Equilibrium reinsurance and investment strategies for insurers with random risk aversion under Heston's SV model

Basic Information

• Paper ID: 2412.19050
• Title: Equilibrium reinsurance and investment strategies for insurers with random risk aversion under Heston's SV model
• Authors: Jian-hao Kang, Zhun Gou, Nan-jing Huang
• Classification: math.OC (Mathematical Optimization and Control)
• Publication Date: January 2, 2025 (arXiv v2 version)
• Paper Link: https://arxiv.org/abs/2412.19050

Abstract

This study employs the expected certainty equivalents approach to explore reinsurance and investment problems for insurers under random risk aversion. The surplus process of the insurance company is modeled using Brownian motion with drift, while the financial market comprises a risk-free asset and a risky asset whose price is described by the Heston stochastic volatility (SV) model. Within a game-theoretic framework, rigorous verification theorems are established to characterize equilibrium reinsurance and investment strategies along with corresponding value functions. By solving a pseudo Hamilton-Jacobi-Bellman (HJB) system, semi-analytical expressions for equilibrium reinsurance and investment strategies are obtained under exponential utility functions. Additionally, numerical experiments are conducted to demonstrate the characteristics of equilibrium strategies.

Research Background and Motivation

Problem Background

1. Limitations of classical investment problems: Since Merton's pioneering work, dynamic investment problems based on expected utility maximization have been a popular research direction in financial economics. However, traditional research assumes a constant risk aversion coefficient, which is insufficiently accurate in practice.
2. Time inconsistency problem: In many cases (such as non-exponential discount factors), reinsurance-investment problems under the expected utility maximization criterion exhibit time inconsistency, where currently optimal strategies may no longer be optimal in the future.
3. Importance of stochastic volatility: Extensive empirical evidence shows that the constant volatility assumption contradicts the volatility smile phenomenon. Stochastic volatility models such as the Heston model better explain these phenomena.

Research Motivation

1. Randomness of risk aversion: In reality, investors/insurers find it difficult to determine the precise value of their risk aversion coefficient and should treat it as a random variable.
2. Model realism: Incorporating random risk aversion into reinsurance-investment problems under the Heston stochastic volatility environment enhances the practical applicability of the model.
3. Theoretical refinement: Extending existing time-inconsistent control theory to provide more practical decision-making tools for insurance companies.

Core Contributions

1. First introduction of random risk aversion into reinsurance-investment problems under the Heston model: This is pioneering work in the field, extending the research of Desmettre and Steffensen (2013).
2. Establishment of rigorous verification theorems: Providing a complete theoretical framework for general random risk aversion reinsurance-investment optimization problems.
3. Derivation of semi-analytical solutions: Under exponential utility functions, semi-closed-form expressions for equilibrium reinsurance and investment strategies are derived with rigorous proof of relevant conditions.
4. Numerical analysis and practical insights: Through numerical experiments, the impact of model parameters on equilibrium strategies is analyzed, yielding valuable practical conclusions.

Methodology Details

Task Definition

Study the optimal reinsurance and investment decision problem for insurance companies under random risk aversion:

• Input: Initial wealth, market parameters, risk aversion distribution
• Output: Equilibrium reinsurance strategy q(t) and investment strategy π(t)
• Constraints: Dynamic constraints of wealth process, acceptability conditions

Model Architecture

1. Surplus Process Model

The surplus process of the insurance company is approximated by Brownian motion with drift:

dR(t) = a[η + η₂q(t)]dt + bq(t)dW₀(t)

where q(t) is the proportional reinsurance strategy.

2. Financial Market Model

The Heston stochastic volatility model is adopted:

dS₁(t) = S₁(t)[(r + ξV(t))dt + √V(t)dW₁(t)]
dV(t) = κ(θ - V(t))dt + σ√V(t)dW₂(t)

3. Wealth Process Dynamics

dX^u(t) = [rX^u(t) + aη + aη₂q(t) + ξV(t)π(t)]dt + bq(t)dW₀(t) + π(t)√V(t)dW₁(t)

4. Objective Function

Using the expected certainty equivalents approach, the objective function is:

J^u(t,x,v) = ∫(φ^γ)⁻¹(E_t[φ^γ(X^u(T))])dΓ(γ)

Technical Innovations

1. Random Risk Aversion Treatment

• Risk aversion coefficient γ is treated as a random variable
• The expected certainty equivalents method is employed to handle weighted multiple utility functions

2. Time Inconsistency Resolution

• A game-theoretic framework is adopted to find equilibrium strategies
• A pseudo-HJB system is established to characterize equilibrium solutions

3. Semi-analytical Solution Method

• For exponential utility functions, an ansatz method is employed: Y^γ(t,x,v) = -1/γ * exp(g₁^γ(t)x + g₂^γ(t)v + g₃^γ(t))
• The infinite-dimensional ODE system is reduced to a finite-dimensional coupled equation system

Experimental Setup

Parameter Configuration

Based on standard parameters from references 4 and 25:

• Insurance market parameters: η₁=0.3, η₂=0.5, λ₁=1, μ₁=0.1, μ₂=0.2
• Financial market parameters: r=0.05, ξ=7/15, κ=5, θ=(0.15)², σ=0.25, ρ=-0.5
• Time horizon: T=10 years and T=100 years

Risk Aversion Distribution

Two-point distributions are considered:

• Case (I): γ₁=0.5, γ₂=4, p₁=p₂=0.5 (Eγ=2.25)
• Case (II): γ₁=0.5, γ₂=4, p₁=0.8, p₂=0.2 (Eγ=1.2)

Numerical Methods

The Predictor-Corrector method is employed to solve the coupled forward ODE system, converting the backward ODE into forward form for numerical computation.

