# Sales Projection Document

Document Sample

```					  Chapter 13

Analyzing and
Forecasting Time
Series Data
Chapter 13 - Chapter Outcomes
After studying the material in this chapter, you
should be able to:
•Apply the basic steps in developing and
implementing forecasting models.
•Identify the components present in a time series.
•Use smoothing-based forecasting models
including, single and double exponential
smoothing.
•Apply trend-based forecasting models, including
linear trend, nonlinear trend, and seasonally
Forecasting

Model specification refers to the
process of selecting the forecasting
technique to be used in a particular
situation.
Forecasting

Model fitting refers to the process of
determining how well a specified
model fits its past data.
Forecasting

Model diagnosis refers to the process
of determining how well the model
fits the past data and how well the
model’s assumptions appear to be
satisfied.
Forecasting

The forecasting horizon refers to the
number of future periods covered by
the forecast, sometimes referred to as
Forecasting

The forecasting period refers to the
unit of time for which the forecasts
Forecasting

The forecasting interval refers to the
frequency with which the new
forecasts are prepared.
Forecasting

Time-Series data are data which are
measured over time. In most applications
the period between measurements is
uniform.
Components of Time Series
Data

• Trend Component
• Seasonal Component
• Cyclical Component
• Random Component
Time Series Forecasting

A time-series plot is a two-dimensional
plot of the time series. The vertical axis
measures the variable of interest and
the horizontal axis corresponds to the
time periods.
\$ x 1,000
Ja
nu

1000

0
100
200
300
400
500
600
700
800
900
ar
y
M
ar
ch

M
ay

Se Jul
pt     y
em
No ber
ve
m
be
Ja r
nu
ar
y
M
ar
ch

M
ay

Se Jul
pt     y
em
No ber
ve
m
be
Ja r
nu
ar
y
M
ar
ch

M
ay
(Figure 13-1)

Se Jul
pt     y
em
No ber
ve
m
be
Ja r
nu
ar
Time-Series Plot

y
M
ar
ch

M
ay

Se Jul
pt     y
em
No ber
ve
m
be
r
Time Series Forecasting

A linear trend is any long-term
increase or decrease in a time
series in which the rate of
change is relatively constant.
Time Series Forecasting

A seasonal component is a pattern
that is repeated throughout a time
series and has a recurrence period
of at most one year.
Time Series Forecasting

A cyclical component is a pattern
within the time series that repeats
itself throughout the time series and
has a recurrence period of more than
one year.
Time Series Forecasting

The random component refers to
changes in the time-series data that are
unpredictable and cannot be associated
with the trend, seasonal, or cyclical
components.
Trend-Based Forecasting
Techniques
LINEAR TREND MODEL

yt   0  1t   t
where:
yi = Value of trend at time t
0 = Intercept of the trend line
1 = Slope of the trend line
t = Time (t = 1, 2, . . . )
Linear Trend Model
(Example 13-2)
Taft Ice Cream Sales
Year       t         Sales
1991          1   \$300,000
1992          2   \$295,000
1993          3   \$330,000
1994          4   \$345,000
1995          5   \$320,000
1996          6   \$370,000
1997          7   \$380,000
1998          8   \$400,000
1999          9   \$385,000
2000         10   \$430,000
Linear Trend Model
(Example 13-2)

\$500,000
Taft Sales
\$450,000

\$400,000

\$350,000

\$300,000
Sales

\$250,000

\$200,000

\$150,000

\$100,000

\$50,000

\$0
1990   1991   1992   1993   1994   1995   1996   1997   1998   1999   2000   2001
Year
Linear Trend Model
(Example 13-2)
LEAST SQUARES EQUATIONS
 t y
 ty  n                         t
t
b1   
t   ty2
2

n

b0   
y       t
 b1
t
where:            n         n
n = Number of periods in the time series
t = Time period independent variable
yt = Dependent variable at time t
Linear Trend Model
(Example 13-2)

SUMMARY OUTPUT

Regression Statistics
Multiple R              0.955138103
R Square                0.912288796
Standard Error          14513.57776
Observations                     10

ANOVA
df            SS           MS          F          Significance F
Regression                        1    17527348485 17527348485 83.20841575 1.67847E-05
Residual                          8     1685151515 210643939.4
Total                             9    19212500000

Coefficients Standard Error t Stat  P-value      Lower 95%   Upper 95%   Lower 95.0% Upper 95.0%
Intercept                277333.3333 9914.661116 27.97204363 2.88084E-09   254470.069 300196.5977 254470.069 300196.5977
t                        14575.75758 1597.892322 9.121864708 1.67847E-05 10891.00889 18260.50626 10891.00889 18260.50626
Linear Trend Model
(Example 13-2)
Taft Linear Trend Model

\$500,000

\$450,000

\$400,000

\$350,000

\$300,000
y = 14576t + 277333
Sales

\$250,000

\$200,000

\$150,000

\$100,000

\$50,000

\$0
1990   1991   1992   1993   1994   1995   1996   1997   1998   1999   2000   2001
Year
Linear Trend Model
- Forecasting -

Trend Projection:

Ft  277 ,333 .33  14 ,575 .76 (t )
Forecasting Period t = 11:

Ft  277 ,333 .33  14 ,575 .76 (11)
\$437 ,666 .69
Linear Trend Model
- Forecasting -

MEAN SQUARE ERROR

MSE 
(y       t    Ft )   2

n
where:
yt = Actual value at time t
Ft = Predicted value at time t
n = Number of time periods
Linear Trend Model
- Forecasting -

MEAN ABSOLUTE DEVIATION

| y         t    Ft |
n
where:
yt = Actual value at time t
Ft = Predicted value at time t
n = Number of time periods
Linear Trend Model
- Forecasting -

MEAN ABSOLUTE DEVIATION

Forecast Bias 
(y   t    Ft )
or:
n

Forecast Bias 
 (error )
n
Nonlinear Trend Models
(Example)

yt   0  1t   t
2
Trend-Based Forecasting

A seasonal index is a number used
to quantify the effect of seasonality
for a given time period.
