A.

1 Forecast Calculation Methods Month 2004 Sales 2005 Sales 2006 Forecast
Simulated 2005
Forecast
Twelve methods of calculating forecasts are available. Most of these methods provide for limited user control.
For example, the weight placed on recent historical data or the date range of historical data used in the August 128 140 161
calculations might be specified. The following examples show the calculation procedure for each of the
available forecasting methods, given an identical set of historical data. September 118 131 151

The following examples use the same 2004 and 2005 sales data to produce a 2006 sales forecast. In addition October 123 114 131 141.45
to the forecast calculation, each example includes a simulated 2005 forecast for a three month holdout period
November 139 119 137 159.85
(processing option 19 = '3') which is then used for percent of accuracy and mean absolute deviation
calculations (actual sales compared to simulated forecast).
December 133 137 158 152.95

A.2 Forecast Performance Evaluation Criteria

Depending on your selection of processing options and on the trends and patterns existing in the sales data,
some forecasting methods will perform better than others for a given historical data set. A forecasting method A.3.1 Forecast Calculation
that is appropriate for one product may not be appropriate for another product. It is also unlikely that a User specified factor (processing option 1a) = 1.15 in this example.
forecasting method that provides good results at one stage of a product's life cycle will remain appropriate
throughout the entire life cycle.
A.3.2 Simulated Forecast Calculation
You can choose between two methods to evaluate the current performance of the forecasting methods. These October, 2004 sales = 123 * 1.15 = 141.45
are Mean Absolute Deviation (MAD) and Percent of Accuracy (POA). Both of these performance evaluation
methods require historical sales data for a user specified period of time. This period of time is called a holdout November, 2004 sales = 139 * 1.15 = 159.85
period or periods best fit (PBF). The data in this period is used as the basis for recommending which of the
forecasting methods to use in making the next forecast projection. This recommendation is specific to each
December, 2004 sales = 133 * 1.15 = 152.95
product, and may change from one forecast generation to the next. The two forecast performance evaluation
methods are demonstrated in the pages following the examples of the twelve forecasting methods.
A.3.3 Percent of Accuracy Calculation
POA = (141.45 + 159.85 + 152.95) / (114 + 119 + 137) * 100 = 454.25 / 370 = 122.770
A.3 Method 1 - Specified Percent Over Last Year
This method multiplies sales data from the previous year by a user specified factor; for example, 1.10 for a 10% A.3.4 Mean Absolute Deviation Calculation
increase, or 0.97 for a 3% decrease.
MAD = (|141.45 - 114| + |159.85 - 119| + |152.95 - 137|) / 3 = (27.45 + 40.85 + 15.95) / 3 = 84.25/3 = 28.08
Required sales history: One year for calculating the forecast plus the user specified number of time periods for
evaluating forecast performance (processing option 19).
A.4 Method 2 - Calculated Percent Over Last Year
Simulated 2005 This method multiplies sales data from the previous year by a factor calculated by the system.
Month 2004 Sales 2005 Sales 2006 Forecast Forecast
Required sales history: One year for calculating the forecast plus the user specified number of time periods for
January 125 128 147 evaluating forecast performance (processing option 19).

February 132 117 135

Simulated 2005
March 115 115 132 Month 2004 Sales 2005 Sales 2006 Forecast Forecast

April 137 125 144 January 125 128 120

May 122 122 140 February 132 117 110

June 130 137 158 March 115 115 108

July 141 129 148 April 137 125 117

 November, 2004 sales = 139 * 1.033591731 = 143.66925
Simulated 2005
Month 2004 Sales 2005 Sales 2006 Forecast Forecast
December, 2004 sales = 133 * 1.033591731 = 137.4677
May 122 122 114

June 130 137 128

A.4.3 Percent of Accuracy Calculation
POA = (127.13178 + 143.66925 + 137.4677) / (114 + 119 + 137) * 100 = 408.26873 / 370 * 100 = 110.3429
July 141 129 121

August 128 140 131 A.4.4 Mean Absolute Deviation Calculation

MAD = (|127.13178 - 114| + |143.66925 - 119| + |137.4677- 137|) / 3 = (13.13178+ 24.66925 + 0.4677)/3 =
September 118 131 123
12.75624

October 123 114 107 127.13178

November 139 119 111 143.66925 A.5 Method 3 - Last year to This Year
December 133 137 128 137.4677 This method copies sales data from the previous year to the next year.

Required sales history: One year for calculating the forecast plus the number of time periods specified for
evaluating forecast performance (processing option 19).

