Paper: arXiv 2311.05781
Authors: Hansjoerg Albrecher, Pablo Azcue, Nora Muler
Abstract
In this paper we study the problem of optimally paying out dividends from an insurance portfolio, when the criterion is to maximize the expected discounted dividends over the lifetime of the company and the portfolio contains claims due to natural catastrophes, modelled by a shot-noise Cox claim number process. The optimal value function of the resulting two-dimensional stochastic control problem is shown to be the smallest viscosity supersolution of a corresponding Hamilton-Jacobi-Bellman equation, and we prove that it can be uniformly approximated through a discretization of the space of the free surplus of the portfolio and the current claim intensity level. We implement the resulting numerical scheme to identify optimal dividend strategies for such a natural catastrophe insurer, and it is shown that the nature of the barrier and band strategies known from the classical models with constant Poisson claim intensity carry over in a certain way to this more general situation, leading to action and non-action regions for the dividend payments as a function of the current surplus and intensity level. We also discuss some interpretations in terms of upward potential for shareholders when including a catastrophe sector in the portfolio.
Complexity vs Empirical Score
- Math Complexity: 8.5/10
- Empirical Rigor: 3.0/10
- Quadrant: Lab Rats — high math complexity, limited empirical testing
Why this score: The paper involves high-level mathematics including stochastic control, Hamilton-Jacobi-Bellman equations, and viscosity solutions, but the implementation is limited to numerical schemes on discretized spaces without actual backtesting or empirical datasets.
Research Flowchart
flowchart TD A["Research Goal:<br>Maximize expected discounted dividends<br>for catastrophe insurer using optimal payout"] --> B["Methodology:<br>Stochastic Control & Hamilton-Jacobi-Bellman Analysis"] B --> C["Model Specification:<br>Shot-noise Cox claim process<br>(Surplus & Claim Intensity State Vars)"] C --> D["Analytical Step:<br>Prove value function as smallest viscosity supersolution"] D --> E["Computational Step:<br>Discretize space & apply uniform approximation scheme"] E --> F["Data/Implementation:<br>Numerical scheme identifies optimal strategies<br>(Barrier & Band strategies)"] F --> G["Key Findings/Outcomes:<br>Action/Non-action regions defined by surplus & intensity.<br>Catastrophe sector increases upward shareholder potential"]
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper proposes a numerical scheme that discretizes both surplus and claim‑intensity spaces to uniformly approximate the optimal dividend value function for a catastrophe insurer.
Source sentence
The optimal value function of the resulting two-dimensional stochastic control problem is shown to be the smallest viscosity supersolution of a corresponding Hamilton-Jacobi-Bellman equation, and we prove that it can be uniformly approximated through a discretization of the space of the free surplus of the portfolio and the current claim intensity level. This will then allow us to determine numerically the value
Assumes 6
- Claims arrive according to a Cox (shot‑noise) process with intensity λt. model
Source sentence
We assume that the intensity λt is a shot-noise process, that is λt = λc t + e Nt X k=1 Yk e−d(t−Tk), (2) where λc t = λ + e−dt (λ −λ) .
- Claim sizes Ui are i.i.d. positive random variables with finite expectation. model
Source sentence
All claims are assumed to be i.i.d. positive random variables with distribution function FU and finite expectation.
- The claim‑intensity process λt can be observed by the insurer. model
Source sentence
Under the assumption that the intensity process can be observed (which we tacitly assume here, but it is not unrealistic, as the jumps in the shot-noise process of the intensity are the documented catastrophes and the used decay function may be assumed known as a consequence of a modelling approach on past claim settlement experiences),
- The surplus process follows Xt = x + pt – Σ_{j=1}^{Nt} Uj. model
Source sentence
2 The model Consider a free surplus process of an insurance portfolio given by Xt = x + pt − Nt X j=1 Uj, (1) where x is the initial surplus, p is the premium rate and Ui is the size of the i-th claim (arriving at time τi).
Show 2 more
- Premium rate p is set as p = (1+η)E(U1)λav with η>0. model
Source sentence
That is, p = (1 + η)E(U1) lim t→∞E(Λt t ) := (1 + η)E(U1)λav, (10) where η > 0 is the relative safety loading.
- The intensity jumps Yk are i.i.d. positive with finite expectation. model
Source sentence
Tk corresponds to the arrival times of catastrophes that produce Figure 2.1: Sample path of λt jumps in the intensity λt and the upward jumps Yk are assumed to be i.i.d. positive random variables with distribution function FY and finite expectation.
Data
No dataset named in the text.
Costs
Not applicable
Finds 5
- The optimal value function is identified as the smallest viscosity supersolution of the associated HJB equation. theoretical
Source sentence
9 Conclusion and Outlook In this paper we solved the two-dimensional stochastic control problem of optimizing expected discounted dividends until ruin of an insurance portfolio, when the claim number process is a Cox process with shot-noise intensity function. We identified the optimal value function as the smallest viscosity supersolution of a Hamilton-Jacobi-Bellman equation and provided a numerical scheme to
- Barrier and band dividend strategies from classical constant‑intensity models extend to the shot‑noise intensity setting, with levels depending on the current intensity. theoretical
Source sentence
We implement the resulting numerical scheme to identify optimal dividend strategies for such a natural catastrophe insurer, and it is shown that the nature of the barrier and band strategies known from the classical models with constant Poisson claim intensity carry over in a certain way to this more general situation, leading to action and non-action regions for the dividend payments as a function of the current
- The additional variability of claim intensity creates upward potential for shareholders, allowing higher expected discounted dividends. theoretical
Source sentence
It will turn out that the additional variability of the claim intensity process leads to an upward potential for the shareholders of the insurance company. As it turns out, the additional variability of the claim occurrence pattern can be used to the advantage of the insurer as far as a valuation of the company in terms of expected discounted future dividends is concerned.
- The proposed discretization scheme converges uniformly to the optimal value function as the grid is refined. theoretical
Source sentence
Theorem 6.9 For any δ > 0, the functions V δk ↗V uniformly as k goes to infinity. We identified the optimal value function as the smallest viscosity supersolution of a Hamilton-Jacobi-Bellman equation and provided a numerical scheme to uniformly approximate it via a discretization of the surplus space and the intensity space.
- Numerical implementation confirms that optimal dividend strategies deviate from classical ones but retain the barrier/band structure. performance
Source sentence
We implement the resulting numerical scheme to identify optimal dividend strategies for such a natural catastrophe insurer, and it is shown that the nature of the barrier and band strategies known from the classical models with constant Poisson claim intensity carry over in a certain way to this more general situation, leading to action and non-action regions for the dividend payments as a function of the current
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 21 related papers). Coverage is partial, so this is neither replication nor refutation.
1 paper in the archive : Optimal dividend and capital injection under self-exciting …
Code
No repository by the authors found in the paper.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.