De Finetti's problem with fixed transaction costs and regime switching

Wenyuan Wang, Zuo Quan Xu, Kazutoshi Yamazaki, Kaixin Yan, Xiaowen ZhouarXiv 2025 · DOI · arXiv

Abstract

This paper investigates a modified version of de Finetti's optimal dividend problem, incorporating fixed transaction costs and a surplus process characterized by regime-switching drift and volatility coefficients. By introducing a threshold-based transition in the company's financial dynamics, we capture the shift between different operational phases, such as high-growth startups and mature enterprises. We demonstrate that the optimal dividend strategy—which maximizes the expected total net dividend payments until ruin—is a two-barrier impulsive strategy. This paper provides explicit mathematical characterizations for these barriers across various financial scenarios and explores how changes in drift, volatility, and transaction costs influence the optimal policy.

Introduction

In actuarial science and mathematical finance, De Finetti’s problem addresses a fundamental corporate finance question: How should a company distribute dividends to maximize the total expected value for shareholders before the eventual occurrence of ruin? While the classic model provides a foundation, it often relies on two oversimplifications: continuous, costless payouts and static growth/risk dynamics.

In reality, issuing dividends incurs fixed administrative and transaction costs, making "lump sum" payments more practical than continuous streams. Furthermore, a company’s risk profile (volatility) and growth rate (drift) are rarely constant; they often shift based on the company's current capital reserves. This paper bridges these gaps by analyzing an impulse control problem with endogenous regime switching, where the surplus process dynamics change once a specific threshold is crossed.

The Model and Methodology

The researchers model the company's surplus using a Stochastic Differential Equation (SDE) where the drift ($\mu$) and volatility ($\sigma$) coefficients switch values when the surplus crosses a fixed threshold $a$. This creates a piecewise-defined environment:

  • Regime 1 ($X_t < a$): Characterized by $(\mu_1, \sigma_1)$.
  • Regime 2 ($X_t \geq a$): Characterized by $(\mu_2, \sigma_2)$.

To identify the optimal strategy, the authors employ several advanced mathematical frameworks:

  • Impulsive Control Theory: This accounts for discrete "jumps" in the surplus process (the dividend payments) rather than continuous flows.
  • Quasi-Variational Inequalities (QVI): The authors utilize a "Verification Lemma" to prove that their candidate value function is the global maximum by ensuring it satisfies specific differential inequalities across the regime boundary.
  • Martingale Approach: This is used to solve the "exit problem" for the surplus process, allowing for the derivation of explicit expressions for the expected dividend function.

Key Insights: The Two-Barrier Strategy

The core finding of the research is that the optimal approach is an $(s, S)$ type policy, specifically a two-barrier impulsive strategy $(z_1, z_2)$. Under this policy, the company allows its surplus to grow until it hits an upper barrier $z_2$. At that moment, it pays out a dividend of size $(z_2 - z_1)$, instantly resetting the surplus to a lower level $z_1$.

The Impact of Regime Switching

Unlike models where regimes switch due to external economic shocks, this model features endogenous switching triggered by the company's own success or failure in accumulating cash. This leads to several strategic nuances:

  • Startup vs. Mature Logic: If the lower regime represents a high-growth "startup" phase, the optimal strategy may involve a lower $z_2$ to frequently "reset" the company back into that high-growth zone.
  • Interdependence: The optimal levels for $z_1$ and $z_2$ are not independent; they are tightly coupled with each other and the regime-switching threshold $a$.

Results and Sensitivity Analysis

The paper provides closed-form expressions for the optimal dividend barriers for almost all parameter combinations. Key observations include:

  • Sensitivity to Costs: As the fixed transaction cost ($\beta$) increases, the distance between the two barriers $(z_2 - z_1)$ grows. When it is expensive to issue a dividend, the company waits longer to accumulate a larger surplus and pays out a larger lump sum.
  • Threshold Effects: The authors identify a "sharp kink" in strategy when the switching threshold $a$ changes. A company might shift from a conservative strategy (holding high cash reserves) to an aggressive one (paying dividends immediately before entering a less profitable regime) based on the location of $a$.
  • Optimality Proof: The study successfully proves that the two-barrier strategy remains optimal even when the underlying process has discontinuous growth and risk rates, representing a significant mathematical advancement over spatially homogeneous models.

Limitations and Future Directions

While the model provides a robust framework for understanding dividend optimization, it currently considers only two regimes and a fixed threshold $a$. Real-world scenarios may involve multiple thresholds or a continuum of regimes. Additionally, the current model focuses on fixed costs; future research could incorporate proportional costs (such as taxes) to further enhance the model's realism. Finally, exploring scenarios where the company can influence the threshold $a$ through strategic investment remains an open area for investigation.

Summary updated

Abstract

In this paper, we examine a modified version of de Finetti's optimal dividend problem, incorporating fixed transaction costs and altering the surplus process by introducing two-valued drift and two-valued volatility coefficients. This modification aims to capture the transitions or adjustments in the company's financial status. We identify the optimal dividend strategy, which maximizes the expected total net dividend payments (after accounting for transaction costs) until ruin, as a two-barrier impulsive dividend strategy. Notably, the optimal strategy can be explicitly determined for almost all scenarios involving different drifts and volatility coefficients. Our primary focus is on exploring how changes in drift and volatility coefficients influence the optimal dividend strategy.
BibTeX
@misc{wang2025de,
  author = {Wenyuan Wang and Zuo Quan Xu and Kazutoshi Yamazaki and Kaixin Yan and Xiaowen Zhou},
  title = {{De Finetti's problem with fixed transaction costs and regime switching}},
  year = {2025},
  eprint = {2502.05839},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2502.05839},
  howpublished = {arXiv},
}
Figure