FORECASTING Random Variations QUALITATIVE FORECASTING

Forecasting “Blips” in the data Jury of Executive Opinions

caused by chance and Under this method, the view of the top managers or a group with a high
⮚ the process of making future assumptions by analyzing unusual
trends, situations.
including historical and present data. level of expertise, often in combination with statistical models, are
⮚ the first step in the planning and scheduling process for most goods and synthesized to generate a consensual forecast.
service organizations. Delphi Method
Forecast The Delphi method is a procedure for acquiring informed judgments and
⮚ a prediction of what will occur in the future. opinions from knowledgeable individuals, using a series of questionnaires to
develop a consensus forecast about what will occur in the future. Responses
•Encompass the immediate acquired from the experts are analyzed by an independent party that will
Short
future and are concerned provide the feedback to the panel members.
Range

Medium
Forecasts
with the daily operations of Delphi Method
a Range
firm. The Delphi method is a procedure for acquiring informed judgments and
Forecasts
•Encompasses anywhere opinions from knowledgeable individuals, using a series of questionnaires to
from 1 or 2 months to 1 year develop a consensus forecast about what will occur in the future. Responses
Forecasting
Components and are concerned to yearly acquired from the experts are analyzed by an independent party that will
production and budgeting. provide the feedback to the panel members.
Long
Range •Encompasses a period longer Consumer Market Survey
Forecasts
than 1 or 2 years and are related Consumer market survey is an organized approach that uses surveys and
to expansion of facilities, secure other research techniques to determine what products and services
long-term financing, etc. customers want and will purchase, and to identify new markets and sources
of customers.
ORGANIZATIONAL DEPARTMENTS QUANTITATIVE FORECASTING
Marketing - They require forecasts to determine which new products or Time Series Forecasting
services to introduce or discontinue. Time series forecasting assumes that the future is a function of the past.
Supply Chain - They use forecasts to make production, logistical plans and Thus, historical data are used to predict the future using sequences with
procurement. equal periods.
Sales - They use forecasts to make sales plans because quotas are generally Causal Forecasting Methods
based on estimates of future sales. Causal forecasting methods involve the determination of factors that relate
Finance and Accounting - They use forecasts to make financial plans such as to the variable you are trying to forecast.
budgeting and capital expenditures. COMPONENTS OF TIME SERIES FORECASTING

Cycle
FORECASTING APPROACHES Trend Seasonality
A pattern of the data that
Qualitative Forecasting - Qualitative forecasting methods are considered to Gradual upward or A data pattern that varies in length and occurs
repeats itself after a every several years usually
be highly subjective and judgmental. downward movement period of days, weeks,
of the data over time. 2 to 10 years.
Quantitative Forecasting - Quantitative forecasting methods make use of months or quarters
historical data. Cycle
A pattern of the data that varies in length and occurs every
several years usually 2 to 10 years.
 QUANTITATIVE FORECASTING MODELS NAÏVE MODEL

ed MonthActual Sales Naïve Absolute Absolute Percent

i ght Naï
We ving ve Mod Forecast DeviationError
Mo rage el
Ave Jan199 0.00%
Feb200 199 1 = (200-199) 0.5% = 1/200*100

Ex moo
Mar150 200 50 = (150-200) 33.3% = 50/150*100
era g

S
ge

po thin
Av ovin

ne g
Apr110 150 40 = (110-150) 36.4% = 40/110*100

nti
M

al
Month Actual Naïve Absolute % Month ActualNaïve Absolute %
MOVING AVERAGE Sales Forecast Error SalesForecast Error
Jan199 July200
Month Actual Sales3-Month Moving Average Feb200 Aug100
Jan15
Mar150 Sep51
Feb12
March15 Apr110 Oct100
April1815 + 12 + 15 = 14 May300 Nov53
3 June400 Dec120
May 2118 + 15 + 1 = 15
3 EXPONENTIAL SMOOTHING
WEIGHTED MOVING AVERAGE ⮚ This method uses a series of exponentially weighted moving averages
techniques where more weight is given to the recent data.
Example:
Weighted Applied Period ⮚ The weights assigned to the values change so that the most recent value
20 3 months ago receives the highest weight, the previous value receives the second-
.302 months ago
.50last month
highest weight, and so on, with the first value receiving the lowest
1.00sum of weights weight.
⮚ Throughout the series, each exponentially smoothed value is supported
Month Actual Sales 3-month WMA
Jan 15 by the belief that the future is more dependent on the recent past than
Feb 12 on the distant past.
Mar 15
April 18 (15x.50)+(12x.30)+(15x.20) = 14 ⮚ This method is known to be useful on random historical data with no
May 21 (18x.50)+(15x.30)+(12x.20) = 15.9 seasonal fluctuations.

Month Actual Sales Month Actual Sales

Jan15 July28
Feb12 Aug32
March15 Sep30
April18 Oct20
May21 Nov18
June25 Dec16