A Wavelet Whittle Estimator of Generalized Long-memory Stochastic Volatility

Michael Hauser

A Wavelet Whittle Estimator of Generalized Long-memory Stochastic Volatility

Michael Hauser

Statistics and Mathematics

Publikation: Konferenzbeitrag › Konferenzposter

Abstract

We propose a wavelet based Whittle estimator, WWE, for $k$-GARMA and generalized stochastic long-memory volatility models, GLMSV. It is shown that the decorrelation properties of wavelets for FI hold also for $k$-GARMA and GLMSV. Using this we show the consistency of the WWE. Small sample properties are compared to Whitcher's(04) and Whittle's estimator. It clearly dominates Whichter's, and is essentially indistinguishable to Whittle's. A GLMSV application for Microsoft realized volatilities is given.

Originalsprache	Englisch
Publikationsstatus	Veröffentlicht - 2010

Verfügbarkeit prüfen

Katalog der Universitätsbibliothek

Zitat

@conference{6be27a47a5e04b7da7a86cd6489a3f32,

title = "A Wavelet Whittle Estimator of Generalized Long-memory Stochastic Volatility",

abstract = "We propose a wavelet based Whittle estimator, WWE, for $k$-GARMA and generalized stochastic long-memory volatility models, GLMSV. It is shown that the decorrelation properties of wavelets for FI hold also for $k$-GARMA and GLMSV. Using this we show the consistency of the WWE. Small sample properties are compared to Whitcher's(04) and Whittle's estimator. It clearly dominates Whichter's, and is essentially indistinguishable to Whittle's. A GLMSV application for Microsoft realized volatilities is given.",

author = "Michael Hauser",

year = "2010",

language = "English",

}

TY - CONF

T1 - A Wavelet Whittle Estimator of Generalized Long-memory Stochastic Volatility

AU - Hauser, Michael

PY - 2010

Y1 - 2010

N2 - We propose a wavelet based Whittle estimator, WWE, for $k$-GARMA and generalized stochastic long-memory volatility models, GLMSV. It is shown that the decorrelation properties of wavelets for FI hold also for $k$-GARMA and GLMSV. Using this we show the consistency of the WWE. Small sample properties are compared to Whitcher's(04) and Whittle's estimator. It clearly dominates Whichter's, and is essentially indistinguishable to Whittle's. A GLMSV application for Microsoft realized volatilities is given.

AB - We propose a wavelet based Whittle estimator, WWE, for $k$-GARMA and generalized stochastic long-memory volatility models, GLMSV. It is shown that the decorrelation properties of wavelets for FI hold also for $k$-GARMA and GLMSV. Using this we show the consistency of the WWE. Small sample properties are compared to Whitcher's(04) and Whittle's estimator. It clearly dominates Whichter's, and is essentially indistinguishable to Whittle's. A GLMSV application for Microsoft realized volatilities is given.

M3 - Conference poster

ER -