University of Notre Dame
Browse

File(s) under permanent embargo

Essays on Asset Returns and Portfolio Allocation

thesis
posted on 2016-04-15, 00:00 authored by Linlin Liu

In this dissertation, I focus on two broad questions. First, what are the interconnections and the underlying driving forces of stock return and bond yields? Understanding the cross-assets dynamics and their economic determinants is important to both policymakers’ and investors’ decision-making, since the interest rates and stock return reflect the macroeconomic expectations for the future state of the economy. This leads to my second question: what are the implications of the empirical behavior of asset returns for the optimal decisions made by investors? Along these lines, in the first chapter of the dissertation, I provide an introduction to the development of the asset pricing models and the portfolio allocation models. In the second chapter, I build on the arbitrage-free asset pricing model to simultaneously price U.S. aggregate stock and the cross-section of Treasury bonds, and account for the time-varying equity risk premium and bond risk premium in a unified framework. The model imposes restrictions on the cross-assets dynamics to preclude arbitrage opportunities at any point in time. With the risk factors constructed from the asset return data and macroeconomic data, the model is able to generate considerably small pricing errors for aggregate stock return and Treasury yields. Also, the model-implied equity and bond risk premiums show marked countercyclical patterns. Since a rational utility maximizing investor requires higher compensation for the exposure to risk during recessions, the estimated countercyclical patterns of the risk premiums are consistent with the theoretical models. In the third chapter, I explore whether the horizon based asset allocation strategy, that is popular among investment professionals, is optimal for investors. To do so, I develop a dynamic portfolio choice model that incorporates two sources of risks that investors face: the time-varying equity risk premium and the uncertainty in labor income. The optimal portfolio weight on stock depends on the relative magnitude of the correlation between aggregate stock return and equity risk premium, and the correlation between aggregate stock return and labor income.

History

Date Modified

2017-06-02

Defense Date

2016-04-05

Research Director(s)

Thomas Cosimano

Degree

  • Doctor of Philosophy

Degree Level

  • Doctoral Dissertation

Program Name

  • Economics

Usage metrics

    Dissertations

    Categories

    No categories selected

    Keywords

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC