College of Accountancy

Western Leyte College of Ormoc City, Inc.

Operations Management Ms. Edelyn P. Aterado, CPA Class of 2011-2012

CHAPTER

Forecasting

What is Forecasting?
FORECAST:

A statement about the future value of a variable of interest such as demand. Forecasts affect decisions and activities throughout an organization Accounting, finance Human resources Marketing MIS Operations Product / service design

Uses of Forecasts
Accounting Finance Human Resources Marketing MIS Operations Product/service design Cost/profit estimates Cash flow and funding Hiring/recruiting/training Pricing, promotion, strategy IT/IS systems, services Schedules, MRP, workloads New products and services

Common in all forecasts

Assumes causal system past ==> future Forecasts rarely perfect because of randomness Forecasts more accurate for groups vs. individuals Forecast accuracy decreases as time horizon increases
I see that you will get an A this semester.

Elements of a Good Forecast

Timely

Reliable
l fu g in an e

Accurate
se u

Written

sy Ea

to

Steps in the Forecasting Process

The forecast

Step 6 Monitor the forecast Step 5 Prepare the forecast Step 4 Gather and analyze data Step 3 Select a forecasting technique Step 2 Establish a time horizon Step 1 Determine purpose of forecast

Types of Forecasts

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

Judgmental Forecasts

Executive opinions Sales force opinions Consumer surveys Outside opinion

Time Series Forecasts

Trend - long-term movement in data Seasonality - short-term regular variations in data Cycle wavelike variations of more than one years duration Irregular variations - caused by unusual circumstances Random variations - caused by chance

Forecast Variations
Irregular variation

Trend

Cycles
90 89 88 Seasonal variations

Naive Forecasts

Uh, give me a minute.... We sold 250 wheels last week.... Now, next week we should sell....
The forecast for any period equals the previous periods actual value.

Uses for Naive Forecasts

Stable time series data

F(t) = A(t-1) F(t) = A(t-n) F(t) = A(t-1) + (A(t-1) A(t-2))

Seasonal variations

Data with trends

Naive Forecasts

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

Techniques for Averaging

Moving average Weighted moving average Exponential smoothing

Moving Averages

Moving average A technique that averages a number of recent actual values, updated as new values become available.

The demand for tires in a tire store in the past 5 weeks were as n follows. Compute a three-period moving average forecast for demand in week 6. 83 80 85 90 94

MA

1 Ai i= n

Moving average & Actual demand

Moving Averages

Weighted moving average More recent values in a series are given more weight in computing the forecast.

Example:

For the previous demand data, compute a weighted average forecast using a weight of .40 for the most recent period, .30 for the next most recent, .20 for the next and .10 for the next. If the actual demand for week 6 is 91, forecast demand for week 7 using the same weights.

Exponential Smoothing

Ft = Ft-1 + (At-1 - Ft-1 )

The most recent observations might have the highest predictive value. Therefore, we should give more weight to the more recent time periods when forecasting.

Exponential Smoothing

Ft = Ft-1 + (At-1 - Ft-1 )

Weighted averaging method based on previous forecast plus a percentage of the forecast error A-F is the error term, is the % feedback

Example - Exponential Smoothing

Period Actual 1 83 2 80 3 85 4 89 5 92 6 95 7 91 8 90 9 88 10 93 11 92 12 0.1 83 82.70 82.93 83.54 84.38 85.44 86.00 86.40 86.56 87.20 87.68 Error -3.00 2.30 6.07 8.46 10.62 5.56 4.00 1.60 6.44 4.80 0.4 83 81.80 83.08 85.45 88.07 90.84 90.90 90.54 89.53 90.92 91.35 Error -3 3.20 5.92 6.55 6.93 0.16 -0.90 -2.54 3.47 1.08

Picking a Smoothing Constant

Exponential Smoothing
Actual Alpha=0.10 Alpha=0.40

100 95 Demand 90 85 80 75 70 2 3 4 5 6 7 8 9 10 11 Period

Problem 1

National Mixer Inc. sells can openers. Monthly sales for a seven-month period were as follows: Forecast September sales volume using each of the following:

Month Feb Mar Apr May Jun Jul Aug

Sales (1000) 19 18 15 20 18 22 20

A five-month moving average Exponential smoothing with a smoothing constant equal to .20, assuming a March forecast of 19. The naive approach A weighted average using .60 for August, .30 for July, and .10 for June.

