MAAUTOLAGS Procedure

@MAAUTOLAGS quickly scans a set of moving average models looking for the one which minimizes one of four information criteria. This can be used in ARMA model identification, and also for diagnostics (best lag should be zero if the series is serially uncorrelated). @ARAutoLags does a similar "quick" calculation for choosing a pure MA process. @BJAUTOFIT and @GMAUTOFIT do more time-consuming calculations (they do exact maximum likelihood fits) for choosing full ARMA models. (@GMAUTOFIT is for seasonal ARMA models).

@MAAutoLags( options )series start end

Parameters

series	series to analyze
start, end	range of series to use (by default, the maximum possible)

Options

MAXLAGS=maximum number of lags to consider [25]

CRIT=[AIC]/BIC/CAIC/HQ

Criterion to use:

AIC is (uncorrected) Akaike information criterion.

BIC is the Bayesian or Schwarz criterion.

HQ is Hannan-Quinn.

CAIC is the AIC-corrected for degrees of freedom

TABLE/[NOTABLE]

Show full table of results (not just best)

Variables Defined

%%AUTOQ

number of parameters selected (INTEGER)

Example

open data dowj.dat

data(format=free,org=columns) 1 78 dowj

set ddow = dowj-dowj{1}

@maautolags(maxlag=17,crit=bic,table) ddow

boxjenk(ma=1,maxl,demean) ddow

boxjenk(ma=2,maxl,demean) ddow

boxjenk(ma=3,maxl,demean) ddow

Sample Output

In the output, small is good. The chosen (starred) model is the MA(1).

Lags IC

0 -1.659

1 -1.798*

2 -1.755

3 -1.699

4 -1.669

5 -1.613

6 -1.569

7 -1.515

8 -1.482

9 -1.427

10 -1.371

11 -1.336

12 -1.297

13 -1.291

14 -1.244

15 -1.231

16 -1.186

17 -1.172