Forecasting

BY: MARK JASON P. JIMENEZ

What is Forecasting?
Forecasting is a technique that uses historical data as inputs to make informed estimates that are
predictive in determining the direction of future trends. Businesses utilize forecasting to determine
how to allocate their budgets or plan for anticipated expenses for an upcoming period of time.
Principles of Forecasting
Forecasts are rarely perfect.
Forecasts are more accurate for groups or families of items rather
than for individual items.
Forecasts are more accurate for shorter than longer time horizons.
 5 Basic steps in Forecasting Process
1. Determine the purpose of the forecast
2. Evaluate and analyze appropriate data
3. Select and test the forecasting model
4. Generate the forecast
5. Monitor forecast accuracy
Types of Forecasting Method

Qualitative Method
Quantitative Method
Types of Qualitative Forecasting
Methods
1. Salesforce estimates
2. Executive Opinion
3. Market Research
4. The Delphi Method
 Type Characteristics Strengths Weaknesses

Executive A group of managers meet Good for strategic or new- One person’s
Opinion and come up with a product forecasting. opinion can
forecast. dominate the
forecast.

Market Uses surveys and interviews Good determinant of It can be difficult to

Research to identify customer customer preferences. develop a good
preferences. questionnaire.

Delphi method Seeks to develop a Excellent for forecasting long term Time – consuming
consensus among a group of product demand, technological to develop.
experts. changes and scientific advances

Salesforce Based on sales Have basis data to set forecast figures Uncontrollable
Estimate data. for planning purposes. change of demand.
Two main approaches to Quantitative
Forecasting

Time series analysis

Causal modelling techniques
Five basic patterns of Time Series Model
• Horizontal. The fluctuation of data around a constant mean.
• Trend. The systematic increase or decrease in the mean of the series over time.
• Seasonal. A repeatable pattern of increases or decreases in demand, depending
on the time of day, week, month, or season.
• Cyclical. The less predictable gradual increases or decreases in demand over
longer periods of time (years or decades).
• Random. The unforecastable variation in demand.
Forecast errors
are defined as providing important clues for making better
forecasts.
Cumulative sum of Forecast Errors
(CFE)
Measures the total forecast error
CFE is also called the bias error and results from consistent
mistakes—the forecast is always too high or too low.
This type of error typically causes the greatest disruption to planning
efforts.
CFE = -31
Average Forecast Error
sometimes called the mean bias
n = number of periods
-31 / 10
= -3.1
mean bias
Absolute Error
The absolute value (or modulus) | x | of a real number x is the non-negative value
of x without regard to its sign.
Absolute Percent Error
l E l / Dt x 100 = l Pct Error l
2 / 39 = 0.05128205128205128205128205128205
0.05128205128205128205128205128205 x 100 =

5.1282051282051282051282051282051

Or

5.128%
mean squared error (MSE)
mean absolute percent error
(MAPE)
mean absolute deviation
(MAD)
n = number of periods
MSE
879 / 10 = 87.9

MAD
81 / 10 = 8.1

MAPE
170.621 / 10 =
17.062
The naïve method
is one of the simplest forecasting models. It assumes that the next period’s forecast is equal to the
current period’s actual.
Example
A restaurant is forecasting sales of chicken dinners for the month of April. Total sales of chicken
dinners for March were 450. If management uses the naïve method to forecast, what is their forecast of
chicken dinners for the month of April?
 The statistical techniques that do have an
adaptive quality in estimating the average in
a time series are; simple moving averages,
weighted moving averages, and
exponential smoothing.
Simple Moving Average Method
A time-series method used to estimate the average of a demand time
series by averaging the demand for the n most recent time periods.
Example
New Tools Corporation is forecasting sales for its classic product, Handy-Wrench. Handy- Wrench
sales have been steady, and the company uses a simple mean to forecast. Weekly sales over the past
five weeks are available. Use the mean to make a forecast for week 8.
F7+1 = 53 + 52 + 50 + 65 + 35
5

