segmented Cost Functions:

   • For activity up to 6,500 units: $\text{Total Cost} = \text{\textdollar}15,000 + (\text{\textdollar}8.00 \times x)$
   • For activity exceeding 6,500 units: $\text{Total Cost} = (\text{\textdollar}15,000 + \text{\textdollar}5,000) + (\text{\textdollar}8.00 \times x) = \text{\textdollar}20,000 + (\text{\textdollar}8.00 \times x)$

2. Forecasting Example: If budgeted production for next quarter is 8,500 units:

$$\text{Forecasted Cost} = \text{\textdollar}20,000 + (\text{\textdollar}8.00 \times 8,500) = \text{\textdollar}20,000 + \text{\textdollar}68,000 = \text{\textdollar}88,000$$

---

Advanced High-Low: Adjusting for Price Inflation and Wage Shifts

When observation periods span several quarters or years, general price inflation, utility tariff adjustments, or statutory wage increases can inflate costs in later periods. If historical costs reflect different price levels, applying the high-low formula directly will falsely inflate the variable cost slope.

Indexation Methodology

To correct for price changes:

1. Obtain the relevant cost price index for each period.
2. Convert all historical costs to a common price level (typically current period prices) prior to applying high-low:

$$\text{Standardized Cost} = \text{Historical Cost} \times \frac{\text{Base Price Index}}{\text{Period Price Index}}$$

1. Apply the standard high-low calculation using the standardized costs.
2. Once the linear equation is derived, re-inflate the variable and fixed components to the expected price level of the future budgeting period.

---

Strategic Evaluation: Strengths and Technical Limitations

Management accountants must critically appraise the high-low method when selecting cost estimation tools for organizational forecasting:

Key Advantages

• Simplicity and Speed: Requires only straightforward basic arithmetic. No specialized statistical software or complex matrix modeling is needed.
• Objective and Replicable: Two analysts analyzing the same historical dataset will arrive at the exact same cost equation, eliminating subjective graphical bias.
• Intuitive Presentation: Readily understood by non-financial operational directors, workshop supervisors, and commercial executives.

Critical Technical Limitations

• Sensitivity to Outliers: Because the method relies entirely on the two extreme data points, any operational anomaly in those two periods—such as a machinery breakdown, an industrial strike, an adverse weather closure, or an unusual overtime surge—will severely distort the resulting cost formula.
• Disregards Intermediate Data: The method completely discards all observations recorded between the highest and lowest activity points. Valuable operational intelligence from intermediate periods is ignored.
• Assumption of Strict Linearity: Forces a straight-line relationship onto operational processes that may naturally follow a curved trajectory due to learning efficiencies or plant congestion.

Comparison with Linear Regression

Where accurate forecasting is paramount, linear regression analysis (the method of ordinary least squares) is superior to the high-low method. Regression utilizes every available historical observation, minimizes the sum of squared errors, and provides statistical measures of reliability (such as the coefficient of determination, $R^2$) to evaluate how well the cost equation fits the actual data.