Experimental Results

Main Results

1. Characteristics of Equilibrium Investment Strategy

• Interest rate impact: As the risk-free rate r increases, the equilibrium investment strategy decreases
• Risk aversion impact: The higher the expected risk aversion Eγ, the less is invested in risky assets
• Time effect: The equilibrium investment strategy decreases over time, supporting the decreasing life-cycle investment theory under exponential utility

2. Characteristics of Equilibrium Reinsurance Strategy

• State independence: The equilibrium strategy is deterministic and state-independent
• Parameter sensitivity:
  • Decreasing in r: ∂q̂(t)/∂r < 0
  • Increasing in η₂: ∂q̂(t)/∂η₂ > 0
  • Increasing in μ₁: ∂q̂(t)/∂μ₁ > 0
  • Decreasing in μ₂: ∂q̂(t)/∂μ₂ < 0

3. Impact of Volatility Parameters

• Role of correlation coefficient ρ: When ρ>0, risky asset prices and volatility move in the same direction; when ρ<0, they move in opposite directions
• Mean reversion rate κ: Under different Eγ and ρ signs, its impact on investment strategy exhibits opposite trends

Numerical Findings

1. Complexity of Investment Strategy

Numerical experiments reveal the necessity of considering random risk aversion in incomplete markets: insurance companies with different expected risk aversion levels adopt opposite investment strategies to manage risk.

2. Parameter Condition Verification

Numerical experiments verify that the theoretical condition -8γᵢξπ̄(t) + 32γᵢ²π̄²(t) ≤ κ²/(2σ²) holds within reasonable parameter ranges.

Related Work

Main Research Directions

1. Classical investment theory: Merton's (1969, 1971) pioneering work
2. Reinsurance-investment problems: Richard (1975) first incorporated life insurance into investment problems
3. Time inconsistency theory: Game-theoretic approaches by Björk et al.
4. Stochastic volatility models: Applications of the Heston model in option pricing and portfolio management

Unique Contributions of This Paper

Compared to existing research, this paper is the first to address reinsurance-investment problems with random risk aversion under the Heston stochastic volatility environment, extending the theoretical framework and providing practical solutions.

Conclusions and Discussion

Main Conclusions

1. Theoretical completeness: A complete verification theorem and solution framework is established
2. Practicality: Semi-analytical solutions and numerical algorithms are provided
3. Economic insights: The significant impact of random risk aversion on strategy selection is revealed
4. Strategy characteristics: Equilibrium strategies are deterministic and state-independent

Limitations

1. Model assumptions: The insurance market and financial market are assumed to be uncorrelated
2. Utility function restrictions: Main results focus on exponential utility functions
3. Parameter conditions: Specific technical conditions must be satisfied to ensure solution existence

Future Directions

1. Consider other stochastic volatility models (e.g., CEV model)
2. Incorporate default risk management
3. Study game problems between insurance companies and reinsurance companies
4. Extend to other types of reinsurance contracts

In-Depth Evaluation

Strengths

1. Theoretical innovation: First introduction of random risk aversion into reinsurance-investment problems under the Heston model
2. Rigorous methodology: Complete mathematical proofs and verification theorems are provided
3. Practical value: Semi-analytical solutions facilitate practical applications
4. Comprehensive numerical analysis: Thorough analysis of parameter impacts

Weaknesses

1. Computational complexity: Solution remains difficult for general utility functions
2. Parameter sensitivity: Solution existence depends on strong technical conditions
3. Insufficient empirical validation: Lacks verification with real data

Impact

1. Theoretical contribution: Provides new application scenarios for time-inconsistent control theory
2. Practical value: Offers theoretical support for insurance company decision-making
3. Foundation for future research: Establishes a foundation for further research in related fields

Applicable Scenarios

This method is particularly suitable for:

1. Strategic decision-making of large insurance companies
2. Investment environments with uncertain risk aversion levels
3. Financial markets requiring consideration of stochastic volatility

References

The paper cites 38 important references, primarily including:

• Merton (1969, 1971): Foundation of classical investment theory
• Björk et al. series: Time-inconsistent control theory
• Desmettre & Steffensen (2023): Random risk aversion investment problems
• Heston (1993): Stochastic volatility model

Overall Assessment: This is a high-quality theoretical research paper with significant theoretical value and practical significance in mathematical finance. The paper has high technical content, rigorous proofs, and provides valuable theoretical tools for insurance company decision-making under uncertain environments.