Abstract:This paper treats of discrete-time Markov decision processes with recursive utility defined using a non-linear aggregator of Epstein-Zin type with the Kreps-Porteus certainty equivalent operator. According to the classification introduced by Marinacci and Montrucchio, the aggregators that we consider are of Thompson type. We focus on the existence and uniqueness of a solution to the Bellman equation. If the action spaces of an agent are singletons, the problem concerns the existence and uniqueness of a recursive utility in the sense of Koopmans. The models of this type were recently studied by Ren and Stachurski who applied Du's fixed point theorem for increasing and convex or concave operators acting on an ordered Banach space. We show that, in two cases considered by Ren and Stachurski, the existence and uniqueness of solution to the Bellman equation follows from the Banach contraction mapping principle. We allow the per-period utilities in the models to be unbounded. Therefore, we work with the weighted supremum norm. Since we apply the Banach fixed point theorem for contraction mappings acting on a standard complete metric space, we need not assume any boundary conditions, which are present when the Thompson metric or Du's theorem are used. Moreover, our results give better bounds for the geometric convergence of the value iteration algorithm leading to the solution of the Bellman equation than those obtained by Du's fixed point theorem. Some illustrative examples are given. This paper also contains also new results on the existence and uniqueness of solutions to the Bellman equation in two cases of parameters in the Epsten-Zin aggregator, where the application of Du's theorem is not possible. Our results are based on a relatively simple reformulation of the problem exploiting the structure of Epstein-Zin preferences.

What problem does this paper attempt to address?

This paper attempts to solve the existence and uniqueness of the solution to the Bellman equation in stochastic dynamic programming under the framework of Epstein - Zin preferences. Specifically: 1. **Research Background and Motivation**: - Epstein - Zin preference is a non - linear recursive utility function, which can distinguish risk attitude and the elasticity of intertemporal substitution (EIS), which is of great significance in economic, financial and other fields. - Traditional linear utility functions cannot handle these complex risk and time - preference relationships well. 2. **Problem Description**: - The author focuses on the recursive utility problem in discrete - time Markov decision processes (MDP), where the utility function is defined through a non - linear aggregator of Epstein - Zin type. - The main problem is to prove the existence and uniqueness of the solution to the Bellman equation, especially under certain parameter conditions (such as \(0 < \gamma < \rho < 1\), \(0 < \rho < \gamma < 1\), etc.). 3. **Methods and Contributions**: - The author uses the Banach contraction mapping principle to prove the existence and uniqueness of the solution to the Bellman equation without assuming boundary conditions. - This method is applicable to certain specific types of Epstein - Zin aggregators, especially Thompson - type aggregators. - Compared with the previous use of Du's fixed - point theorem, this method provides a faster geometric convergence rate and makes weaker regularity assumptions on the utility function per period. 4. **Main Results**: - For different parameter ranges (such as \(0 < \rho < \gamma < 1\), \(0 < \gamma < \rho < 1\), \(\gamma > \rho > 1\), \(\rho > \gamma > 1\)), the author respectively proves the existence and uniqueness of the solution to the Bellman equation. - Specific convergence rate estimates are provided, and the fast convergence of the numerical iteration algorithm is demonstrated. In summary, this paper solves the existence and uniqueness of the solution to the Bellman equation under Epstein - Zin preferences by introducing the Banach contraction mapping principle, and in some cases provides better convergence performance than existing methods. This provides a new theoretical tool for understanding and applying recursive utility models.

Stochastic dynamic programming under recursive Epstein-Zin preferences

Stochastic Dynamic Programming with Non-linear Discounting

An approximation approach to dynamic programming with unbounded returns

Dynamic Programming Principle and Associated Hamilton-Jacobi-Bellman Equation for Stochastic Recursive Control Problem with Non-Lipschitz Aggregator

Recursive Utility and Turnpike Theory for GMM Thompson Aggregators

Computing Dsge Models with Recursive Preferences and Stochastic Volatility

Computing DSGE Models with Recursive Preferences

A Nonparametric GMM Series Approach to Solving Multi-equation Asset Pricing Models with Recursive Preferences

Dynamic Coalition Portfolio Selection with Recursive Utility

Utility maximization in constrained and unbounded financial markets: Applications to indifference valuation, regime switching, consumption and Epstein-Zin recursive utility

Some Limit Properties of Markov Chains Induced by Stochastic Recursive Algorithms

Existence and uniqueness of solutions to the Bellman equation in stochastic dynamic programming

Consumption and Portfolio Optimization with Generalized Stochastic Differential Utility in Incomplete Markets.

Proper solutions for Epstein-Zin Stochastic Differential Utility

Do not Blame Bellman: It Is Koopmans' Fault

Optimal consumption and investment under relative performance criteria with Epstein-Zin utility

Utility Maximization with Habit Formation: Dynamic Programming and Stochastic PDEs

Consumption-investment optimization with Epstein-Zin utility in unbounded non-Markovian markets

Unbounded Markov Dynamic Programming with Weighted Supremum Norm Perov Contractions

Dynamic Volatility Strategy with Recursive Utility

Dynamic programming principle for delayed stochastic recursive optimal control problem and HJB equation with non-Lipschitz generator