Cost Behaviour and Estimation Flashcards
Why do we need to estimate costs?
- assure profitability of a product/service
- evaluate the profit margin
Cost behaviour, estimation and prediction
CB: existing relationship between cost and activity
CE: process of estimating relationship of costs and cost driver activities
CP: use of CE to forecast level of costs at particular activity levels
Types of cost behaviour
- step fixed costs and step variable costs
- mixed costs
- semi-variable costs
Cost estimation methods
- Regression analysis (statistical method)
- account analysis
- engineering method
Regression Analysis equation
TC = F + VX or Y = a + bX TC/Y: total costs, dependent variable F/a: fixed costs, intercept V/b: variable cost per unit, slope/gradient X: cost driver, independent variable
Regression Analysis methodology
To create an equation that relates a dependent and one or more independent variables, using data from the past in order to estimate relationships between costs and cost driver activities
Line of ‘best fit’ or ‘least squares’
Find the line through data points that minimizes sum of squared distances from the data points to the line
Measures of correlation and ‘best fit’
Correlation coefficient (r): strength of relationship between variables
- [r = (-)1]: strong positive or negative relationship
- [r = 0]: no relationship
Coefficient of determination (R^2): measure of goodness of fit
- R^2 close to 1: data points can be strongly fitted into a straight line
Considerations of regression analysis
- More data points means more accurate results
- Only estimate costs in the relevant range of activities (e.g. estimation of FC can result in negative)
Regression equation may be a poor predictor of future costs if
- cost-activity relationships have changed (e.g. through time, after decisions)
- Costs themselves have changed independently from activity changes (external factors, e.g. supplier prices)
Advantages of Regression analysis
- permits the use of more than one cost driver activity to explain costs
- predict costs for new products that are similar to past products with its regression model
- equation can be reused at any times to calculate costs at different activity levels (giving that no external factors or internal changes have taken place)
Problems with regression analysis
- these are only estimations, not exact
- can’t be 100% sure that variables have a good relationship (even though logical)
- cannot have more than one Independent variable that are highly correlated
- Historical costs: mismatched periods of highly inflationary periods
Account analysis methodology
- review each activity that makes up total cost
- relate costs and activity in the form of general equation (TC = F + VX)
Account analysis: process
- identify cost drivers and their costs
- sum fixed costs
- sum VC for each cost driver activity
- Divide total VC calculated beforehand by total cost driver units to obtain VC per unit
- Divide FC by number of time periods in data
- Substitute each number in equation
Engineering method process
Direct labour: find labour cost per unit
Direct material: material cost per unit
Overhead costs: estimate total OH costs