Problem 2

A dry cleaner uses exponential smoothing to forecast equipment usage at its main plant. August usage was forecast to be 88% of capacity. Actual usage was 89.6%. A smoothing constant of 0.1 is used.

Prepare a forecast for September Assuming actual September usage of 92%, prepare a forecast of October usage

Problem 3

An electrical contractors records during the last five weeks indicate the number of job requests: Week: 1 2 3 4 5 Requests: 20 22 18 21 22 Predict the number of requests for week 6 using each of these methods:

Nave A four-period moving average Exponential smoothing with a smoothing constant of .30. Use 20 for week 2 forecast.

Review: forecast

Assumes causal system past ==> future Forecasts rarely perfect because of randomness Forecasts more accurate for groups vs. individuals Forecast accuracy decreases as time horizon increases

Review: forecast

Nave technique

Stable time series data Seasonal variations Data with trends

Averaging

Moving average Weighted moving average Exponential smoothing

Techniques for Trend

Develop an equation that will suitably describe trend, when trend is present. The trend component may be linear or nonlinear We focus on linear trends

Common Nonlinear Trends

Parabolic

Exponential

Growth

Linear Trend Equation

Ft

Ft = a + bt

Ft = Forecast for period t 0 1 t = Specified number of time periods a = Value of Ft at t = 0 b = Slope of the line

2 3 4 5

Example: Ft =10+2t. Interpret 10 and 2. Plot F

Example

Sales for over the last 5 weeks are shown below: 1 150 2 157 3 162 4 166 5 177

Week: Sales:

Plot the data and visually check to see if a linear trend line is appropriate. Determine the equation of the trend line Predict sales for weeks 6 and 7.

Line chart
Sales 180 175 170 165 Sales 160 155 150 145 140 135 1 2 3 Week 4 5 Sales

Calculating a and b
n (ty) - t y b = n t 2 - ( t) 2

y - b t a = n

Linear Trend Equation Example

t W 1 2 3 4 5 t 2 ( t ) = =
2

e e k

t 1 4 9 1 6 2 5

y S 1 1 1 1 1 5 5y a 5 5 6 6 7 = le s t y 0 1 5 0 7 3 1 4 2 4 8 6 6 6 6 4 7 8 8 5 8 1 y 2 = t 2 4 9

1 5t = 2 2 5

Linear Trend Calculation

5 (2499) - 15(812) 12495-12180 b = = = 6.3 5(55) - 225 275 -225

812 - 6.3(15) a = = 143.5 5

y = 143.5 + 6.3t

Linear Trend plot

Actual data 180 175 170 165 160 155 150 145 140 135 1 2 3 4 5 Linear equation

Recall: Problem 1

National Mixer Inc. sells can openers. Monthly sales for a seven-month period were as follows:

Month Feb Mar Apr May Jun Jul Aug

Sales (1000) 19 18 15 20 18 22 20

Plot the monthly data Forecast September sales volume using a line trend equation Which method of forecast seems least appropriate? What does use of the term sales rather than demand presume?

Line chart
Sales 20

M J

A Month A

M S

Problem 4

A cosmetics manufacturers marketing department has developed a linear trend equation that can be used to predict annual sales of its popular Hand & Foot Cream:

Ft = 80 + 15t where Ft = Annual sales (1000 bottles) t = 0 corresponds to 1990

Are annual sales increasing or decreasing? By how much? Predict annual sales for the year 2006 using the equation.

Techniques for Seasonality

Seasonality may refer to regular annual variation. There are two models:

Additive: expressed as a quantity (e.g., 20 units), which is added or subtracted from the series average Multiplicative: a percentage of the average or seasonal relative (e.g., 1.10), which is used to multiply the value of a series to incorporate seasonality.