A.4.1 Forecast Calculation Simulated 2005

Month 2004 Sales 2005 Sales 2006 Forecast Forecast
Range of sales history to use in calculating growth factor (processing option 2a) = 3 in this example.
January 125 128 128
Sum the final three months of 2005: 114 + 119 + 137 = 370
February 132 117 117
Sum the same three months for the previous year: 123 + 139 + 133 = 395
March 115 115 115
The calculated factor = 370/395 = 0.9367
April 137 125 125
Calculate the forecasts:
May 122 122 122
January, 2005 sales = 128 * 0.9367 = 119.8036 or about 120
June 130 137 137

February, 2005 sales = 117 * 0.9367 = 109.5939 or about 110

July 141 129 129

March, 2005 sales = 115 * 0.9367 = 107.7205 or about 108 August 128 140 140

A.4.2 Simulated Forecast Calculation September 118 131 131

Sum the three months of 2005 prior to holdout period (July, Aug, Sept): October 123 114 114 123

129 + 140 + 131 = 400 November 139 119 119 139

Sum the same three months for the previous year: December 133 137 137 133

141 + 128 + 118 = 387

The calculated factor = 400/387 = 1.033591731

A.5.1 Forecast Calculation
Calculate simulated forecast: January 2005 sales = January 2006 forecast = 128

October, 2004 sales = 123 * 1.033591731 = 127.13178

February 2005 sales = February 2006 forecast = 117
March 2005 sales = March 2006 forecast = 115
A.6.1 Forecast Calculation
Number of periods to be included in the average (processing option 4a) = 3 in this example
A.5.2 Simulated Forecast Calculation
October 2004 sales = 123 For each month of the forecast, average the previous three month's data.

November 2004 sales = 139 January forecast: 114 + 119 + 137 = 370, 370 / 3 = 123.333 or 123

December 2004 sales = 133 February forecast: 119 + 137 + 123 = 379, 379 / 3 = 126.333 or 126

March forecast: 137 + 123 + 126 = 379, 386 / 3 = 128.667 or 129

A.5.3 Percent of Accuracy Calculation
POA = (123 + 139 + 133) / (114 + 119 + 137) * 100 = 395/370 * 100 = 106.7567 A.6.2 Simulated Forecast Calculation
October 2005 sales = (129 + 140 + 131)/3 = 133.3333
A.5.4 Mean Absolute Deviation Calculation
MAD = (|123-114| + |139 - 119| + |133 - 137|) / 3 = (9 + 20 + 4)/3 = 11 November 2005 sales = (140 + 131 + 114)/3 = 128.3333

December 2005 sales = (131 + 114 + 119)/3 = 121.3333

A.6 Method 4 - Moving Average

This method averages a user specified number of months (processing option 4a) to project the next months
A.6.3 Percent of Accuracy Calculation
demand. POA = (133.3333 + 128.3333 + 121.3333) / (114 + 119 + 137) * 100 = 103.513

Required sales history: Twice the number of periods to be included in the average (processing option 4a), plus
number of time periods for evaluating forecast performance (processing option 19).
A.6.4 Mean Absolute Deviation Calculation
MAD = (|133.3333 - 114| + |128.3333 - 119| + |121.3333 - 137|) / 3 = 14.7777
Simulated 2005
Month 2004 Sales 2005 Sales 2006 Forecast Forecast
A.7 Method 5 - Linear Approximation
January 125 128 123
Linear Approximation calculates a trend based upon two sales history data points.
February 132 117 126
Those two points define a straight trend line that is projected into the future. Use this method with caution, as
March 115 115 129 long range forecasts are leveraged by small changes in just two data points.

April 137 125 126 Required sales history: The number of periods to include in regression (processing option 5a), plus 1 plus the
number of time periods for evaluating forecast performance (processing option 19).
May 122 122 127

June 130 137 127 Simulated 2005

Month 2004 Sales 2005 Sales 2006 Sales Forecast
July 141 129 127
January 125 128 149
August 128 140 127
February 132 117 160
September 118 131 127
March 115 115 172
October 123 114 127 133.3333
April 137 125 183
November 139 119 127 128.3333
May 122 122 195
December 133 137 127 121.3333
June 130 137 206

July 141 129 218

 Forecast specifications: n = identifies the periods of sales history that will be used in calculating the values for a
Simulated 2005
Month 2004 Sales 2005 Sales 2006 Sales Forecast and b. For example, specify n = 3 to use the history from October through December, 2005 as the basis for the
calculations. When data is available a larger n (such as n = 24) would ordinarily be used. LSR will define a line
August 128 140 229 for as few as two data points. For this example, a small value for n (n = 3) was chosen to reduce the manual
calculations required to verify the results.
September 118 131 241
Required sales history: The number of periods to include in regression (processing option 6a) plus the number
October 123 114 252 132 of time periods for evaluating forecast performance (processing option 19).

