Exponential Smoothing

Exponential smoothing is a simple method of adaptive forecasting. It is an effective way of forecasting when you have only a few observations on which to base your forecast. Unlike forecasts from regression models which use fixed coefficients, forecasts from exponential smoothing methods adjust based upon past forecast errors. For additional discussion, see Bowerman and OConnell (1979). To obtain forecasts based on exponential smoothing methods, choose Proc/Exponential Smoothing. The Exponential Smoothing dialog box appears: You need to provide the following information: Smoothing Method. You have the option to choose one of the five methods listed. Smoothing Parameters. You can either specify the values of the smoothing parameters or let EViews estimate them. To estimate the parameter, type the letter e (for estimate) in the edit field. EViews estimates the parameters by minimizing the sum of squared errors. Don't be surprised if the estimated damping parameters are close to on eit is a sign that the series is close to a random walk, where the most recent value is the best estimate of future values. To specify a number, type the number in the field corresponding to the parameter. All parameters are constrained to be between 0 an d 1; if you specify a number outside the unit interval, EViews will estimate the parameter. Smoothed Series Name. You should provide a name for the smoothed series. By default, EViews will generate a name by appending SM to the original series name, but you can enter any valid EViews name. Estimation Sample. You must specify the sample period upon which to base your forecasts (whether or not you choose to estimate the parameters). The default is the current workfile sample. EViews will calculate forecasts starting from the first observation after the end of the estimation sample. Cycle for Seasonal. You can change the number of seasons per year from the default of 12 for monthly or 4 for quarterly series. This option allows you to forecast from unusual data such as an undated workfile. Enter a number for the cycle in this field.

Single Smoothing (one parameter)

yt = ayt (1 a)yt 1
where is the damping (or smoothing) factor. The smaller is the , the smoother is the series. By repeated substitution, we can rewrite the recursion as (11.39) This shows why this method is called exponential smoothing the forecast of is a weighted average of the past values of , where the weights decline exponentially with time. The forecasts from single smoothing are constant for all future observations. This constant is given by: (11.40) where is the end of the estimation sample. To start the recursion, we need an initial value for and a value for . EViews uses the mean of the initial observations of to start the recursion (where is the number of observations in the sample). Bowerman and OConnell (1979) suggest that values of around 0.01 to 0.30 work quite well. You can also let EViews estimate to minimize the sum of squares of one-step forecast errors.

Double Smoothing (one parameter)

This method applies the single smoothing method twice (using the same parameter) and is appropriate for series with a linear trend. Double smoothing of a series is defined by the recursions: (11.41) where is the single smoothed series and is the double smo othed series. Note that double smoothing is a single parameter smoothing method with damping factor . Forecasts from double smoothing are computed as: (11.42)

The last expression shows that forecasts from double smoothing lie on a linear trend with intercept and slope .

y t ayt (1 a)y t 1= + 0< a 1a y

y t a (1 a)syt s
s= 0 t 1

yt yt y

T + k y = T for all k > 0 T y ta (T + 1) 2 yt T aa y St = ayt + (1 a)St 1 Dt aSt (1 a)Dt 1 = + SD 0< a 1 y T + k 2 ak 1a + ------------ ST 1 ak 1a + ------------ = DT 2ST DT a 1a + ------------ (ST DT)k = 2ST DT a(ST DT) (1 a)

Exponential Smoothing357

Holt-WintersMultiplicative (three parameters)

This method is appropriate for series with a linear time trend and multiplicative seasonal variation. The smoothed series is given by, (11.43) where (11.44) These three coefficients are defined by the following recursions: (11.45) where are the damping factors and is the seasonal frequency specified in the Cycle for Seasonal field box. Forecasts are computed by: (11.46) where the seasonal factors are used from the last estimates.

Holt-WintersAdditive (three parameter)

This method is appropriate for series with a linear time trend and additi ve seasonal variation. The smoothed series is given by: (11.47)

where and are the permanent component and trend as defined above in Equation (11.44) and are the additive seasonal factors. The three coefficients are defined by the following recursions: (11.48) where are the damping factors and is the seasonal frequency specified in the Cycle for Seasonal field box.

