Library: | times |
See also: | armacls |
Quantlet: | armalik | |
Description: | estimates an ARMA(1,1) process with mean zero by maximum likelihood using the innovation algorithm |
Usage: | y = armalik(x) | |
Input: | ||
x | n-vector, the process | |
Output: | ||
y | list containing 1. the estimated parameters, 2. the corresponding asymptotic standard deviations, 3. the asymptotic covariance, and 4. the estimate of the white noise variance |
library("times") ; Loads the quantlets from Times Library randomize(0) ; Sets random seed x = genarma(0.7,0.3,normal(100)) ; Generates ARMA(p,q) with White Noise timeplot(x,100) ; Plots the ARMA process in a single display {a, stderr, covp, s2} = armalik(x) ; Estimation procedure for ARMA(1,1) process z=(a)|(stderr)|(covp)|(s2) ; Estimation output z
Contents of z ; Vector of parameter Estimates [1,] 0.87487 ; AR(1) [2,] -0.069252 ; MA(1) [3,] 0.056481 ; Corresponding Standard Error [4,] 0.11633 ; Corresponding Standard Error [5,] -0.0033796 ; Corresponding Asymptotic Convariance [6,] 1.023 ; Estimate of White Noise Variance "For reading convenience the graphical output has been omitted."
library("times") ; Loads the quantlets from Times Library randomize(101) ; Sets random seed dax=read("dax") ; monthly DAX 1979:1-2000:10 daxreturn=tdiff(log(dax)) ; generates the monthly return timeplot(daxreturn,261) ; Plots the monthly return series for DAX {a, stderr, covp, s2} = armalik(daxreturn) ; Estimation procedure for return process z=(a)|(stderr)|(covp)|(s2) ; Estimation output z
Contents of z ; Vector of parameter Estimates [1,] -0.066774 ; AR(1) [2,] 0.088562 ; MA(1) [3,] 2.8178 ; Corresponding Standard Error [4,] 2.8131 ; Corresponding Standard Error [5,] -7.9248 ; Corresponding Asymptotic Convariance [6,] 0.0032548 ; Estimate of White Noise Variance "For reading convenience the graphical output has been omitted."