Operations Management (MD021)

1
Operations Management(MD021)

• Forecasting

2
Agenda

• Background on Forecasting
• Forecasting Techniques
• Assessing Forecast Accuracy

3
Background on Forecasting
4
Forecasts can be made for many phenomena

• Forecasts are a statement (a prediction) about
  the future value of a variable of interest
• Customer demand for goods/services
• Aggregate demand for material inputs
• Income
• Interest rate
• Technology shifts
• Grades/school performance

5
Good forecasting is based on science, art, and
luck

• Forecasting is both a science and an art
• Science
• forecasting equations
• statistics/regression analysis
• Art
• good at guessing (judgmental forecasting)
• good at picking the correct type of forecasting
  equation

6
Two operational environments in which we might
use forecasts

• Forecast future demand and build to forecast
  (PLAN AND BUILD)
• OR
• Dont forecast Use flexible operations and wait
  for demand to occur before building anything
  (SENSE AND RESPOND)

7
Performance objectives when forecasting

• Cost
• Generally, it takes more to create better
  forecasts
• Time
• Want forecast fast, to be able to respond quickly
• Faster forecasting costs more
• Accuracy
• More accurate forecast usually takes more time
  and more

8
Functional areas create many different forecasts
9
Forecasts from one functional area can affect
other functional areas

• Forecasts developed within one functional area
  can affect decisions and activities throughout an
  organization
• Marketing forecast average demand develops new
  ads with lower price updates forecasts ? demand
  is inspected to increase a lot
• Operations will need to have sufficient
  machines to satisfy demand
• Accounting, Finance need to provide capital for
  additional machines
• Human Resources need to hire more people
• MIS need to have sufficient computer resources
  to process transactions

10
Critical assumptions behind forecasts

• Assume that the same underlying causal system
  existing in the past will exist in the future
• Previous phenomena work the same way as future
  phenomena
• Forecasts are rarely perfect
• Randomness in data
• Weird, unexpected events can take place
• Aggregate forecasts (for groups) tend to be more
  accurate than forecasts for individual items
• All Barbie dolls vs. Vegas Barbie doll
• Quarterly demand vs. Daily demand
• Forecast accuracy decreases as time horizon
  increases
• One quarter forecast vs. Five year forecast

11
Elements of a Good Forecast
12
Steps in the Forecasting Process
13
Forecasting Techniques
14
Choosing a Forecasting Technique

• No single technique works in every situation
• Two most important factors
• Cost
• Accuracy
• Other factors include the availability of
• Historical data
• Computers
• Time needed to gather and analyze the data
• Forecast horizon

15
Types of Forecasts

• Judgmental - uses subjective inputs
• Time series - uses historical data assuming the
  future will be like the past
• Associative models - uses historical explanatory
  variables to predict the future

16
Judgmental Forecasting
17
Judgmental Forecasts

• Executive opinions
• Sales force opinions
• Consumer surveys
• Outside opinion
• Delphi method
• Opinions of managers and staff
• Achieves a consensus forecast

18
Time Series Forecasting
19
Time series data can be broken up into several
components

• Data Trend Cyclical Seasonality Irregular
  Random
• Trend - long-term movement in data
• Cycle wavelike variations of more than one
  years duration
• Seasonality - short-term regular variations in
  data
• Irregular variations - caused by unusual
  circumstances
• Random variations - caused by chance white
  noise residual variation

20
Time series data contain several components
Irregularvariation
Trend
Cycles
90
89
88
Seasonal variations
21
Naive Forecasts
22
Advantages of Naïve Forecasts

• Simple to use
• Virtually no cost
• Quick and easy to prepare
• Data analysis is nonexistent
• Easily understandable
• Cannot provide high accuracy
• Can be a standard for accuracy

23
Uses for Naïve Forecasts

• Stable time series data
• F(t) A(t-1)
• Forecast at time t is the Actual value from time
  t-1
• Time t tomorrow Time t-1 today
• Seasonal variations
• F(t) A(t-n)
• Data with trends
• F(t) A(t-1) (A(t-1) A(t-2))

24
Techniques for Averaging

• Moving average
• Weighted moving average
• Exponential smoothing

25
Moving Averages

• Moving average A technique that averages a
  number of recent actual values, updated as new
  values become available.
• Weighted moving average More recent values in a
  series are given more weight in computing the
  forecast.

