What Is the High-Low Method?
The high-low method is a cost accounting technique used to separate a mixed cost — a cost containing both fixed and variable components — into its fixed and variable elements using only two data points: the highest and lowest activity levels within a given period. By comparing the total cost at the highest activity level to the total cost at the lowest activity level, and dividing the cost difference by the activity difference, the variable cost per unit of activity is estimated. The fixed cost is then derived by subtracting the total variable cost from the total cost at either the high or low point. While simple and quick, the high-low method is the least sophisticated cost estimation technique and can produce misleading results when the high and low points are outliers or when the cost-activity relationship is not linear.
How the High-Low Method Works
The calculation follows three straightforward steps. First, identify the periods with the highest and lowest activity levels — not the highest and lowest costs. This distinction is critical: the method uses activity level (units produced, machine hours, labor hours, miles driven) as the independent variable driving cost changes. Second, compute the variable cost per unit: Variable Cost per Unit = (Cost at High Activity - Cost at Low Activity) / (High Activity Level - Low Activity Level). Third, compute the fixed cost: Fixed Cost = Total Cost at High (or Low) Activity - (Variable Cost per Unit × High (or Low) Activity Level). For example, a factory's electricity costs are $50,000 in the month when 10,000 units were produced (the high) and $30,000 when 5,000 units were produced (the low). Variable cost per unit = ($50,000 - $30,000) / (10,000 - 5,000) = $4 per unit. Fixed cost = $50,000 - ($4 × 10,000) = $10,000. The cost equation is: Total Electricity Cost = $10,000 + $4 per unit produced.
Limitations and When to Use Caution
The high-low method's simplicity is also its greatest weakness. By using only two data points, it discards all other available information and is extremely sensitive to outliers. If the high-activity month coincidentally had unusually high costs due to a one-time equipment repair, or the low-activity month was an atypical holiday period with artificially low costs, the resulting cost function will be distorted. The method assumes a perfectly linear relationship between cost and activity within the relevant range — an assumption that may not hold if costs exhibit step-function behavior (jumping at certain thresholds) or economies of scale (variable cost per unit declining with volume). The method also assumes that a single activity driver adequately explains cost behavior, when in reality costs may be driven by multiple factors. For these reasons, the high-low method is best suited for quick approximations, preliminary analyses, and situations where the data is sparse and the relationship is known to be approximately linear. For anything requiring precision — budgeting, pricing, make-or-buy decisions — more sophisticated methods such as least-squares regression analysis should be used.
Real-World Application: Estimating Delivery Costs
A small wholesale distributor is trying to understand its monthly delivery cost structure to plan for an anticipated increase in delivery volume. Over the past year, monthly delivery miles ranged from 3,000 (low, in a slow January) to 8,000 (high, in a busy November), with corresponding delivery costs of $12,000 and $27,000. Using the high-low method: variable cost per mile = ($27,000 - $12,000) / (8,000 - 3,000) = $3 per mile; fixed cost = $27,000 - ($3 × 8,000) = $3,000. The estimated cost function is $3,000 + $3 per mile. If the distributor expects 10,000 miles next month, the estimated delivery cost is $3,000 + $3 × 10,000 = $33,000. This quick estimate helps with near-term planning, but the distributor recognizes its limitations: the cost behavior between 8,000 and 10,000 miles may differ from the 3,000-8,000 range, and factors like fuel price fluctuations and driver overtime pay at higher mileages could cause actual costs to diverge from the simple linear estimate.
The High-Low Method in Context
The high-low method exists within a hierarchy of cost estimation techniques. At the simplest level are account analysis and engineering estimates — subjective but informed by operational knowledge. The high-low method introduces quantitative rigor with minimal data requirements. Scattergraph (visual fit) analysis plots all data points and fits a line by eye, using all available data but introducing subjectivity in line placement. Least-squares regression analysis uses all data points to fit the line that minimizes the sum of squared errors, producing the most statistically reliable estimates and enabling measures of goodness-of-fit and confidence intervals. For serious cost analysis — particularly when decisions involve significant financial commitments — regression analysis is the standard. The high-low method is useful primarily as a rapid diagnostic tool and as a pedagogical stepping stone to more advanced techniques. Understanding its limitations is as important as understanding its calculation.
FAQ
Why use the highest and lowest activity levels rather than the highest and lowest costs?
The high-low method assumes that activity level is the independent variable driving cost changes. If costs spike in a period due to an unusual event (like a repair) rather than high activity, using that period as the "high" point would distort the variable cost estimate. By selecting points based on activity, the method attempts to capture the systematic relationship between volume and cost, though outliers in cost at those activity levels can still cause distortion.
When is the high-low method appropriate to use?
The high-low method is appropriate for quick initial estimates when data is limited, when the cost-activity relationship is believed to be strongly linear within the relevant range, and when the consequences of estimation error are minor. It is also useful in educational settings to illustrate cost behavior concepts. It is NOT appropriate for significant financial decisions, regulatory filings, or when data is abundant — in those cases, regression analysis should be used.
Related Terms
- Fixed Cost — a cost that remains constant in total regardless of changes in activity level within the relevant range
- Variable Cost — a cost that changes proportionally with changes in activity level
- Mixed Cost (Semi-Variable Cost) — a cost containing both fixed and variable components
- Regression Analysis — a statistical technique for estimating the relationship between a dependent variable and one or more independent variables
- Relevant Range — the range of activity within which the assumed cost behavior patterns remain valid
Related MoneyBestPal Guides
Use these related MoneyBestPal resources to compare High-Low Method with nearby finance, economics, investing, and business concepts.
 |
| Image: Moneybestpal.com |
In accounting and finance, the fixed and variable parts of a cost function are estimated using the high-low method. For the purpose of calculating the overall variable cost and the fixed cost component, the procedure entails comparing the highest and lowest levels of activity.
A business must first determine the times when activity was highest and lowest, as well as the overall costs incurred during those times, before applying the high-low approach. The procedure then entails figuring out the variable cost per unit by dividing the variation in total costs by the variation in activity level between the two periods. It is possible to estimate the variable component of costs for additional activity levels once the variable cost per unit has been established.
The fixed cost component is then calculated by subtracting the entire variable cost from the overall cost for the high or low period. The fixed cost component can then be used to predict the fixed portion of expenses for other activity levels.
The high-low approach is a useful resource for cost assessment when data is scarce or more involved procedures are impracticable. It should be emphasized that the high-low method assumes a linear relationship between activity and costs, which may not always be the case in reality. So, it's important to use caution and other procedures to check the predictions made by the high-low method.