Additive vs. multiplicative

Example

A furniture manufacturer wants to predict quarterly demand for a certain loveseat for periods 15 and 16, which happen to be the second and third quarters of a particular year. The series consists of both trend and seasonality. The trend portion of demand is projected using the equation

Ft = 124 + 7.5t

Quarter relatives are

Q1 = 1.20, Q2 = 1.10, Q3 = 0.75, Q4 = 0.95

Use this information to predict demand for periods 15 and 16.

Problem

A manager is using the equation below to forecast quarterly demand for a product:
Y(t) = 6,000 + 80t

where t = 0 at Q2 of last year

Quarter relatives are Q1 = .6, Q2 = .9, Q3 = 1.3, and Q4 = 1.2.

What forecasts are appropriate for the last quarter of this year and the first quarter of next year?

Problem

A manager of store that sells and installs hot tubs wants to prepare a forecast for January, February and March of 2007. Her forecasts are a combination of trend and seasonality. She uses the following equation to estimate the trend component of monthly demand:

Ft = 70 + 5t
Where t=0 is June of 2005. Seasonal relatives are 1.10 for Jan, 1.02 for Feb, and .95 for March. What demands should she predict?

Computing seasonal relatives

120 100 80 60 40 20 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

If your data appears to have seasonality, how do you compute the seasonal relatives?

Computing seasonal relatives

Calculate centered moving average for each period. Obtain the ratio of the actual value of the period over the centered moving average. Number of periods needed in a centered moving average = Number of seasons involved:

Monthly data: a 12-period moving average Quarterly data: a 4-period moving average

Example

The manager of a parking lot has computed the number of cars per day in the lot for three weeks. Using a seven-period centered moving average, calculate the seasonal relatives. Note that a seven period centered moving average is used because there are seven days (seasons) per week. See seasonal relatives1.xls

Problem 5

Obtain estimates of quarter relatives for these data:

Year:

3 1 2 3 4 45 54 84 88

4 1 58

Quarter: 1 2 3 4 1 2 3 4 Demand: 14 18 35 46 28 36 60 71

Problem

The manager of a restaurant believes that her restaurant does about 10% of its business on Sunday through Wednesday, 15% on Thursday night, 25% on Friday night, and 20% on Saturday night.

What seasonal relatives would describe this situation?

Note:

An alternative to deal with seasonality is to deseasonalize data. Deseasonalize = Remove seasonal component from data Gives clearer picture of the trend (nonseasonal component) Deseasonalize can be done by dividing each data point by its seasonal relative.

Forecasts: review

Judgmental - uses subjective inputs Time series - uses historical data assuming the future will be like the past

Nave approach Averaging Techniques for trend Trend and seasonality

Associative models - uses explanatory variables to predict the future

Associative Forecasting

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

LINEAR REGRESSION LINEAR REGRESSION

Suppose that J&T has a new product called AppleGlo, which is a household cleaner. This new product has been introduced into 14 sales regions over the last two years. The Advertising expenditure vs. the first year sales are shown in the table for each region. The company is considering introducing AppleGlo into two new regions, with the advertising campaign of $2.0 and $1.5 million. The company would like to predict what the expected first year sales of AppleGlo would be in each region.
F irs tY e a r A p p le G lo A d v e rtis in g F irs tY e a r E x p e n d itu re s S a le s ($ m illio n s ) ($ m illio n s ) R e g io n x y M a in e 1 .8 104 N e w H a m p s h ir e 1 .2 68 V e rm o n t 0 .4 39 M a s s a c h u s e tts 0 .5 43 C o n n e c tic u t 2 .5 127 R h o d e Isla n d 2 .5 134 N e w Y o r k 1 .5 87 N e w Je rs e y 1 .2 77 P e n n s y lv a n ia 1 .6 102 D e la w a re 1 .0 65 M a ry la n d 1 .5 101 W e s tV ir g in ia 0 .7 46 V irg in ia 1 .0 52 O h io 0 .8 33

LINEAR REGRESSION LINEAR REGRESSION

160 140 120

Sales ($Millions)

100 80 60 40 20 0 0 0.5 1 1.5 2 2.5 Advertising Expenditures ($Millions)

Questions:

How to relate advertising to sales? What is expected first-year sales if advertising expenditure is $1M? How confident are you in the estimate? How good is the fit?

