CASE STUDY 1

THE SWANK RETAILERS

Mary Demerick, chief operating officer of Swank Retailers in Phoenix, Arizona, is busy looking over the most recent sales information for the company. She has called a meeting of all salespersons in the region for one week from today, and she is attempting to estimate the sales levels that should be expected for their company over the next three months. She needs to have this information so that sales quotas can be set for the individual salespersons. Her staff has accumulated these historical sales data:

═════════════════════════════════════════════════════════════════

Sales Sales Sales

Year 1 ($ Millions) Year 2 ($ Millions) Year 3 ($Millions)

─────────────────────────────────────────────────────────────────

Jan. 4.9 Jan. 5.1 Jan. 5.4 Feb. 6.1 Feb. 6.3 Feb. 7.5 Mar. 7.5 Mar. 7.9 Mar. 8.2 Apr. 7.4 Apr. 8.0 Apr. 8.7 May 5.2 May 5.5 May 6.1

Jun. 5.3 Jun. 5.9 Jun. 6.3

Jul. 5.6 Jul. 6.3 Jul. 6.9

Aug. 7.1 Aug. 7.7 Aug. 8.5

Sept. 8.0 Sept. 8.5 Sept. 9.0

Oct. 6.7 Oct. 7.1 Oct. 8.1

Nov. 8.2 Nov. 8.9 Nov. 10.2

Dec. 7.5 Dec. 8.7 Dec. 9.5

Ms. Demerick expects these sales patterns and trends to continue.

  1. PLOT THE SALES DATA ON A GRAPH AND EXAMINE THE DATA TO IDENTIFY SPECIFIC PATTERNS AND STATE WHICH FORECASTING METHOD(S) SEEM APPROPRIATE FOR THESE PATTERNS.

2.IF EXPONENTIAL SMOOTHING WERE APPLIED TO THESE DATA FOR THE PURPOSE OF DEVELOPING SHORT-RANGE FORECASTS, WOULD YOU RECOMMEND A LOW VALUE OF ALPHA (THE SMOOTHING CONSTANT), SAY .1, OR A HIGH VALUE, SAY .5? WHY?

3.TRY OUT DIFFERENT VALUES OF ALPHA (THE SMOOTHING CONSTANT)AND CHOOSE THE ONE THAT RESULTS IN THE LEAST VALUE OF THE MEAN ABSOLUTE DEVIATION (MAD) OVER THE LAST 12 MONTHS OF THIS DATA? USE THIS VALUE OF ALPHA TO FORECAST NEXT MONTH'S SALES:

4.USE THE SEASONALIZED TIME SERIES REGRESSION ANALYSIS TO DEVELOP A FORECAST FOR THE NEXT THREE MONTHS' SALES. HOW CONFIDENT ARE YOU IN THESE FORECASTS? DEVELOP A STATISTICAL STATEMENT ABOUT NEXT MONTH'S FORECAST (YEAR 4, JANUARY) THAT REFLECTS YOUR LEVEL OF CONFIDENCE IF YOU USE A 95 PERCENT CONFIDENCE INTERVAL:

5.BASED ON YOUR FINDINGS IN PARTS 3 AND 4, WOULD YOU RECOMMEND THAT SWANK USE EXPONENTIAL SMOOTHING OR SEASONALIZED TIME SERIES REGRESSION ANALYSIS? WHAT ARE THE PROS AND CONS OF EACH METHOD IN THIS CASE?