Paper: arXiv 2212.00391

Authors: Hyungbin Park, Heejun Yeo

Abstract

This paper investigates dynamic and static fund separations and their stability for long-term optimal investments under three model classes. An investor maximizes the expected utility with constant relative risk aversion under an incomplete market consisting of a safe asset, several risky assets, and a single state variable. The state variables in two of the model classes follow a 3/2 process and an inverse Bessel process, respectively. The other market model has the partially observed state variable modeled as an Ornstein-Uhlenbeck state process. We show that the dynamic optimal portfolio of this utility maximization consists of m+3 portfolios: the safe asset, the myopic portfolio, the m time-independent portfolios, and the intertemporal portfolio. Over time, the intertemporal portfolio eventually vanishes, leading the dynamic portfolio to converge to m+2 portfolios, referred to as the static portfolio. We also prove that the convergence is stable under model parameter perturbations. In addition, sensitivities of the intertemporal portfolio with respect to small parameters perturbations also vanish in the long run. The convergence rate for the intertemporal portfolio and its sensitivities are computed explicitly for the presented models.

Complexity vs Empirical Score

  • Math Complexity: 8.5/10
  • Empirical Rigor: 2.0/10
  • Quadrant: Lab Rats — theoretically deep, empirically untested

Why this score: The paper presents advanced continuous-time stochastic control and utility maximization involving specialized processes (3/2, inverse Bessel, partial observation OU). However, it is purely theoretical with no empirical backtesting or real market data validation.

Research Flowchart

  flowchart TD
    A[Goal: Analyze stability & dynamic/static fund separation for long-term optimal CRRA investment] --> B[Define Asset Market & State Variable Models]
    
    B --> C[Input Models: Incomplete Market with Safe Asset, Risky Assets, & State Variable<br><i>3/2 process, Inverse Bessel, or Partially Observed OU process</i>]
    
    C --> D[Utility Maximization Computation<br><i>Solve CRRA expected utility over time</i>]
    
    D --> E[Derive Dynamic Portfolio: m+3 Funds<br><i>Safe Asset + Myopic + m Time-Independent + Intertemporal</i>]
    
    E --> F[Long-Term Limit & Sensitivity Analysis<br><i>Evaluate convergence rate & parameter perturbations</i>]
    
    F --> G[Key Outcomes:<br>1. Intertemporal fund vanishes, converging to Static Portfolio m+2 funds<br>2. Convergence & sensitivities are stable under parameter perturbations]