Correlation

The correlation coefficient is a quantitative measure of the strength of the linear relationship between two variables. The correlation ranges from + 1.0 to - 1.0. A correlation of 1.0 indicates a perfect linear relationship, whereas a correlation of 0 indicates no linear relationship.

An algebraic formula for correlation coefficient

r=

[n( x ) ( x) ][n( y ) ( y ) ]
2 2 2 2

n xy x y

Simple Linear Regression

Simple linear regression analysis analyzes the linear relationship that exists between two variables.

y = a + bx
where: y = Value of the dependent variable x = Value of the independent variable a = Populations y-intercept b = Slope of the population regression line

Simple Linear Regression

The coefficients of the line are

b=

n xy x y n x ( x )
2 2

or

y b x a=
n

a = y bx

Problem 7

The manager of a seafood restaurant was asked to establish a pricing policy on lobster dinners. Experimenting with prices produced the following data:

Sold (y) 200 190 188 180 170 162 160 155 156 148 140 133

Price (x) 6.00 6.50 6.75 7.00 7.25 7.50 8.00 8.25 8.50 8.75 9.00 9.25

Create the scatter plot and determine if a linear relationship is appropriate. Determine the correlation coefficient and interpret it Obtain the regression line and interpret its coefficients.

Forecast Accuracy

Source of forecast errors:

Model may be inadequate Irregular variations Incorrect use of forecasting technique Random variation

Key to validity is randomness

Accurate models: random errors Invalid models: nonrandom errors

Key question: How to determine if forecasting errors are random?

Error measures

Error - difference between actual value and predicted value Mean Absolute Deviation (MAD)

Average absolute error Average of squared error Average absolute percent error

Mean Squared Error (MSE)

Mean Absolute Percent Error (MAPE)

MAD, MSE, and MAPE

MAD = Actual forecast n MSE = ( Actual forecast) n -1
2

MAPE =

Actual Forecast 100 Actual n

Example
Period 1 2 3 4 5 6 7 8 Actual 217 213 216 210 213 219 216 212 Fore cast 215 216 215 214 211 214 217 216 (A-F) 2 -3 1 -4 2 5 -1 -4 -2 |A-F| 2 3 1 4 2 5 1 4 22

(A-F)^2 (|A-F|/Actual)*1 4 0.9 9 1.4 1 0.4 16 1.9 4 0.9 25 2.2 1 0.4 16 1.8 76 10.2

MAD= MSE= MAPE=

2.75 10.86 1.28

Controlling the Forecast

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

Controlling the forecast

Control charts

Control charts are based on the following assumptions: when errors are random, they are Normally distributed around a mean of zero. Standard deviation of error is MSE 95.5% of data in a normal distribution is within 2 standard deviation of the mean 99.7% of data in a normal distribution is within 3 standard deviation of the mean Upper and lower control limits are often determine via 0 2 MSE or0 3 MSE

Example

Compute 2s control limits for forecast errors of previous example and determine if the forecast is accurate.
5.41

s = MSE = 3.295 2 s = 6.59

3.41 1.41 -0.59 0 -2.59 -4.59 -6.59

Errors are all between -6.59 and +6.59 No pattern is observed Therefore, according to control chart criterion, forecast is reliable

10

Problem 8

The manager of a travel agency has been using a seasonally adjusted forecast to predict demand for packaged tours. The actual and predicted values are

Pe rio d De m a nd Predic te d

Compute MAD, MSE, and MAPE. Determine if the forecast is working using a control chart with 2s limits. Use data from the first 8 periods to develop the control chart, then evaluate the remaining data with the control chart.

1 2 3 4 5 6 7 8 9 10 11 12 13

1 29 1 94 156 91 85 1 32 1 26 1 26 95 1 49 98 85 137

1 24 20 0 150 94 80 1 40 1 28 1 24 100 150 94 80 1 40

Problem

Given the following demand data, prepare a nave forecast for periods 2 through 10. Then determine each forecast error, and use those values to obtain 2s control limits. If demand in the next two periods turns out to be 125 and 130, can you conclude that the forecasts are in control?

Period Demand

10

118 117 120 119 126 122 117 123 121 124

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