F8 = 255
5

F8 = 51
Example
Three-period Moving Average
Weighted Moving Average Method
A time-series method in which each historical demand in the average can have
its own weight; the sum of the weights equals 1.0.
Example
A manager at Fit Well department store wants to forecast sales of swimsuits for August using a
three- period weighted moving average. Sales for May, June, and July are as follows:

The manager has decided to weight May (0.50), June (0.30), and July (0.20).
Note: Remember that to compute a weighted moving average you need to multiply
each observation by its corresponding weight. These values are then summed in
order to get a weighted average.

The forecast for August is computed as follows:

Exponential Smoothing Method
Example
The Crab Tamales Mexican restaurant uses exponential smoothing to forecast
monthly usage of tabasco sauce. Its forecast for September was 250 bottles,
whereas actual usage in September was 450 bottles. If the restaurant’s managers
use an α of 0.70, what is their forecast for October?
Given:
October
450

250
0.70
 Trend-Adjusted Exponential Smoothing
1. The first smooths out the level of the series,
2. The second smooths out the trend, and
3. The third generates a forecast by adding up the findings from the
first two equations.
Step 1 Smoothing the Level of the Series
𝑆𝑡 = α𝐴𝑡 + (1- α) (𝑆𝑡−1 + 𝑇𝑡−1)

Step 2 Smoothing the Trend

𝑇𝑡 = β (𝑆𝑡 - 𝑆𝑡−1) + (1 – β)
𝑇
𝑡−1

Step 3 Forecast Including Trend

𝐹𝐼𝑇𝑡+1 = 𝑆𝑡 + 𝑇𝑡

Where:
𝐹𝐼𝑇𝑡+1 = forecast including trend for next period, t + 1
𝑆𝑡 = exponentially smoothed average of the time series in period t
𝑇𝑡 = exponentially smoothed trend of the time series in period t
α = smoothing coefficient of the level
β = smoothing coefficient of the trend
Example
Note: When solving this type of problem, always begin by identifying the
information that is given in the problem.

Given information:
𝑆𝐽𝑢𝑛𝑒 = 59 gallons/month
𝑇𝐽𝑢𝑛𝑒 = 17 gallons/month
𝐴𝐽𝑢𝑙𝑦 = 64 gallons α = 0.20
β = 0.10
 Step 1 Smoothing the Level of the Series
𝑆𝑡 = α𝐴𝑡 + (1- α) (𝑆𝑡−1 + 𝑇𝑡−1)
𝑆
𝐽𝑢𝑙𝑦 = α𝐴𝐽𝑢𝑙𝑦 + (1- α) (𝑆𝐽𝑢𝑛𝑒 + 𝑇𝐽𝑢𝑛𝑒 )
𝑆𝐽𝑢𝑙𝑦 = (0.20) (64) + (0.80) (59+17)
𝑆𝐽𝑢𝑙𝑦 = 12.8 + (0.80) (76)
𝑆𝐽𝑢𝑙𝑦 = 12.8 + 60.8
𝑆𝐽𝑢𝑙𝑦 = 73.6

Given information:

𝑆𝐽𝑢𝑛𝑒 = 59 gallons/month

𝑇𝐽𝑢𝑛𝑒 = 17 gallons/month
𝐴𝐽𝑢𝑙𝑦 = 64 gallons

α = 0.20

β = 0.10
 Step 2 Smoothing the Trend
𝑇𝑡 = β (𝑆𝑡 - 𝑆𝑡−1) + (1 – β) 𝑇𝑡−1
𝑇
𝐽𝑢𝑙𝑦 = β (𝑆𝐽𝑢𝑙𝑦 - 𝑆𝐽𝑢𝑛𝑒) + (1 – β) 𝑇𝐽𝑢𝑛𝑒
𝑇𝐽𝑢𝑙𝑦 = (0.1) (73.6 - 59) + (0.90) 17
𝑇𝐽𝑢𝑙𝑦 = (0.1) (14.6) + 15.3
𝑇𝐽𝑢𝑙𝑦 = 1.46 + 15.3
𝑇𝐽𝑢𝑙𝑦 = 16.76

Step 3 Forecast Including Trend

𝐹𝐼𝑇𝑡+1 = 𝑆𝑡 + 𝑇𝑡
Given information: 𝐹𝐼𝑇
𝐴𝑢𝑔𝑢𝑠𝑡 = 𝑆𝐽𝑢𝑙𝑦 + 𝑇𝐽𝑢𝑙𝑦
𝑆𝐽𝑢𝑛𝑒 = 59 gallons/month 𝐹𝐼𝑇𝐴𝑢𝑔𝑢𝑠𝑡 = 73.6 + 16.76
𝑇𝐽𝑢𝑛𝑒 = 17 gallons/month 𝐹𝐼𝑇𝐴𝑢𝑔𝑢𝑠𝑡 = 90.36 gallons
𝐴𝐽𝑢𝑙𝑦 = 64 gallons

α = 0.20

β = 0.10
Linear trend line is a time series technique that computes a forecast
with trend by drawing a straight line through a set of data.

A linear trend line uses the following equation to generate a

forecast:
Y = a + bX

Where:
Y = forecast for period X
X = the number of time periods from X = 0
a = value of Y at X = 0 (Y intercept)
b = slope of the line
The steps for computing the forecast using a linear trend line are as
follows
Example:
A manufacturer has plotted product sales over the past four
weeks. Use a linear trend line to generate a forecast for week 5.

Note: Remember to follow the four steps given in the text for
generating a forecast using a linear trend line.
 SEASONAL PATTERN
are regularly repeating upward or downward movements in demand
measured in periods of less than one year (hours, days, weeks, months,
or quarters).

• For example, enrollment is usually much higher in the fall than

in the summer.
• Other examples of seasonality include sales of turkeys before
Thanksgiving or ham before Easter, sales of
The amount of seasonality is the extent to which
actual values deviate from the average or mean of
the data.

• multiplicative seasonality, in which the

seasonality is expressed as a percentage of the
average. The percentage by which the value for
each season is above or below the mean is a
seasonal index.
• For example, if enrollment for the fall semester at
your university is 1.30 of the mean, the fall
enrollment is 30 percent above the average.
Similarly, if enrollment for the summer semester is
0.70 of the mean, then summer enrollment is 70
percent of the average
Procedure for computing quarterly seasonality that
lasts a year consists of the following steps:
Step 1 Calculate the Average Demand for Each Quarter or “Season.”
This is done by dividing the total annual demand by 4 (the number of
seasons per year).

Step 2 Compute a Seasonal Index for Every Season of

Every Year for Which You
Have Data. This is done by dividing the actual demand for each season by
the average demand per season (computed in Step 1).
Step 3 Calculate the Average Seasonal Index for Each Season. For each
season, compute the average seasonal index by adding up the seasonal index
values for that season and dividing by the number of years.

Step 4 Calculate the Average Demand per Season for Next Year. This
could be done by using any of the methods used to compute annual demand.
Then we would divide that by the number of seasons to determine the
average demand per season for next year.

Step 5 Multiply Next Year’s Average Seasonal Demand by Each

Seasonal Index.
This will produce a forecast for each season of next year.
Example:
CVs University wants to develop forecasts for next year’s quarterly
enrollment. It has collected quarterly enrollments for the past two years. It has
also forecast total annual enrollment for next year to be 90,000 students. What
is the forecast for each quarter of next year?

