Forecasting inventory needs using index analysis and linear regression

Forecasting is the process of using data from previous intervals to determine future data. Meteorologists use data from previous weather events to predict future weather patterns. In a similar way, sales can help to predict future inventory stocking needs by accumulated data from previous years. The first step in the process is to create an index for each month by dividing the current month by the index (or first) month. For example, month one of the first year is equal to 55,200. Month one of the second year is equal to 39,800. Dividing the second year by the first year gives an index result of 0.721014. An index number smaller than one indicates a decrease in the number from the first year to the second, and an index number greater than one shows an increase from one year to the next. The following chart shows the resulting index for each month.

Index Year 2 Index Year 3 Index Year

January

February

March

April

May 1 .836449

June

July

August

September 1.407643 1.378545 1.268657

October 0.771455 1.378545 1.268657

November 0.552885 0.723558 0.81851

December 0.573388 0.757202 0.838134

The next step in the forecasting process is to plot each of the monthly indices onto a scatter plot and to use the trend line to determine the function for finding the next month's indices. By using linear regression, the plot yields a solvable slope intercept formula to determine future inventory needs.

The resulting formula, y = 0.203x + 0.403, can be extrapolated to find the index for the fourth year using x = 4. The result of this equation is 1.215, which means that the next number in the series will be greater than the first number or index number. The following below shows the indices for each month with the addition of Year 5.

Indices

Index Y2 Index Y3 Index Y4 Index Y5

January 0.721014 0.582971 1.128623 1.215

February 1.117698 0.67306 1.159547 1.021

March 3.090909 1.624675 2.038961 1.199

April 1.554152 1.851986 1.31769 1.339

May 1 .836449 1.485514 0.785047 0.32

June 0.602339 1.818713 1.105263 1.676

July 2.505556 3.322222 1.972222 2.069

August 2.3-1.552525 2.588384 2.401

September 1.407643 1.378545 1.268657 2.995

October 0.771455 1.378545 1.268657 1.634

November 0.552885 0.723558 0.81851 0.96

December 72,900 41,800 55,200 61,100 71879

This is just one method of forecasting. This method of forecasting might not help a company to foresee outlying problems such as surges in the economy that may influence buying habits. This system is also limited because it requires collection of data for several years before it becomes truly efficient. However, by using mathematical equations and computer software, this company now has a forecast of how much inventory to have on hand for each month for the next year. Forecasting can help the company avoid overstocking costly materials. Making this data available can also help to ensure that the company orders enough supplies to fill all orders based on the stated demand.