26
Simple Moving Average
Actual
MA5
MA3
27
Exponential Smoothing
Ft Ft-1 ?(At-1 - Ft-1)

• Premise-- The most recent observation may have
  the highest predictive value. Therefore, we
  should give more weight to more recent time
  periods when forecasting.
• Weighted averaging method based on previous
  forecast plus a percentage of the forecast error
• A-F is the error term,
• ? is the feedback, and is between 0 and 1

28
Picking a Smoothing Constantfor Exponential
Smoothing
29
Trend Forecast

• Linear Trend a long-term movement up, or a
  long-term movement down
• Curvilinear Trend parabolic patterns,
  exponential patterns, growth curve (S-curve)

30
Common Nonlinear Trends
POTENTIAL USES
Demand growth and decline (and vice versa)
End of product life
Product introduction Technology adoption
31
Linear Trend Equation

• Ft Forecast for period t
• t Specified number of time periods
• a Value of Ft at t 0
• b Slope of the line

32
Linear Trend Calculating a and b
33
Linear Trend Equation Example
34
Linear Trend Calculation
35
Trend-Adjusted Exponential Smoothing

• Adjusts the Exponential Smoothing forecast for a
  visible trend pattern

TAFt1 St Tt
where
St TAFt ?(At - TAFt)
Tt Tt-1 ?(TAFt TAFt-1 - Tt-1)
36
Forecasts Incorporating Seasonal Multipliers

• When seasonality is present, seasonal multipliers
  can be used to create seasonally adjusted
  forecasts (SAF)
• Multipliers (seasonal relatives)
  increase/decrease the forecast based on a
  periods seasonality

SAFt Ft (SeasonalRelativet)
37
Associative Forecasting Using Linear Regression
38
Associative Forecasting

• Predictor variables - used to predict values of
  variable interest
• Linear Regression - technique for fitting a line
  to a set of points
• Least squares line - minimizes sum of squared
  deviations around the line

39
Linear Regression Equation
y a bx

• Ft Forecast
• x predictor variable
• a constant
• b Slope of the line

40
Linear Model Seems Reasonable
A straight line is fitted to a set of sample
points.
41
Assessing Forecast Accuracy
42
Many Potential Sources of Forecast Errors

• Model may be inadequate
• Irregular variations
• Incorrect use of forecasting technique

Forecasters need to make sure that the above are
not affecting their forecast
43
Forecast Accuracy

• Forecast Error - difference between the actual
  value and predicted value for a given time period
• Mean Absolute Deviation (MAD)
• Average absolute error
• Mean Squared Error (MSE)
• Average of squared error
• Mean Absolute Percent Error (MAPE)
• Average absolute percent error

et At - Ft
44
MAD, MSE, and MAPE
Actualt
/ Actualt100)
?(
Forecastt
?
MAPE


n
45
Example 10
46
Controlling the Forecast with Control Charts

• Control chart
• A visual tool for monitoring forecast errors
• Used to detect non-randomness in errors
• Forecasting errors are in control if
• All errors are within the control limits
• No patterns, such as trends or cycles, are present

47
Control Charts for Forecast Errors
s (MSE)0.5
UCL 0 zs
LCL 0 - zs
48
Tracking Signal

• Ratio of cumulative error to MAD
• Tracks period-by-period whether there is a
  systematic bias in the forecast
• Bias tendency for forecast to be persistently
  above or below actual values
• Zero is ideal value for TSt.
• If TSt gt 4 or TSt lt -4 then there appears to be
  bias in the forecast, and corrective action
  should be taken.