Note: You can see that the data exhibit a seasonal pattern, with each quarter
representing a “season.” To compute the forecast for each quarter of next year, follow
the five steps given in the text on forecasting with seasonality.
Step 1 Calculate the Average Demand for Each Quarter or “Season”

We do this by dividing the total annual demand for each year Year
1: 80/4 = 20
Year 2: 84/4 = 21

Step 2 Compute a Seasonal Index for Every Season of Every Year

for Which You Have Data.
To do this we divide the actual demand for each season by the
average demand per season.
Step 3 Calculate the Average Seasonal Index for Each
Season.
You can see that the seasonal indexes vary from year to year for the same
season. The simplest way to handle this is to compute an average index, as
follows:

Step 4 Calculate the Average Demand per Season for Next Year. We are told
that the university forecast annual enrollment for the next year to be 90,000
students. The average demand per season, or quarter, is
90,000/4 = 22,500
Step 5 Multiply Next Year’s Average Seasonal Demand by Each
Seasonal Index. This last step will give us the forecast for each quarter
of next year:
 Causal Methods

❖ used when historical data are available and the relationship between the factor to be
forecasted and other external or internal factors (e.g., government actions or advertising
promotions) can be identified.

❖ It also good for predicting turning points in demand and for preparing long-range forecast
Linear regression
✔ A causal method in which one variable
dependent variable) is (the to one
independent variables by a linear equation.
related or more

▪ dependent variable. The variable that one wants to forecast

▪ independent variables. Variables that are assumed to affect the

dependent variable and thereby “cause” the results observed in the
past.
In the simplest linear regression
models, the dependent variable is a
function of only one independent
variable and, therefore, the
theoretical relationship is a straight
line:
Correlation coefficient - Statistic that measures the direction
and strength of the linear relationship between two variables.
• which measures the direction and strength of the linear
relationship between the independent and dependent
variables.

Three measures commonly reported in linear regression;

(1) the sample correlation coefficient, (r)

(2) the sample coefficient of determination, (r²) and

(3) the standard error of the estimate. (𝑆𝑥𝑦)

MULTIPLE REGRESSION is an extension of linear regression.

• unlike in linear regression where the dependent variable is related to

one independent variable, multiple regression develops a relationship
between a dependent variable and multiple independent variables.

The general formula for multiple regression is as follows:

Y = 𝐵0 + 𝐵1 𝑋1 + 𝐵2 𝑋2 + …… + 𝐵𝐾 𝑋𝐾
Y = dependent variable
𝐵0 = the Y intercept
𝐵1 . . . 𝐵𝐾 = coefficients that represent the influence of the independent
variables on the dependent variable
𝑋1 . . . 𝑋𝐾 = independent variables
Tracking Signals

is a measure that indicates whether a method of forecasting is

accurately predicting actual changes in demand.
Tool used to monitor the quality of a forecast.

• It is computed as the ratio of the algebraic sum of the

forecast errors divided by MAD:
Tracking signal = algebraic sum of forecast errors / MAD Tracking
signal = Σ (actual – forecast) / MAD
Collaborative Planning, Forecasting, And
Replenishment (CPFR)

- is a collaborative process between two trading

partners that establishes formal guidelines for joint
forecasting and planning.

- The premise behind CPFR is that companies can be

more successful if they join forces to bring value to their
customers, share risks of the marketplace, and improve
their performances.
CPFR a nine-step process:
1. Establish collaborative relationships.
2. Create a joint business plan.
3. Create a sales forecast.
4. Identify exceptions for sales forecasts.
5. Resolve/collaborate on exceptions to sales
forecasts.
6. Create order forecast.
7. Identify exceptions for order forecast
8. Resolve/collaborate on exceptions to order forecast.
9. Generate order.
Forecast across the Organization
Finance needs forecasts to project cash flows and capital
requirements.

Human resources uses forecasts to anticipate hiring and training

needs.

Marketing is an important source for sales forecast information

because it is closest to external customers.

Operations and supply chain managers need forecasts to plan output

levels, purchases of services and materials, workforce and output
schedules, inventories, and long-term capacities.