The high-low method in accounting is the simplest and easiest way to separate mixed costs into their fixed and variable components. By using this method, we observe only the highest and lowest points in the data set with the assumption that all the data have a linear relationship. We use the high-low method accounting formula to calculate the variable unit per cost as the change in total cost divided by the change in units produced (or other measure of activity).
Variable Cost per unit | = | Highest Activity Total Cost – Lowest Activity Total Cost |
Highest Units Produced – Lowest Units Produced |
Key takeaways
- The high-low method estimates variable costs using the total costs incurred while at the highest level of production minus the total costs incurred at the lowest level of production.
- The high-low method is the fastest way to separate mixed costs into fixed and variable components.
- Using the high-low method may produce less reliable and less accurate results because it only uses two extreme data points in calculations. In other words, it’s only a rough estimate of actual variable costs.
- Regression analysis or least squares method is an alternative to the high-low method since it produces finer and more accurate results and considers all observations.
- Regardless of the method used, the information derived from either the high-low or regression analysis method is only as good as the data used. Using a more sophisticated method won’t compensate for inaccurate data.
How To Use the High-low Method & Example
Follow the steps below to perform the high-low method by using our sample data from Fusion Company. Let’s assume that the company wants to project client support costs for next year’s budgeting.
Month | Number of Client Calls | Total Cost of Client Support |
---|---|---|
January | 140 | $15,500 |
February | 120 | $13,500 |
March | 150 | $16,500 |
April | 160 | $17,500 |
May | 130 | $14,500 |
June | 190 | $22,500 |
July | 200 | $24,500 |
August | 220 | $27,500 |
September | 300 | $31,500 |
October | 100 | $11,500 |
November | 110 | $12,500 |
December | 120 | $13,500 |
In the sample data above, the number of client calls refers to the activity level. The activity level can pertain to any measurable business activity, such as documents processed, units produced, finished goods inspected, or services rendered. The total cost of client support is the activity cost. It is presented in total, so we can’t immediately determine the fixed or variable components.
Step 1: Determine the Highest and Lowest Activity Levels
The highest and lowest activity levels are September at 300 client calls and October at 100 client calls. As far as the high-low method is concerned, these are the only data points that we’ll use in the calculation. The rest of the data will be ignored.
Hence, our high and low points are:
- High: 300 calls at $31,500
- Low: 100 calls at $11,500
Sometimes, outliers—which are activity levels or costs that are abnormally high or low if compared to the rest of the observations—may exist in the data set. For instance, if the number of client calls in December reaches 1,000 calls, such is considered an outlier since it’s too far from the other observations.
When you encounter an outlier, simply remove it from the dataset and use the high-low method for the remaining observations.
Step 2: Compute Variable Cost per Unit
By using the formula in computing the variable cost per unit, let’s substitute the figures we gathered from Step 1.
Variable Cost per Unit | = | 31,500 – 11,500 = 20,000 |
300 – 100 = 200 |
Variable Cost per Unit = $100
This means each client support call costs $100. Now that we have this figure, let’s proceed to Step 3 to determine the total fixed cost.
Step 3: Compute Fixed Cost Using the Variable Cost per Unit in Step 2 and the Cost Equation
The total fixed cost in this manner is regardless of activity level. Whether the activity level is high or low, fixed costs remain constant. And if the activity level is zero, the total costs will just be equal to the total fixed costs. To compute total fixed costs, let’s use the cost equation.
y = a + bX
where:
- y = total cost or the dependent variable
- a = total fixed cost or the y-intercept
- b = variable cost per unit or the slope
- X = activity level or the independent variable
In Step 2, we know that b is $100. To substitute the rest except a, we pick either the high or low point as reference. Regardless of our choice, the fixed cost will be the same.
Using the High Point | Using the Low Point |
---|---|
31,500 = a + (100)(300) 31,500 = a + 30,000 a = 31,500 - 30,000 a = 1,500 | 11,500 = a + (100)(100) 11,500 = a + 10,000 a = 11,500 - 10,000 a = 1,500 |
Therefore, total fixed costs for client support calls is $1,500 per month. In the side-by-side computation above, we’ve proven our point that regardless of which reference point we use, we still arrive at $1,500.
