Cost Management: A Strategic Emphasis 5th Edition by David Stout, Edward Blocher, Gary Cokins

Exercise 52

Cost Estimation, High-Low Method, Regression Analysis DVD Express is a large manufacturer of affordable DVD players. Management recently became aware of rising costs resulting from returns of malfunctioning products. As a starting point for further analysis, Bridget Forrester, the controller, wants to test different forecasting methods and then use the best one to forecast quarterly expenses for 2010. The relevant data for the previous three years follows:

2007 Quarter	Return Expenses	2008 Quarter	Return Expenses	2009 Quarter	Return Expenses
1	$15,000	1	$16,200	1	$16,600
2	17,500	2	17,800	2	18,100
3	18,500	3	18,800	3	19,000
4	18,600	4	17,700	4	19,200

The result of a simple regression analysis using all 12 data points yielded an intercept of $16,559.09 and a coefficient for the independent variable of $183.22. (R-squared = .27, t = 1.94, SE = 1128).

Required

1. Calculate the quarterly forecast for 2010 using the high-low method and regression analyses. Recommend which method Bridget should use and explain why.

---

2. How does your analysis in requirement 1 change if DVD Express manufactures its products in multiple global production facilities to serve the global market?

Step 1 of 2

Cost Estimation; High-Low Method, Regression Analysis (30min)

1.?High-Low Method

An examination of the exhibit below indicates that representative high and low points are the last and first data points, respectively, so these points are used to develop the high-low estimate.

Variable cost  = ($19,200 - $15,000)/(12 - 1) = $381.82

Fixed cost = $ 15,000 – ($381.82 x 1) = $14,618.18

    [also:   $19,200 – ($381.82 x 12) = $14,618.18]

Quarterly Predictions are:

??$14,618 + $381.82 x 13 = $19,582

??$14,618 + $381.82 x 14 = $ 19,963

??$14,618 + $381.82 x 15 = $ 20,345

??$14,618 + $381.82 x 16 = $ 20,727

Regression

The regression equation from the spreadsheet is:

Return Expense

= $16,559+ (quarter number x 183.22)

Predicted Expense for the next four quarters using regression analysis:

??1   $16,559+ (13 x $183.22)   =  $  18,940.86

??2   $16,559 + (14 x $183.22)   =  $ 19,124.08

??3   $16,559 + (15 x $183.22)   =  $ 19,307.30

??4   $16,559 + (16 x $183.22)   =  $ 19,490.52

Since there is noticeable seasonality in the data (lower for periods 1, 5 and 9, the first quarters of the year) it is possible to improve on the regression model by adding a dummy variable with 0s in all periods and 1s in the periods 1,5, and 9. The result is below, which shows a much higher R squared, improved SE, and a significant  t-value  for the dummy variable.

??1   $17,556.3 + (13 x $114.1791) – 2,193.86   =  $  16,847

??2   $17,556.3 + (14 x $114.l791)   =  $ 19,155

??3   $17,556.3 + (15 x $114.1791)  =  $ 19,269

??4   $17,556.3 + (16 x $114.1791)  =  $ 19,383