VMAorder: VMA Order Specification

Description Usage Arguments Details Value Author(s) References Examples

Description

Performs multivariate Ljung-Box tests to specify the order of a VMA process

Usage

1
VMAorder(x, lag = 20)

Arguments

x

Data matrix of the observed k-dimensional time series. Each column represents a time series.

lag

The maximum VMA order entertained. Default is 20.

Details

For a given lag, the command computes the Ljung-Box statistic for testing rho_j = ... = rho_lag = 0, where j = 1, 2, ..., lag. For a VMA(q) process, the Ljung-Box statistics should be significant for the first q lags, and insignificant thereafter.

Value

The Q-statistics and p-value plot

Author(s)

Ruey S. Tsay

References

Tsay (2014). Multivariate Time Series Analysis with R and Financial Applications. John Wiley. Hoboken, NJ.

Examples

1
2
zt=matrix(rnorm(600),200,3)
VMAorder(zt)

Example output

Q(j,m) Statistics:  
         j     Q(j,m)   p-value
 [1,]   1.00    197.13     0.18
 [2,]   2.00    191.57     0.13
 [3,]   3.00    185.95     0.10
 [4,]   4.00    181.99     0.05
 [5,]   5.00    171.40     0.06
 [6,]   6.00    157.63     0.09
 [7,]   7.00    145.80     0.11
 [8,]   8.00    133.41     0.14
 [9,]   9.00    124.90     0.13
[10,]  10.00    118.27     0.09
[11,]  11.00    102.37     0.18
[12,]  12.00     98.64     0.09
[13,]  13.00     88.90     0.09
[14,]  14.00     73.46     0.17
[15,]  15.00     69.96     0.07
[16,]  16.00     59.70     0.07
[17,]  17.00     43.91     0.17
[18,]  18.00     37.06     0.09
[19,]  19.00     24.65     0.13
[20,]  20.00      8.53     0.48

MTS documentation built on May 29, 2017, 5:15 p.m.