Optimización de dividendos bajo una tasa estocástica y con cambio de régimen
No hay miniatura disponible
Fecha
2018-10-22
Autores
Título de la revista
ISSN de la revista
Título del volumen
Editor
Pontificia Universidad Católica del Perú
DOI
Resumen
En el presente trabajo, estudiaremos el problema de optimización de dividendos para
una compañía de seguros cuya reserva de efectivo y la tasa de interés de descuento
son modelados por procesos de difusión con los coeficientes de la tendencia y la volatilidad
dependiendo del régimen económico externo (condiciones macroeconómicas). Este
cambio de régimen está modelado por una cadena de Markov observable de estados finitos.
El objetivo es encontrar un esquema de distribución de dividendos que maximice
el valor esperado de los dividendos acumulados descontados hasta el tiempo de ruina.
Consideramos dos escenarios:
(I) Cuando el proceso de dividendos tiene una tasa y esta es uniformemente
acotada. En este caso, probaremos un Teorema de verificación que indica que la
soluci´on de la ecuación Hamilton-Jacobi-Bellman correspondiente coincide con
la función de valor asociada a nuestro problema y que bajo ciertas condiciones
una estrategia óptima existe. Además, encontraremos una forma explícita de una
estrategia óptima, en el caso de dos regímenes. Esta estrategia consiste en que la
compañía pagar´a dividendos con la tasa máxima siempre y cuando el proceso de
reservas después de pagar dividendos sea igual o mayor a algunos niveles críticos
(barreras) y no pagar nada cuando se encuentre por debajo de estas barreras.
(II) En general, cuando el proceso de dividendos es solo cadlag. En este caso,
obtenemos una cota superior para la función de valor asociada a nuestro problema.
Adema´s, a partir de los resultados obtenidos en la literatura existente en problemas
similares y de los resultados obtenidos en el presente trabajo conjeturamos
una posible forma de la estrategia óptima.
In the present work, we will study the dividend optimization problem for an insurance entity whose cash surplus process and the discounting interest rate are modeled by diffusion processes with drift and volatility coefficients dependent on the extern economic regime (macroeconomic conditions). This regime switching is modeled by an observable finite-sate Markov chain. The aim is to find a dividend distribution policy that maximizes the expected total discounted amount of dividend payments up to bankruptcy. We consider two situations: (I) When the dividend process has a rate and this is uniformly bounded. In this case, we will prove a verification Theorem which indicates that the solution of the Hamilton-Jacobi-Bellman equation corresponding coincides with the value function associated with our problem and that under certain conditions an optimal strategy exists. Also, we will find an explicit form for optimal dividend strategy, in the case of two regimenes. This consists in that the company will pay out dividends at the maximun rate as long as the reserve process after the payment of pay dividends is bigger than or equal to than some critical levels (barriers) and do not pay dividends when is below these barriers. (II) In general, when the dividends process is cadlag only. In this case, we get an upper bound for the value function associated with our problem. Also, from the results obtained in the existing literature in similar problems and the results obtained in the present work, we conjecture a possible form of an optimal strategy.
In the present work, we will study the dividend optimization problem for an insurance entity whose cash surplus process and the discounting interest rate are modeled by diffusion processes with drift and volatility coefficients dependent on the extern economic regime (macroeconomic conditions). This regime switching is modeled by an observable finite-sate Markov chain. The aim is to find a dividend distribution policy that maximizes the expected total discounted amount of dividend payments up to bankruptcy. We consider two situations: (I) When the dividend process has a rate and this is uniformly bounded. In this case, we will prove a verification Theorem which indicates that the solution of the Hamilton-Jacobi-Bellman equation corresponding coincides with the value function associated with our problem and that under certain conditions an optimal strategy exists. Also, we will find an explicit form for optimal dividend strategy, in the case of two regimenes. This consists in that the company will pay out dividends at the maximun rate as long as the reserve process after the payment of pay dividends is bigger than or equal to than some critical levels (barriers) and do not pay dividends when is below these barriers. (II) In general, when the dividends process is cadlag only. In this case, we get an upper bound for the value function associated with our problem. Also, from the results obtained in the existing literature in similar problems and the results obtained in the present work, we conjecture a possible form of an optimal strategy.
Descripción
Palabras clave
Optimización de dividendos, Control óptimo esctocástico
Citación
Colecciones
item.page.endorsement
item.page.review
item.page.supplemented
item.page.referenced
Licencia Creative Commons
Excepto se indique lo contrario, la licencia de este artículo se describe como info:eu-repo/semantics/openAccess