The paper deals with a generalization of the risk model with stochastic premiums where dividends are paid according to a multi-layer dividend strategy. First of all, we derive piecewise integro-differential equations for the Gerber–Shiu function and the expected discounted dividend payments until ruin. In addition, we concentrate on the detailed investigation of the model in the case of exponentially distributed claim and premium sizes and find explicit formulas for the ruin probability as well as for the expected discounted dividend payments. Lastly, numerical illustrations for some multi-layer dividend strategies are presented.
Risk model with stochastic premiumsmulti-layer dividend strategyGerber–Shiu functionexpected discounted dividend paymentsruin probabilitypiecewise integro-differential equation91B3060G51Introduction
The ruin measures such as the ruin probability, the surplus prior to ruin and the deficit at ruin have attracted great interest of researchers recently (see, e.g., [2, 28, 33] and references therein). Gerber and Shiu [16] introduced the expected discounted penalty function for the classical risk model, which enabled to study those risk measures together by combining them into one function. After that, the so-called Gerber–Shiu function has been investigated by many authors in more general risk models (see, e.g., [6–8, 10, 11, 17, 18, 36, 39, 45, 47]).
In particular, a lot of attention has been paid to the study of risk models where shareholders receive dividends from their insurance company. De Finetti [14], who first considered dividend strategies in insurance, dealt with a binomial model. For the classical risk model and its different generalizations, different dividend strategies have been studied in a number of papers (see, e.g., [9, 12, 13, 20, 22–24, 26, 27, 35, 37, 40]). In addition, the monograph by Schmidli [34] is devoted to optimal dividend problems in insurance risk models.
Applying multi-layer dividend strategies enables to change the dividend payment intensity depending on the current surplus. Albrecher and Hartinger [1] consider the modification of the classical risk model where both the premium intensity and the dividend payment intensity are assumed to be step functions depending on the current surplus level. The authors derive algorithmic schemes for the determination of explicit expressions for the Gerber–Shiu function and the expected discounted dividend payments. A similar risk model is considered by Lin and Sendova [25], who derive a piecewise integro-differential equation for the Gerber–Shiu function and provide a recursive approach to obtain general solutions to that equation and its generalizations. Developing a recursive algorithm to calculate the moments of the expected discounted dividend payments for a class of risk models with Markovian claim arrivals, Badescu and Landriault [3] generalize some of the results obtained in [1] (see also [4] for some results related to the class of Markovian risk models with a multi-layer dividend strategy).
The absolute ruin problem in the classical risk model with constant interest force and a multi-layer dividend strategy is investigated in [43], where a piecewise integro-differential equation for the discounted penalty function is derived, some explicit expressions are given when claims are exponentially distributed and an asymptotic formula for the absolute ruin probability is obtained for heavy-tailed claim sizes. The dual model of the compound Poisson risk model with a multi-layer dividend strategy under stochastic interest is considered in [44]. Results related to perturbed compound Poisson risk models under multi-layer dividend strategies can be found in [31, 42]. In addition, different classes of more general renewal risk models are investigated in [15, 19, 40, 41], and some recent papers deal with risk models that incorporate various dependence structures (see, e.g., [21, 38, 46, 48]).
The present paper generalizes the risk model with stochastic premiums introduced and investigated in [5] (see also [28]). In that risk model, both claims and premiums are modeled as compound Poisson processes, whereas premiums arrive with constant intensity and are not random in the classical compound Poisson risk model (see also [29, 30] for a generalization of the classical risk model where an insurance company gets additional funds whenever a claim arrives). In [5], claim sizes and inter-claim times are assumed to be mutually independent, and the same assumption is made concerning premium arrivals. In contrast to [5], the recent paper [32] deals with the risk model with stochastic premiums where the dependence structures between claim sizes and inter-claim times as well as premium sizes and inter-premium times are modeled by the Farlie–Gumbel–Morgenstern copulas, and dividends are paid according to a threshold dividend strategy. The Gerber–Shiu function, a special case of which is the ruin probability, and the expected discounted dividend payments until ruin are studied in [32]. In the present paper, we develop those results and make the assumption that dividends are paid according to a multi-layer dividend strategy and all random variables and processes are mutually independent.
The rest of the paper is organized as follows. In Section 2, we give a description of the risk model with stochastic premiums and a multi-layer dividend strategy. In Sections 3 and 4, we derive piecewise integro-differential equations for the Gerber–Shiu function and the expected discounted dividend payments until ruin. Next, in Section 5, we deal with exponentially distributed claim and premium sizes and obtain explicit formulas for the ruin probability and the expected discounted dividend payments. Finally, Section 6 provides some numerical illustrations.
