Cyclical fractional cointegration

Michelle Voges; Philipp Sibbertsen

doi:10.1016/j.ecosta.2020.05.004

Cyclical fractional cointegration

Michelle Voges^*, Philipp Sibbertsen

^*Corresponding author for this work

Institute of Statistics

Research output: Contribution to journal › Article › Research › peer review

Abstract

The concept of cyclical long memory is extended to a multivariate setting and definitions of cyclical fractional cointegration are provided. Furthermore, cyclical long-memory models that exhibit these characteristics are proposed and a cyclical multiple local Whittle estimator for the cyclical memory parameters and the cyclical cointegrating vector is derived. A series of Monte Carlo studies shows that the proposed method works well in finite samples. Finally, an application to financial high-frequency data underlines the usefulness of the method in practical applications where cyclical fractional cointegration between realized volatility and trading volume is found for a daily cycle.

Original language	English
Pages (from-to)	114-129
Number of pages	16
Journal	Econometrics and Statistics
Volume	19
Early online date	24 Jun 2020
DOIs	https://doi.org/10.1016/j.ecosta.2020.05.004
Publication status	Published - Jul 2021

Keywords

C52, C58)
Fractional cointegration (C32
Multivariate time series
Seasonal/Cyclical long memory

ASJC Scopus subject areas

Statistics and Probability
Economics and Econometrics
Statistics, Probability and Uncertainty

Access to Document

10.1016/j.ecosta.2020.05.004

Cite this

@article{1abdd35c53444389864f153a4d37fdb2,

title = "Cyclical fractional cointegration",

abstract = "The concept of cyclical long memory is extended to a multivariate setting and definitions of cyclical fractional cointegration are provided. Furthermore, cyclical long-memory models that exhibit these characteristics are proposed and a cyclical multiple local Whittle estimator for the cyclical memory parameters and the cyclical cointegrating vector is derived. A series of Monte Carlo studies shows that the proposed method works well in finite samples. Finally, an application to financial high-frequency data underlines the usefulness of the method in practical applications where cyclical fractional cointegration between realized volatility and trading volume is found for a daily cycle.",

keywords = "C52, C58), Fractional cointegration (C32, Multivariate time series, Seasonal/Cyclical long memory",

author = "Michelle Voges and Philipp Sibbertsen",

note = "Funding Information: We are very grateful to Christian Leschinski for helpful comments. Special thanks are also due to two anonymous referees whose suggestions improved the paper a lot. Financial support of DFG under grant SI 745/9-2 is gratefully acknowledged. ",

year = "2021",

month = jul,

doi = "10.1016/j.ecosta.2020.05.004",

language = "English",

volume = "19",

pages = "114--129",

journal = "Econometrics and Statistics",

publisher = "Elsevier BV",

}

TY - JOUR

T1 - Cyclical fractional cointegration

AU - Voges, Michelle

AU - Sibbertsen, Philipp

N1 - Funding Information: We are very grateful to Christian Leschinski for helpful comments. Special thanks are also due to two anonymous referees whose suggestions improved the paper a lot. Financial support of DFG under grant SI 745/9-2 is gratefully acknowledged.

PY - 2021/7

Y1 - 2021/7

N2 - The concept of cyclical long memory is extended to a multivariate setting and definitions of cyclical fractional cointegration are provided. Furthermore, cyclical long-memory models that exhibit these characteristics are proposed and a cyclical multiple local Whittle estimator for the cyclical memory parameters and the cyclical cointegrating vector is derived. A series of Monte Carlo studies shows that the proposed method works well in finite samples. Finally, an application to financial high-frequency data underlines the usefulness of the method in practical applications where cyclical fractional cointegration between realized volatility and trading volume is found for a daily cycle.

AB - The concept of cyclical long memory is extended to a multivariate setting and definitions of cyclical fractional cointegration are provided. Furthermore, cyclical long-memory models that exhibit these characteristics are proposed and a cyclical multiple local Whittle estimator for the cyclical memory parameters and the cyclical cointegrating vector is derived. A series of Monte Carlo studies shows that the proposed method works well in finite samples. Finally, an application to financial high-frequency data underlines the usefulness of the method in practical applications where cyclical fractional cointegration between realized volatility and trading volume is found for a daily cycle.

KW - C52, C58)

KW - Fractional cointegration (C32

KW - Multivariate time series

KW - Seasonal/Cyclical long memory

UR - http://www.scopus.com/inward/record.url?scp=85088223034&partnerID=8YFLogxK

U2 - 10.1016/j.ecosta.2020.05.004

DO - 10.1016/j.ecosta.2020.05.004

M3 - Article

AN - SCOPUS:85088223034

VL - 19

SP - 114

EP - 129

JO - Econometrics and Statistics

JF - Econometrics and Statistics

ER -

Cyclical fractional cointegration

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

The effect of structural changes to inference in long-memory time series

Cite this

Cyclical fractional cointegration

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Projects

The effect of structural changes to inference in long-memory time series

Cite this