Applying Time Series Methodologies

| Blues Inc. |

Memo

To: Myra Reid

From: Nichole Noble

Date: August 3, 2011

Re: Forecasting

Forecasting

After completing three weeks of forecasting for Blues Inc., I have come to some determinations and want to explain the process and ways in which I was able to forecast for the company.

In week one of my forecasting I looked at the sales in millions of dollars versus the advertising budget that was set. We expected the market size for this year to be $40 billion. Using the regression equation to set the advertising budget I stated that we should see the sales at $2,400 million and $162 million for advertising. Using the sales as the variable in the regression analysis was a good decision because the correlation coefficient of sales with the advertising budget is 0.96. This means that sales have a strong positive relationship with the advertising budget.

In week two I studied the fluctuations in the market size to arrive at a sales forecast. To complete the forecasting I chose to use the simple moving average and set the production levels at 47 million units. I should have used the weighted moving average model to arrive at a better forecasting for the company. Using the weighted moving average would have allowed me to arrive at a reasonably accurate forecast. While using the wrong moving average model, I accurately predicted to keep up with the market size, Blues Inc. would need to produce 47.00 million units.

Finally in week three I needed to use the center moving average model to forecast the quarterly results. Using the centered moving average model allowed me to look at the results of months past to predict what will happen in the future. This information also allowed me to fix production levels for all 4 quarters. I decided to set production at the following levels: 12.00 million for the first quarter, 14.00 million for the second quarter, 13.00 million for the third quarter, and 20.00 million units for the fourth quarter. This was a good...