Description of the model
Let (Ω,F,P) be a probability space satisfying the usual conditions, and let all the stochastic objects we use below be defined on it.
In the risk model with stochastic premiums introduced in [5] (see also [28]), claim sizes form a sequence (Yi)i≥1 of non-negative independent and identically distributed (i.i.d.) random variables (r.v.’s) with cumulative distribution function (c.d.f.) FY(y)=P[Yi≤y], and the number of claims on the time interval [0,t] is a Poisson process (Nt)t≥0 with constant intensity λ>00$]]>. In addition, premium sizes form a sequence (Y¯i)i≥1 of non-negative i.i.d. r.v.’s with c.d.f. F¯Y¯(y)=P[Y¯i≤y], and the number of premiums on the time interval [0,t] is a Poisson process (N¯t)t≥0 with constant intensity λ¯>00$]]>. Thus, the total claims and premiums on [0,t] equal ∑i=1NtYi and ∑i=1N¯tY¯i, respectively.
It is worth pointing out that, here and subsequently, a sum is always set to 0 if the upper summation index is less than the lower one. In particular, we have ∑i=10Yi=0 if Nt=0, and ∑i=10Y¯i=0 if N¯t=0. In what follows, we also assume that the r.v.’s (Yi)i≥1 and (Y¯i)i≥1 have finite expectations μ>00$]]> and μ¯>00$]]>, respectively. Furthermore, we suppose that (Yi)i≥1, (Y¯i)i≥1, (Nt)t≥0 and (N¯t)t≥0 are mutually independent.
Next, we denote a non-negative initial surplus of the insurance company by x, and let Xt(x) be its surplus at time t provided that the initial surplus is x. Then the surplus process (Xt(x))t≥0 is defined by the equality
Xt(x)=x+∑i=1N¯tY¯i−∑i=1NtYi,t≥0.
In contrast to the risk model considered in [5], we make the additional assumption that the insurance company pays dividends to its shareholders according to a k-layer dividend strategy with k≥2. Let b=(b1,…,bk−1) be a (k−1)-dimensional vector with real-valued components such that 0<b1<⋯<bk−1<∞. Besides that, we set b0=0 and bk=∞. Let (Xtb(x))t≥0 denote the modified surplus process under the k-layer dividend strategy b, which implies that dividends are paid continuously at a rate dj>00$]]> whenever bj−1≤Xtb(x)<bj, i.e. the process (Xtb(x))t≥0 is in the jth layer at time t, where 1≤j≤k. Then
Xtb(x)=x+∑i=1N¯tY¯i−∑i=1NtYi−∫0t∑j=1kdj1(bj−1≤Xsb(x)<bj)ds,t≥0,
where 1(·) is the indicator function.
From now on, we suppose that the net profit condition holds, which in this case means that
λ¯μ¯>λμ+max1≤j≤k{dj}.\lambda \mu +\underset{1\le j\le k}{\max }\{{d_{j}}\}.\]]]>
Let (Dt)t≥0 denote the dividend distributing process. For the k-layer dividend strategy described above, we have
dDt=djdsifbj−1≤Xtb(x)<bj,1≤j≤k.
Next, let τb(x)=inf{t≥0:Xtb(x)<0} be the ruin time for the risk process (Xtb(x))t≥0 defined by (2). In what follows, we omit the dependence on x and write τb instead of τb(x) when no confusion can arise.
For δ0≥0, the Gerber–Shiu function is defined by
m(x,b)=E[e−δ0τbw(Xτb−b(x),|Xτbb(x)|)1(τb<∞)|X0b(x)=x],x≥0,
where w(·,·) is a bounded non-negative measurable function, Xτb−b(x) is the surplus immediately before ruin and |Xτbb(x)| is a deficit at ruin. Note that if w(·,·)≡1 and δ0=0, then m(x,b) becomes the infinite-horizon ruin probability
ψ(x,b)=E[1(τb<∞)|X0b(x)=x].
For δ>00$]]>, the expected discounted dividend payments until ruin are defined by
v(x,b)=E[∫0τbe−δtdDt|X0b(x)=x],x≥0.
For simplicity of notation, we also write m(x), ψ(x) and v(x) instead of m(x,b), ψ(x,b) and v(x,b), respectively. For all 1≤j≤k and bj−1≤x≤bj, we also set mj(x)=m(x,b), ψj(x)=ψ(x,b) and vj(x)=v(x,b). Thus, the functions mj(x), ψj(x) and vj(x) are defined on [bj−1,bj], and we have mj(bj)=mj+1(bj), ψj(bj)=ψj+1(bj) and vj(bj)=vj+1(bj) for all 1≤j≤k−1.