Trend-Based Forecasting

MUTIPLICATIVE TIME SERIES MODELS

yt  Tt  St  Ct  I t
where:
yt = Value of the time series at time t
Tt = Trend value at time t
St = Seasonal value at time t
Ct = Cyclical value at time t
It = Residual or random value at time t
Trend-Based Forecasting

A moving average is the average
of n consecutive values in a time
series.
Trend-Based Forecasting

RATIO-TO-MOVING-AVERAGE
yt
St  I t 
Tt  Ct
Trend-Based Forecasting

DESEASONALIZATION
yt
Tt  Ct  I t 
St
Trend-Based Forecasting

A seasonally unadjusted forecast is a
forecast made for seasonal data that
does not include an adjustment for
the seasonal component in the time
series.
Steps in the Seasonal
• Compute each moving average from the k
appropriate consecutive data values.
• Compute the centered moving averages.
• Isolate the seasonal component by
computing the ratio-to-moving-average
values.
• Compute the seasonal indexes by
averaging the ratio-to-moving-averages
for comparable periods.
Steps in the Seasonal
(continued)

• Normalize the seasonal indexes.
• Deseasonalize the time series.
• Use least-squares regression to develop the
trend line using the deseasonalized data.
• Develop the unadjusted forecasts using trend
projection.
• Seasonally adjust the forecasts by
multiplying the unadjusted forecasts by the
appropriate seasonal index.
Forecasting Using Smoothing
Techniques

Exponential smoothing is a time-series
smoothing and forecasting technique that
produces an exponentially weighted
moving average in which each smoothing
calculation or forecast is dependent upon
all previously observed values.
Forecasting Using Smoothing
Techniques
EXPONENTIAL SMOOTHING MODEL

Ft 1  Ft   ( yt  Ft )
or::
Ft 1  yt  (1   ) Ft
where:
Ft+1= Forecast value for period t + 1
yt = Actual value for period t
Ft = Forecast value for period t
 = Alpha (smoothing constant)
Forecasting Using Smoothing Techniques
DOUBLE EXPONENTIAL SMOOTHING MODEL
Ct  yt  (1   )(Ct 1  Tt 1 )
Tt   (Ct  Ct 1 )  (1   )Tt 1
Ft 1  Ct  Tt
where:
yt = Actual value in time t
 = Constant-process smoothing constant
 = Trend-smoothing constant
Ct = Smoothed constant-process value for period t
Tt = Smoothed trend value for period t forecast
value for period t
Ft+1= Forecast value for period t + 1
t = Current time period
Key Terms
• Alpha ()            • Forecast Error
• Beta ()             • Forecasting
• Cyclical Component   • Forecasting Horizon
• Deseasonalizing      • Forecasting Interval
• Double Exponential   • Forecasting Period
Smoothing            • Linear Trend
• Exponential          • Mean Absolute
• Forecast Bias        • Mean Squared Error
(MSE)
Key Terms
(continued)

• Model Diagnosis       • Ratio-To-Moving-
• Model Fitting           Average Method
• Model Specification   • Residual
• Moving Average        • Seasonal Component
• Nonlinear Trend       • Seasonal Index
• Qualitative           • Seasonally
• Quantitative          • Smoothing Constant
Forecasting           • Splitting Samples
• Random Component
Key Terms
(continued)

• Time-Series Data
• Time-Series Plot
• Trend

```
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
 views: 25 posted: 7/27/2011 language: English pages: 41
Description: Sales Projection Document document sample