You are on page 1of 2

A taxonomy of exponential smoothing methods

Exponential smoothing methods are not restricted to those we have presented so far. By
considering variations in the combination of the trend and seasonal components, fifteen
exponential smoothing methods are possible, listed in Table 7.7. Each method is labelled
by a pair of letters (T,S) defining the type of Trend and Seasonal components. For
example, (A,M) is the method with an additive trend and multiplicative seasonality;
(M,N) is the method with multiplicative trend and no seasonality; and so on.

Trend
Component
N (None)
A (Additive)
Ad (Additive damped)
M (Multiplicative)
Md (Multiplicative damped)

Seasonal Component
N
A
(None) (Additive)
(N,N) (N,A)
(A,N)
(A,A)
(Ad,N) (Ad,A)
(M,N) (M,A)
(Md,N) (Md,A)

M
(Multiplicative)
(N,M)
(A,M)
(Ad,M)
(M,M)
(Md,M)

Some of these methods we have already seen:

(N,N)
(A,N)
(M,N)
(Ad,N)
(Md,N)
(A,A)
(A,M)
(Ad,M)

=
=
=
=
=
=
=
=

simple exponential smoothing


Holts linear method
Exponential trend method
additive damped trend method
multiplicative damped trend method
additive Holt-Winters method
multiplicative Holt-Winters method
Holt-Winters damped method

This type of classification was proposed by Pegels (1969). It was later extended by
Gardner (1985) to include methods with additive damped trend and by Taylor (2003) to
include methods with multiplicative damped trend.
Table 7.8 gives the recursive formulae for applying all possible fifteen exponential
smoothing methods. Each cell includes the forecast equation for generating h-stepahead forecasts and the smoothing equations for applying the method. In Table 7.9 we
present some strategies for selecting initial values for some of the most commonly
applied exponential smoothing methods. We do not recommend that these strategies be
used directly; rather, they are useful in providing starting values for the optimization
process.

Table
7.8: Formulae for recursive calculations and point forecasts. In each case, tdenotes the series level
at time t, bt denotes the slope at time t, st denotes the seasonal component of the series at time t,
and m denotes the number of seasons in a year; , , and are smoothing
parameters, h=+2++h and h+m=(h1)

mod m+1. (Click the table for a

larger version.)

Method

Initial values

(N,N)

0=y1

(A,N) (Ad,N)

0=y1, b0=y2y1

(M,N) (Md,N)

0=y1, b0=y2/y1

(A,A) (Ad,A)

0=1m(y1++ym)
b0=1m[ym+1y1m++ym+mymm]
s0=ym0, s1=ym10, , sm+1=y10

(A,M) (Ad,M)

0=1m(y1++ym)
b0=1m[ym+1y1m++ym+mymm]
s0=ym/0, s1=ym1/0, , sm+1=y1/0

You might also like