November 139 119 264 101

Simulated 2005
Month 2004 Sales 2005 Sales 2006 Forecast Forecast
December 133 137 275 113
January 125 128 146

February 132 117 158

A.7.1 Forecast Calculation March 115 115 169

Number of periods to include in regression (processing option 5a) = 3 in this example

April 137 125 181

For each month of the forecast, add the increase or decrease during the specified periods prior to holdout May 122 122 192
period the previous period.
June 130 137 204
January forecast: (137 - 114)/2 + 137 = 148.5 or 149
July 141 129 215
February forecast: (137 - 114)/2 * 2 + 137 = 160
August 128 140 227
March forecast: (137 - 114)/2 * 3 + 137 = 171.5 or 172
September 118 131 238

A.7.2 Simulated Forecast Calculation October 123 114 250 135.333

October 2004 sales = (131 - 129) / 2 + 131 = 132

November 139 119 261 102.333

November 2004 sales = (114 - 140) / 2 + 114 = 101 December 133 137 273 109.333

December 2004 sales = (119 - 131) /2 + 119 = 113

A.7.3 Percent of Accuracy Calculation

POA = (132 + 101 + 113) / (114 + 119 + 137) * 100 = 93.5135
A.8.1 Forecast Calculation
Number of periods to include in regression (processing option 6a) = 3 in this example

A.7.4 Mean Absolute Deviation Calculation For each month of the forecast, add the increase or decrease during the specified periods prior to holdout
MAD = (|132 - 114| + |101 - 119| + |113 - 137|) / 3 = 20 period the previous period.

January forecast:
A.8 Method 6 - Least Square Regression Average of the previous three months = (114 + 119 + 137)/3 = 123.3333
Linear Regression or Least Squares Regression (LSR) is the most popular method for identifying a linear trend
in historical sales data. The method calculates the values for "a" and "b" to be used in the formula: Y = a + bX. Summary of the previous three months with weight considered
The equation describes a straight line where Y represents sales, and X represents time. Linear regression is
slow to recognize turning points and step function shifts in demand. Linear regression fits a straight line to the = (114 * 1) + (119 * 2) + (137 * 3) = 763
data, even when the data is seasonal or would better be described by a curve. When the sales history data
follows a curve or has a strong seasonal pattern, forecast bias and systematic errors occur. Difference between the values
= 763 - 123.3333 * (1 + 2 + 3) = 23 Value2 = Average - value1 * ratio = 128.3333 - (-12.9999) * 2 = 154.3333

Ratio = (1^2 + 2^2 + 3^2) - {(1 + 2 + 3)/3}^2 * 3 = 14 - 12 = 2 Forecast = 4 * -12.9999 + 154.3333 = 102.3333

Value1 = Difference/Ratio = 23/2 = 11.5 December 2004 sales

Value2 = Average - value1 * ratio = 123.3333 - 11.5 * 2 = 100.3333 Average of the previous three months

Forecast = (1 + n) * value1 + value2 = 4 * 11.5 + 100.3333 = 146.333 or 146 = (131 + 114 + 119)/3 = 121.3333

February forecast: Summary of the previous three months with weight considered

Forecast = 5 * 11.5 + 100.3333 = 157.8333 or 158 = (131 * 1) + (114 * 2) + (119 * 3) = 716

March forecast: Difference between the values

Forecast = 6 * 11.5 + 100.3333 = 169.3333 or 169 = 716 - 121.3333 * (1 + 2 + 3) = -11.9999

Value1 = Difference/Ratio = -11.9999/2 = -5.9999

A.8.2 Simulated Forecast Calculation
October 2004 sales: Value2 = Average - value1 * ratio = 121.3333 - (-5.9999) * 2 = 133.3333

Average of the previous three months Forecast = 4 * (-5.9999) + 133.3333 = 109.3333

= (129 + 140 + 131)/3 = 133.3333

A.8.3 Percent of Accuracy Calculation
Summary of the previous three months with weight considered POA = (135.33 + 102.33 + 109.33) / (114 + 119 + 137) * 100 = 93.78

= (129 * 1) + (140 * 2) + (131 * 3) = 802

A.8.4 Mean Absolute Deviation Calculation
Difference between the values MAD = (|135.33 - 114| + |102.33 - 119| + |109.33 - 137|) / 3 = 21.88

