Optimal Diversification within Mixed-Asset Portfolios using a Conditional Heteroskedasticity Approach: Evidence from the U.S. and the U.K.

Michael Giliberto, Foort Hamelink, Martin Edward Ralph Hoesli, Bryan Duncan MacGregor

Research output: Contribution to journalArticle

Abstract

In this article, portfolio allocation strategies based on a threshold autoregressive conditional heteroskedasticity model (QTARCH) are constructed for the United States and the United Kingdom and compared to a conventional asset allocation. Our procedure is based on partitioning the historical data into ‘states of the world,’ which are used to produce expectations of return and risk. Several approaches are developed to partition an initial in-sample period (1978-1983), using quarterly asset returns and economic data. These partitions are then used to test out-of-sample strategies for the next quarter. Although the conditional results are sensitive to the method of partitioning, we show that the approach can improve portfolio performance in both countries and that most of the performance improvement stems from using conditional variances-covariances.
Original languageEnglish
Pages (from-to)31-45
Number of pages16
JournalJournal of Real Estate Portfolio Management
Volume5
Issue number1
Publication statusPublished - 1998

Fingerprint

Economics
Conditional heteroskedasticity
Partitioning
Assets
Diversification
Economic data
Risk and return
Performance improvement
Portfolio allocation
Portfolio performance
Conditional variance
Asset returns
Autoregressive conditional heteroskedasticity
Asset allocation

Cite this

@article{598660045b884a2cbd0ccec754b4dde9,
title = "Optimal Diversification within Mixed-Asset Portfolios using a Conditional Heteroskedasticity Approach: Evidence from the U.S. and the U.K.",
abstract = "In this article, portfolio allocation strategies based on a threshold autoregressive conditional heteroskedasticity model (QTARCH) are constructed for the United States and the United Kingdom and compared to a conventional asset allocation. Our procedure is based on partitioning the historical data into ‘states of the world,’ which are used to produce expectations of return and risk. Several approaches are developed to partition an initial in-sample period (1978-1983), using quarterly asset returns and economic data. These partitions are then used to test out-of-sample strategies for the next quarter. Although the conditional results are sensitive to the method of partitioning, we show that the approach can improve portfolio performance in both countries and that most of the performance improvement stems from using conditional variances-covariances.",
author = "Michael Giliberto and Foort Hamelink and Hoesli, {Martin Edward Ralph} and MacGregor, {Bryan Duncan}",
year = "1998",
language = "English",
volume = "5",
pages = "31--45",
journal = "Journal of Real Estate Portfolio Management",
issn = "1083-5547",
publisher = "INST INVESTOR INC",
number = "1",

}

TY - JOUR

T1 - Optimal Diversification within Mixed-Asset Portfolios using a Conditional Heteroskedasticity Approach

T2 - Evidence from the U.S. and the U.K.

AU - Giliberto, Michael

AU - Hamelink, Foort

AU - Hoesli, Martin Edward Ralph

AU - MacGregor, Bryan Duncan

PY - 1998

Y1 - 1998

N2 - In this article, portfolio allocation strategies based on a threshold autoregressive conditional heteroskedasticity model (QTARCH) are constructed for the United States and the United Kingdom and compared to a conventional asset allocation. Our procedure is based on partitioning the historical data into ‘states of the world,’ which are used to produce expectations of return and risk. Several approaches are developed to partition an initial in-sample period (1978-1983), using quarterly asset returns and economic data. These partitions are then used to test out-of-sample strategies for the next quarter. Although the conditional results are sensitive to the method of partitioning, we show that the approach can improve portfolio performance in both countries and that most of the performance improvement stems from using conditional variances-covariances.

AB - In this article, portfolio allocation strategies based on a threshold autoregressive conditional heteroskedasticity model (QTARCH) are constructed for the United States and the United Kingdom and compared to a conventional asset allocation. Our procedure is based on partitioning the historical data into ‘states of the world,’ which are used to produce expectations of return and risk. Several approaches are developed to partition an initial in-sample period (1978-1983), using quarterly asset returns and economic data. These partitions are then used to test out-of-sample strategies for the next quarter. Although the conditional results are sensitive to the method of partitioning, we show that the approach can improve portfolio performance in both countries and that most of the performance improvement stems from using conditional variances-covariances.

M3 - Article

VL - 5

SP - 31

EP - 45

JO - Journal of Real Estate Portfolio Management

JF - Journal of Real Estate Portfolio Management

SN - 1083-5547

IS - 1

ER -