Abstract:We find the asymptotics of the value function maximizing the expected utility of discounted dividend payments of an insurance company whose reserves are modeled as a classical Cramér risk process, with exponentially distributed claims, when the initial reserves tend to infinity. We focus on the power and logarithmic utility functions. We perform some numerical analysis as well.
What problem does this paper attempt to address?
The core problem that this paper attempts to solve is to maximize the expected utility of the insurer's discounted dividend payments when the initial reserve tends to infinity in the classical collective risk process. Specifically:
1. **Research Background**:
- The insurer's reserve model is the classical Cramér - Lundberg risk process, and its mathematical expression is:
\[
R_t = x+\mu t-\sum_{i = 1}^{N_t}Y_i
\]
where \(Y_1, Y_2,\ldots\) are independent and identically distributed (i.i.d) positive random variables, representing claim amounts; \(N_t\) is a Poisson process with intensity \(\lambda>0\), representing the time when claims occur; \(x > 0\) represents the initial surplus; \(\mu\) is the premium intensity.
2. **Objective Function**:
- The researchers defined the objective value function \(v(x)\), which represents maximizing the expected utility of discounted dividend payments given the initial reserve \(x\):
\[
v(x)=\sup_{(d_t)_{t\geq0}}\mathbb{E}_x\left(\int_0^{\tau}e^{-\beta t}U(d_t)dt\right)
\]
where \(\beta\) is the discount factor, \(U\) is a fixed differentiable utility function, which is equal to 0 on the negative half - axis, and \(\mathbb{E}_x\) represents the expectation under the condition \(X_0 = x\).
3. **Main Challenges**:
- The paper focuses on utility functions in the form of power functions and logarithmic functions, and analyzes the asymptotic behavior of this value function when the initial reserve \(x\) tends to infinity. Since the claim amounts follow an exponential distribution, this makes the problem more complex.
4. **Methods and Results**:
- The optimal strategy is determined by solving the Hamilton - Jacobi - Bellman (HJB) equation. For different utility functions (such as power functions and logarithmic functions), the corresponding asymptotic solutions are derived.
- For the power utility function \(U(x)=\frac{x^{\alpha}}{\alpha}\), when \(x\rightarrow\infty\), the asymptotic solution is:
\[
v(x)\sim\left(\frac{1 - \alpha}{\beta}\right)^{1-\alpha}\frac{x^{\alpha}}{\alpha}
\]
- For the logarithmic utility function \(U(x)=\ln(x + 1)\), when \(x\rightarrow\infty\), the asymptotic solution is:
\[
v(x)\sim\frac{1}{\beta}(\ln(\beta(x + 1))-1)
\]
In summary, this paper aims to find the optimal strategy for maximizing the expected utility of dividend payments under the classical risk model through strict mathematical analysis, and presents the asymptotic behavior of these strategies when the initial reserve tends to infinity.