= 802 - 133.3333 * (1 + 2 + 3) = 2

Ratio = (1^2 + 2^2 + 3^2) - {(1 + 2 + 3)/3}^2 * 3 = 14 - 12 = 2

A.9 Method 7 - Second Degree Approximation
Linear Regression determines values for a and b in the forecast formula Y = a + bX with the objective of fitting a
Value1 = Difference/Ratio = 2/2 = 1 straight line to the sales history data. Second Degree Approximation is similar. However, this method
determines values for a, b, and c in the forecast formula Y = a + bX + cX2 with the objective of fitting a curve to
Value2 = Average - value1 * ratio = 133.3333 - 1 * 2 = 131.3333 the sales history data. This method may be useful when a product is in the transition between stages of a life
cycle. For example, when a new product moves from introduction to growth stages, the sales trend may
Forecast = (1 + n) * value1 + value2 = 4 * 1 + 131.3333 = 135.3333 accelerate. Because of the second order term, the forecast can quickly approach infinity or drop to zero
(depending on whether coefficient c is positive or negative). Therefore, this method is useful only in the short
November 2004 sales term.

Average of the previous three months Forecast specifications: The formulae finds a, b, and c to fit a curve to exactly three points. You specify n in the
processing option 7a, the number of time periods of data to accumulate into each of the three points. In this
= (140 + 131 + 114)/3 = 128.3333 example n = 3. Therefore, actual sales data for April through June are combined into the first point, Q1. July
through September are added together to create Q2, and October through December sum to Q3. The curve will
Summary of the previous three months with weight considered be fitted to the three values Q1, Q2, and Q3.

= (140 * 1) + (131 * 2) + (114 * 3) = 744 Required sales history: 3 * n periods for calculating the forecast plus the number of time periods required for
evaluating the forecast performance (PBF).
Difference between the values = 744 - 128.3333 * (1 + 2 + 3) = -25.9999

Value1 = Difference/Ratio = -25.9999/2 = -12.9999

 b = (Q2 - Q1) - 3c
Simulated 2005
Month 2004 Sales 2005 Sales 2006 Forecast Forecast
Substitute this equation for b into equation (3)
January 125 128 98
(3) Q3 = a + 3[(Q2 - Q1) - 3c] + c
February 132 117 98
a = Q3 - 3(Q2 - Q1)
March 115 115 98
Finally, substitute these equations for a and b into equation (1)
April 137 125 57
[Q3 - 3(Q2 - Q1)] + [(q2 - Q1) - 3c] + c = Q1
May 122 122 57
c = [(Q3 - Q2) + (Q1 - Q2)]/2
June 130 137 57

The Second Degree Approximation method calculates a, b, and c as follows:

July 141 129 1

August 128 140 1 a = Q3 - 3(Q2 - Q1) = 370 - 3(400 - 384) = 322

September 118 131 1 c = [(Q3 - Q2) + (Q1 - Q2)]/2 = [(370 - 400) + (384 - 400)]/2 = -23

October 123 114 136 b = (Q2 - Q1) - 3c = (400 - 384) - (3 * -23) = 85

November 139 119 136 Y = a + bX + cX^2 = 322 + 85*X + (-23)X^2

December 133 137 136 January thru March forecast (X=4):

(322 + 340 - 368)/3 = 294/3 = 98 per period

April thru June forecast (X=5):

A.9.1 Forecast Calculation
(322 + 425 - 575)/3 = 57.333 or 57 per period
Number of periods to include (processing option 7a) = 3 in this example

Use the previous (3 * n) months in three-month blocks: July thru September forecast (X=6):

(322 + 510 - 828)/3 = 1.33 or 1 per period

Q1(Apr - Jun) = 125 + 122 + 137 = 384

October thru December (X=7)

Q2(Jul - Sep) = 129 + 140 + 131 = 400

(322 + 595 - 1127/3 = -70

Q3(Oct - Dec) = 114 + 119 + 137 = 370

The next step involves calculating the three coefficients a, b, and c to be used in the forecasting formula Y = a A.9.2 Simulated Forecast Calculation
+ bX + cX^2
October, November and December, 2004 sales:
(1) Q1 = a + bX + cX^2 (where X = 1) = a + b + c
Q1(Jan - Mar) = 360
(2) Q2 = a + bX + cX^2 (where X = 2) = a + 2b + 4c
Q2(Apr - Jun) = 384
(3) Q3 = a + bX + cX^2 (where X = 3) = a + 3b + 9c
Q3(Jul - Sep) = 400
Solve the three equations simultaneously to find b, a, and c:
a = 400 - 3(384 - 360) = 328
Subtract equation (1) from equation (2) and solve for b
c = [(400 - 384) + (360 - 384)]/2 = -4
(2) - (1) = Q2 - Q1 = b + 3c
b = (384 - 360) - 3 * (-4) = 36