Step 4: Assemble the Cost Equation
Now that all elements of the equation are complete, let’s assemble the cost equation for client support calls. The equation should look like this:
y = 1,500 + 100X
You can now use this cost equation to project future costs of client support calls for budgeting purposes. If you want to double-check if the equation is correct, try computing for other months and check if your answer and the total client support costs are the same.
To prove our earlier point, let’s assume that there is zero activity. Will the total cost be zero as well? Let’s compute:
y = 1,500 + 100(0)
y = 1,500 + 0
y = 1,500
Therefore, even though we have zero client support calls, we still incur $1,500 client support costs because these are fixed costs.
The equation we presented above is always true in a perfect world. But in reality, it is not always the case due to unforeseen circumstances, uncontrollable events, and uncertainties. For illustration purposes, let’s use the cost equation above to check if the total costs for the following months:
Month | Number of Client Calls | Total Cost of Client Support |
---|---|---|
June | 190 | $22,500 |
July | 200 | $24,500 |
August | 220 | $27,500 |
- June: 1,500 + 100(190) = $20,500 vs $22,500 (variance of $2,000)
- July: 1,500 + 100(200) = $21,500 vs $24,500 (variance of $3,000)
- August: 1,500 + 100(220) = $23,500 vs $27,500 (variance of $4,000)
The computations above show that the actual total costs and computed total costs using the equation don’t match. This scenario best shows that there will be instances where the cost equation won’t hold true.
For the months from June to August, the actual costs are always higher than the computed costs. These variances can stem from different causes, and every business manager should look at the variances.
But more importantly, this scenario shows the weakness of the high-low method. Since our first computation excludes June, July, and August, we could not include its data in our cost equation. This only means that if we use the cost equation to project next year’s cost for June to August, then we may be underestimating costs in the budget.
Highlights and Limitations of the High-low Method
Highlights | Limitations |
---|---|
Is easy to perform | Can be unreliable because the highest and lowest values may be outliers |
Requires only basic understanding of algebra | Can be inaccurate since it only considers two data points |
Provides a quick estimate of fixed and variable costs | |
Is best for business activities that don’t experience sudden fluctuations in costs |
High-low Method vs Regression Analysis
High-low | Regression Analysis | |
---|---|---|
Ease of Use | Easy | Difficult |
Number of Data Points Used | 2 only (excluding outliers) | All (excluding outliers) |
Accuracy of Information | Less accurate | More accurate |
In managerial accounting, both the high-low method and regression analysis separate mixed costs into their fixed and variable components. The main difference between the two would be the approximation of results and difficulty. There’s no problem in using the high-low method in accounting since it still provides actionable information. Choosing between high-low or regression analysis methods is only a matter of capability and expertise.
If you or anyone in your company possesses statistical and data analysis skills, go for regression analysis and make use of other sophisticated methods like linear programming. But if you’re a small business owner with little expertise in data analysis and statistics, the high-low method is easy to use and only requires basic knowledge in algebra.
Overall, the method applied is only as good as the data used. If the data is inaccurate, either method will produce inaccurate results. Good bookkeeping is still essential to ensure high-quality data for analysis. To learn more about bookkeeping, our guide on small business bookkeeping will teach you how to perform small business bookkeeping and how to organize accounting data appropriately.
Frequently Asked Questions (FAQs)
Separating variable and fixed costs can help you understand the business’ cost structure. Both of these costs have an impact on overall profitability and knowing each will help you make better decisions. Differentiating fixed and variable components can also aid in breakeven point analysis wherein you can determine the minimum revenue you need to reach breakeven point or the point at profit is zero.
The high-low method may produce inaccurate results since it only considers two extreme data points, which may not be representative of other data points. It can also be unreliable because it’s possible that the highest and lowest points are outliers.
Yes, because it is a simple tool to compute costs at different activity levels. It can also be used for budgeting purposes, especially for business activities with fixed and variable components.
Bottom Line
The high-low method is an easy way to separate fixed and variable costs. This tool can help you understand the business’ cost structure and aid in rational decision-making. However, it can produce less accurate and unreliable results since it only uses two extreme data points.