y
t

y (a + bk)ct + k a permanent component (intercept) b trend ct multiplicative seasonal factor a(t) a yt ct (t s) = ------------------- + (1 a)(a(t 1) + b(t 1)) b(t) = b(a(t) a(t 1)) + (1 b)b(t 1) ct (t) g yt a(t) = ---------- + (1 g)ct(t s) 0 < a, b, g < 1 s y t + k = (a(T) + b(T)k)cT + k s s y t y t + k = a + bk + ct + k ab c a(t ) = a(yt ct(t s)) + (1 a)(a(t 1) + b(t 1)) b(t ) = b(a(t) a(t 1)) + 1 bb(t 1) ct (t) = g(yt a(t + 1))gct(t s) 0 < a, b, g < 1 s
358Chapter 11. Series Forecasts are computed by: (11.49) where the seasonal factors are used from the last estimates.
t+ k=

Holt-WintersNo Seasonal (two parameters)

This method is appropriate for series with a linear time trend and no seasonal variation. This method is similar to the double smoothing method in that both generate forecasts with a linear trend and no seasonal component. The double smoothing method is more parsimonious since it uses only one parameter, while this method is a two parameter method. The smoothed series is given by: (11.50) where and are the permanent component and trend as defined above in Equation (11.44). These two coefficients are defined by the following recursions:; (11.51) where are the damping factors. This is an exponential smoothing method with two parameters. Forecasts are computed by: (11.52) These forecasts lie on a linear trend with int ercept and slope . It is worth noting that Holt -WintersNo Seasonal, is not the same as additive or multiplicative with . The condition only restricts the seasonal factors from changing over time so there are still (fixed) nonzero seasonal factors in the f orecasts.

Illustration

As an illustration of forecasting using exponential smoothing we forecast data on monthly housing starts (HS) for the period 1985M01 1988M12 using the DRI Basics data for the period 1959M011984M12. These data are provided in the workfile HS.WF1. Load the workfile, highlight the HS series, double click, select Proc/Exponential Smoothing. We use the Holt-Wintersmultiplicative method to account for seasonality, name the smoothed forecasts as HS_SM, and estimate all parameters over th e period 1959M11984M12. When you click OK, EViews displays the results of the smoothing procedure. The first part displays the estimated (or specified) parameter values, the sum of squared residuals, the

y
T+ k=

a(T) + b(T)k + cT +

ks

s y
t

y
t+ k=

a + bk ab a(t) = ayt + (1 a)(a(t 1) + b(t 1)) b(t) = b(a(t) a(t 1)) + 1 bb(t 1) 0 < a, b, g < 1 y T + k = a(T) + b(T)k a(T) b(T) g= 0g= 0
Exponential Smoothing359 root mean squared error of the forecast. The zero values for Beta and Gamma in this example mean that the trend and seasonal components are estimated as fixed and not changing. The second part of the table displays the mean , and trend at the end of the estimation sample that are used for post -sample smoothed forecasts. For seasonal methods, the seasonal factors used in the forecasts are also displayed. The smoothed series in the workfile contains data from the beginning of the estimation sample to the end of the workfile range; all values after the estimation period are forecasts.

6.4.3.2. Forecasting with Single Exponential Smoothing

New forecast is previous forecast plus an error adjustment

This can be written as:

where

is the forecast error (actual - forecast) for period t.

In other words, the new forecast is the old one plus an adjustment for the error that occurred in the last forecast. Bootstrapping of Forecasts Bootstrapping forecasts What happens if you wish to forecast from some origin, usually the last data point, and no actual observations are available? In this situation we have to modify the formula to become:

where yorigin remains constant. This technique is known as bootstrapping. Example of Bootstrapping Example The last data point in the previous example was 70 and its forecast (smoothed value S) was 71.7. Since we do have the data point and the forecast available, we can calculate the next forecast using the regular formula

= .1(70) + .9(71.7) = 71.5

= .1)

But for the next forecast we have no data point (observation).

So now we compute: St+2 =. 1(70) + .9(71.5 )= 71.35 Comparison between bootstrap and regular forecasting Table comparing two methods The following table displays the comparison between the two methods: Period Bootstrap Data Single Smoothing forecast Forecast 13 14 15 16 17 71.50 71.35 71.21 71.09 70.98 75 75 74 78 86 71.5 71.9 72.2 72.4 73.0

Single Exponential Smoothing with Trend Single Smoothing (short for single exponential smoothing) is not very good when there is a trend. The single coefficient is not enough. Sample data set with trend Let us demonstrate this with the following data set smoothed with an of 0.3: Data Fit 6.4 5.6 7.8 8.8 11.0 11.6 16.7 15.3 21.6 22.4 6.4 6.2 6.7 7.3 8.4 9.4 11.6 12.7 15.4

lot demonstrating inadequacy of single exponential smoothing when there is

The resulting graph looks like:

trend