Note that although we consider the interval [bk−1,∞) instead of [bj−1,bj] if j=k, for the sake of convenience and compactness, here and subsequently, we do write [bj−1,bj] for all 1≤j≤k. In addition, in what follows, the derivatives of all functions at the ends of the closed intervals [bj−1,bj] are assumed to be one-sided.
Piecewise integro-differential equation for the Gerber–Shiu function
Let the surplus process(Xtb(x))t≥0be defined by (2) under the above assumptions, and letFY(y)andw(u1,u2)be continuous onR+andR+2, respectively. Then the functionm(x)is differentiable on the intervals[bj−1,bj]for all1≤j≤kand satisfies the piecewise integro-differential equationdjm′(x)+(λ+λ¯+δ0)m(x)=λ∫0xm(x−y)dFY(y)+λ∫x∞w(x,y−x)dFY(y)+λ¯∫0∞m(x+y)dFY¯(y),x∈[bj−1,bj].
We now fix any j such that 1≤j≤k and deal with the case x∈[bj−1,bj]. For all x∈[bj−1,bj], we define the following functions:
a1(x)=(x−bj−1)/dj+(bj−1−bj−2)/dj−1+⋯+(b2−b1)/d2+(b1−b0)/d1,a2(x)=(x−bj−1)/dj+(bj−1−bj−2)/dj−1+⋯+(b2−b1)/d2,…aj−1(x)=(x−bj−1)/dj+(bj−1−bj−2)/dj−1,aj(x)=(x−bj−1)/dj.
From these equalities we conclude that for all x∈[bj−1,bj], the process (Xtb(x))t≥0 up to its first jump is in the jth layer if t∈[0,aj(x)] and in the ith layer if t∈[ai+1(x),ai(x)], where 1≤i≤j−1. Thus, for any x∈[bj−1,bj], the sequence aj(x), aj−1(x),…,a1(x) defines the times when (Xtb(x))t≥0 passes through the values bj−1, bj−2,…,b0 provided that it has no jumps until those times.
It is easily seen that the time of the first jump of (Xtb(x))t≥0 is exponentially distributed with mean 1/(λ+λ¯). Considering the time and the size of the first jump of that process and applying the law of total probability we obtain
m(x)=Ij(x)+Ij−1(x)+⋯+I1(x)+I0(x),x∈[bj−1,bj],
where
Ij(x)=∫0aj(x)e−(λ+λ¯)t(λ∫0x−djte−δ0tm(x−djt−y)dFY(y)+λ∫x−djt∞e−δ0tw(x−djt,y−x+djt)dFY(y)+λ¯∫0∞e−δ0tm(x−djt+y)dFY¯(y))dt,Ij−1(x)=∫aj(x)aj−1(x)e−(λ+λ¯)t(λ∫0bj−1−dj−1(t−aj(x))e−δ0t×m(bj−1−dj−1(t−aj(x))−y)dFY(y)+λ∫bj−1−dj−1(t−aj(x))∞e−δ0tw(bj−1−dj−1(t−aj(x)),y−bj−1+dj−1(t−aj(x)))dFY(y)+λ¯∫0∞e−δ0tm(bj−1−dj−1(t−aj(x))+y)dFY¯(y))dt,…I1(x)=∫a2(x)a1(x)e−(λ+λ¯)t(λ∫0b1−d1(t−a2(x))e−δ0t×m(b1−d1(t−a2(x))−y)dFY(y)+λ∫b1−d1(t−a2(x))∞e−δ0tw(b1−d1(t−a2(x)),y−b1+d1(t−a2(x)))dFY(y)+λ¯∫0∞e−δ0tm(b1−d1(t−a2(x))+y)dFY¯(y))dt,I0(x)=e−(λ+λ¯+δ0)a1(x)w(0,0).
Note that the term Ii(x), 1≤i≤j, in (5) corresponds to the case where (Xtb(x))t≥0 is in the ith layer when its first jump occurs, and the term I0(x) corresponds to the case where there are no jumps of (Xtb(x))t≥0 up to the time a1(x).
Changing the variable x−djt=s in the outer integral in the expression for Ij(x) yields
Ij(x)=1dje−(λ+λ¯+δ0)x/dj∫bj−1xe(λ+λ¯+δ0)s/dj(λ∫0sm(s−y)dFY(y)+λ∫s∞w(s,y−s)dFY(y)+λ¯∫0∞m(s+y)dFY¯(y))ds.
