Package tech.nmfin.portfoliooptimization.lai2010.fit

Class SimpleGARCHFit

java.lang.Object

tech.nmfin.portfoliooptimization.lai2010.fit.SimpleGARCHFit

All Implemented Interfaces:

ResamplerModel

public class SimpleGARCHFit
extends Object
implements ResamplerModel

This class does a quick GARCH(1,1) fitting to the time series, essentially treating the returns as independent. A better way to fit the data is to use a multivariate version of GARCHFit.

Constructor Summary

Constructors

Constructor	Description
SimpleGARCHFit(Matrix mts)

Method Summary

Modifier and Type	Method	Description
Matrix	fittedValues()
Matrix	sigma2()	Gets the conditional variances of residuals over time.
Matrix	standarizedInnovations()	Gets the standarized innovations (normalized by the conditional standard deviation at the time) of the time series.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

SimpleGARCHFit

public SimpleGARCHFit(Matrix mts)

Method Detail

fittedValues

public Matrix fittedValues()

Specified by:

fittedValues in interface ResamplerModel

standarizedInnovations

public Matrix standarizedInnovations()

Description copied from interface: ResamplerModel

Gets the standarized innovations (normalized by the conditional standard deviation at the time) of the time series. Note (haksunli): we should not use instead the non-standarized innovations (have non-unit variances) because when they are scaled by the time dependent standard deviations/variances. We cannot simply shuffle them to get a new time series of innovations.

Specified by:

standarizedInnovations in interface ResamplerModel

Returns:

sigma2

public Matrix sigma2()

Description copied from interface: ResamplerModel

Gets the conditional variances of residuals over time.

Specified by:

sigma2 in interface ResamplerModel

Returns: