Please note that this article is intended for educational purposes only and should not be deemed to be or used as legal, employment, or health & safety advice. For guidance or advice specific to your business, consult with a qualified professional.
High-low method accounting is used to calculate costs at the maximum (high) and minimum (low) levels of production. This makes it possible to calculate (or at least estimate), the break-even point. Businesses can then use this to forecast when and how they might benefit from economies of scale.
Example of the high-low method
The high-low method formula is:
Fixed Cost + (Variable Cost × Unit Activity) = High-Low Cost
The formula to calculate the variable cost is:
Highest Activity Cost - Lowest Activity Cost / Highest Activity Units - Lowest Activity Units = Variable cost per unit
Fixed costs (also known as overheads) stay the same regardless of the level of business activity. Variable costs, by contrast, increase and decrease in line with output (also known as unit activity). The challenge of the high-low method is therefore to calculate, or at least estimate, the variable costs accurately.
Although the high-low method is designed to be used to calculate costs at maximum and minimum output, the formula can be used for any level of output. It can be useful to apply the formula to different levels of production if any of your variable costs increase in a non-linear way. This is standard practice with costs that relate to contracts for goods or services.
The benefit of the high-low method
The main benefit of the high-low method is that it is simple to implement. Techniques such as regression analysis can eliminate the potential limitations of the high-low method, but the problem with the techniques is that they require a far greater understanding of, and confidence with, maths and maybe specialist tools.
By contrast, the high-low method is simple enough to be used by anyone who understands basic maths. Similarly, the only tool you need is a spreadsheet. In fact, very small businesses (e.g. freelancers) could probably do the necessary calculations with just a basic calculator. The end result may not be as accurate as with other approaches but will generally be more than sufficient for most purposes, especially for SMEs.
The limitations of the high-low method
The limitations of the high-low method are essentially a reflection of its simplicity. Fortunately, it is possible to address them.
The issue of outlier data
One potential issue with the basic approach to the high-low model is that it is vulnerable to outlier data. This can be addressed by hygiene-checking the data before it’s used for the calculation. If the business is established, this could be done by comparing the same time period in different years. If it’s new, then the costs could still be checked manually.
The issue of non-linear tiered pricing
If service contracts use variable pricing, there is a strong possibility that this pricing is tiered. There is also a strong possibility that the rate of increase is non-linear. In other words, as your usage increases, the price-per-unit decreases. This can effectively make it impossible to get a true average variable cost.
How much this matters depends on the extent of the variation between the pricing levels. If it’s fairly low, then it might be pragmatic just to accept it as the impact should be minor. If it’s higher, then it could make sense to apply the high-low calculation to each tier as well as to the overall maximum and minimum points.
The issue of changing pricing
The results of high-low modelling are only valid for as long as the data underpinning them is valid. This means that businesses will need to repeat the high-low modelling exercise periodically to refresh the figures.
How often this needs to happen depends on how often and how significantly prices change. Given that all prices tend to increase over time (inflation), businesses should probably look to undertake high-low modelling at least once a year. In sectors where prices change rapidly, businesses may need to undertake high-low modelling more frequently.