Lecture note on cost behavior determination terms in advanced management accounting study

Lecture note on  cost behavior determination terms in advanced management accounting study

Briefly describe the following cost behavior determination terms in advanced management accounting study

(Terms used in Chapter 9)

 

1.    Cost estimation and predictions

2.    Time span and relevant range

3.    Cost estimation approach: industrial engineering method

4.    Cost estimation approach: conference method

5.    Cost estimation approach: account analysis method

6.    Cost estimation approach: quantitative analysis of current or past cost relationships

7.    High-low method

8.    Regression analysis method

9.    Estimating and choosing cost drivers

10.          Cost drivers and activity-based costing

11.          Big-data-focused quantitative analysis to determine cost drivers of activities

12.          Learning curve and experience curve

 

Cost-behaviour determination estimates how a cost changes when an underlying activity changes, usually expressed as Y=a+bX, where Y is total cost, a fixed cost, b variable cost per activity unit, and X the cost-driver volume.

Core concepts

1.    Cost estimation and predictions

Cost estimation develops a cost function from engineering knowledge, managerial judgment, or observed data. Cost prediction then applies that function to expected future activity levels for planning, budgeting, pricing, and control.

2.    Time span and relevant range

The time span matters because more costs become adjustable or variable in the long run—for example, capacity and staffing may be fixed this month but changeable next year. The relevant range is the expected band of activity within which the assumed cost pattern, often linear, remains reasonably valid; outside it, new capacity, overtime, discounts, or step costs may change the relationship.

Judgment-based approaches

3.    Industrial engineering method

This method estimates cost from the physical relationship between inputs and output—for example, materials required, machine time, and labour time per unit. It commonly uses technical specifications and time-and-motion studies, so it is particularly useful for new products or processes with little historical cost data.

4.    Conference method

Managers and specialists from functions such as production, purchasing, engineering, and accounting jointly identify activities, cost drivers, and likely cost behaviour. It pools organisational expertise but is subjective and can be affected by bias or dominant participants.

5.    Account analysis method

Also called account-classification analysis, this method examines general-ledger accounts and classifies each cost as fixed, variable, or mixed relative to a chosen activity measure. It is quick, inexpensive, and practical, but its quality relies heavily on the analyst’s informed judgement.

Data-based approaches

6.    Quantitative analysis of current or past cost relationships

This approach uses observations of costs and possible drivers—such as units produced, machine-hours, deliveries, or production runs—to statistically estimate a cost function. A normal process is to select the cost to predict, propose driver(s), collect and plot data, estimate the equation, then test whether it predicts adequately.

7.    High-low method

A simple technique using only the observations with the highest and lowest activity levels, not necessarily the highest and lowest costs. Variable cost per unit is estimated as:

B = (Cost at high activity − Cost at low activity)/ (High activity − Low activity)  

Fixed cost is then a=Y−bX. It is easy to apply but unreliable if either selected observation is unusual, since it ignores all other data.

8.    Regression analysis method

Regression uses all relevant observations to estimate the relationship between cost and one driver (simple regression) or several drivers (multiple regression). Least-squares regression chooses coefficients that minimise the squared residuals — the gaps between actual and predicted costs — so it is generally more robust than high-low analysis and allows statistical assessment of fit and significance.

Drivers and ABC

9.    Estimating and choosing cost drivers

A cost driver is a measurable factor that causally explains resource consumption or cost changes. Choose drivers using operational logic first, then test them using data: plausible causality, availability and reliability of measurement, predictive power, stability, manageable collection cost, and avoidance of spurious correlations. A useful model should also be parsimonious: add drivers only when they meaningfully improve decisions.

10.                    Cost drivers and activity-based costing

Activity-based costing (ABC) first assigns resource costs to activity cost pools, then assigns activity costs to products, services, customers, or channels according to each object’s use of activity drivers. For instance, warehouse costs may be linked to receiving, inspection, put-away, picking, packing, and shipping—not allocated merely by production volume.

11.                    Big-data-focused quantitative analysis to determine activity cost drivers

Large, detailed operational datasets—such as ERP transactions, barcode scans, machine logs, delivery records, customer orders, and web activity—can identify drivers at a more granular level and in near real time. Regression, data mining, and machine learning can uncover nonlinear effects and interactions, but managerial validation remains essential: a predictive variable is not automatically a causal, decision-useful driver. Research on machine-learning ABC indicates accurate allocations depend on sufficiently informative longitudinal data and avoiding omission of drivers that explain substantial cost variation.

Learning and experience

12.                    Learning curve and experience curve

A learning curve shows that labour hours or labour cost per unit tend to fall as cumulative production rises, because workers learn, processes improve, and waste declines. An experience curve extends this idea beyond direct labour to broader value-chain costs, reflecting accumulated production and sales experience, scale, design improvement, procurement learning, and process innovation. These effects mean unit cost may decline nonlinearly even when output is increasing.

Also study the videos on this topic: video 1; video 2; video 3.

  

A collection of notes on advanced management accounting