Advanced : Regression analys is and an analysis of costs where un it variable cost Is no

Question:

Advanced : Regression analys is and an analysis of costs where un it variable cost Is no t con stant (See note on page 663)

Babel ltd produces a single product. The company's directors want to explore new markets. and they require an accurate analysis of the firm's cost structure for both forecasting and pricing purposes.

An attempt to provide this analysis from the aggregation of individual costs has produced a poor correspondence between actual and predicted costs. You are an accountant employed by Babel ltd, and you have been asked to provide a statistical approach to the problem.

The financial director has given you the following data:

image text in transcribed

You obtain the following further information:
(1) The costs from which the averages have been computed consist of the firm's entire costs for the relevant month.
(2) Fixed costs can be assumed to be unaffected by seasonal factors except for winter heating . In January and February a supplementary heating system was employed; this cost £10000 per month to operate.
You are asked to transform the original data in any way which may be necessary. Work to the nearest £100 for total costs and to the nearest10p for unit cost calculations.
Requirements

(a) Estimate Babel ltd's normal fixed and variable cost of production using linear regression, showing full workings.
(7 marks)

(b) Draw a graph of Babel Ltd 's costs versus output to provide a better analysis of costs than the regression estimates in part (a), estimate the point at which variable costs change and estimate the linear cost relationships over the relevant ranges. (5 marks)

(c) Use your answer from part

(b) to provide an analysis of costs from January to June, set out in spreadsheet format (i.e. rows and columns) . Divide the rows into relevant cost categories to explain observed costs. and treat any residual amounts as unexplained differences. (6 marks)

(d) Discuss the difficulties which may be encountered in preparing and using statistical cost estimation in practice.