Changing the variable bj−1−dj−1(t−aj(x))=s in the outer integral in the expression for Ij−1(x) gives
Ij−1(x)=1dj−1e−(λ+λ¯+δ0)(aj(x)+bj−1/dj−1)×∫bj−2bj−1e(λ+λ¯+δ0)s/dj−1(λ∫0sm(s−y)dFY(y)+λ∫s∞w(s,y−s)dFY(y)+λ¯∫0∞m(s+y)dFY¯(y))ds.
In the same manner we change variables in all the outer integrals on the right-hand side of (5). Finally, changing the variable b1−d1(t−a2(x))=s in the outer integral in the expression for I1(x) yields
I1(x)=1d1e−(λ+λ¯+δ0)(a2(x)+b1/d1)∫b0b1e(λ+λ¯+δ0)s/d1×(λ∫0sm(s−y)dFY(y)+λ∫s∞w(s,y−s)dFY(y)+λ¯∫0∞m(s+y)dFY¯(y))ds.
Thus, from the above and equality (5) we deduce that m(x) is continuous on [bj−1,bj], and hence, on R+. Therefore, (6) implies that Ij(x) is differentiable on [bj−1,bj], and for all x∈[bj−1,bj], we get
Ij′(x)=−λ+λ¯+δ0djIj(x)+1dj(λ∫0xm(x−y)dFY(y)+λ∫x∞w(x,y−x)dFY(y)+λ¯∫0∞m(x+y)dFY¯(y)).
Moreover, it is easily seen, e.g. from (7) and (8), that all the functions Ij−1(x),…,I1(x) and I0(x) are differentiable on [bj−1,bj], and for all x∈[bj−1,bj], we have
Ij−1′(x)=−λ+λ¯+δ0djIj−1(x),…,I1′(x)=−λ+λ¯+δ0djI1(x),I0′(x)=−λ+λ¯+δ0djI0(x).
From (5) it follows that m(x) is also differentiable on [bj−1,bj]. Differentiating (5) and taking into account expressions for Ij′(x), Ij−1′(x),…,I1′(x), I0′(x) we obtain
m′(x)=−λ+λ¯+δ0djm(x)+1dj(λ∫0xm(x−y)dFY(y)+λ∫x∞w(x,y−x)dFY(y)+λ¯∫0∞m(x+y)dFY¯(y)),x∈[bj−1,bj],
from which equation (4) follows immediately. □
To solve equation (4), we use the following boundary conditions. The first k−1 conditions are easily obtained from the equality mj(bj)=mj+1(bj) or, equivalently, limx↑bjm(x)=limx↓bjm(x) for all 1≤j≤k−1. In addition, for the ruin probability, using standard considerations (see, e.g., [28, 30, 33]) we can show that limx→∞ψ(x)=0 provided that the net profit condition holds. Finally, it is evident that ψ(0)=1. Although equation (4) is not solvable analytically in the general case, we can find explicit expressions for the corresponding ruin probability in the case where claim and premium sizes are exponentially distributed (see Section 5). The uniqueness of the required solution to (4) should be justified in each case.
In the assertion of Theorem 1, we require the continuity of FY(y). Note that if FY(y) has positive points of discontinuity, then m(x) may be not differentiable at some interior points of the intervals [bj−1,bj], 1≤j≤k (for details, see [28, 30]). Moreover, it is easily seen from (4) that m(x) is not differentiable at x=bj, 1≤j≤k−1, since its one-sided derivatives do not coincide at those points.
Piecewise integro-differential equation for the expected discounted dividend payments until ruin
Let the surplus process(Xtb(x))t≥0be defined by (2) under the above assumptions, and letFY(y)be continuous onR+. Then the functionv(x)is differentiable on the intervals[bj−1,bj]for all1≤j≤kand satisfies the piecewise integro-differential equationdjv′(x)+(λ+λ¯+δ)v(x)=λ∫0xv(x−y)dFY(y)+λ¯∫0∞v(x+y)dFY¯(y)+dj,x∈[bj−1,bj].
We now fix any j such that 1≤j≤k and deal with the case x∈[bj−1,bj]. As in the proof of Theorem 1, considering the time and the size of the first jump of (Xtb(x))t≥0 and applying the law of total probability we have
v(x)=I1,j(x)+I2,j(x)+I1,j−1(x)+I2,j−1(x)+⋯+I1,1(x)+I2,1(x)+I1,0(x),x∈[bj−1,bj],
where
I1,j(x)=∫0aj(x)(λ+λ¯)e−(λ+λ¯)t(∫0tdje−δsds)dt,I2,j(x)=∫0aj(x)e−(λ+λ¯)t(λ∫0x−djte−δtv(x−djt−y)dFY(y)+λ¯∫0∞e−δtv(x−djt+y)dFY¯(y))dt,I1,j−1(x)=∫aj(x)aj−1(x)(λ+λ¯)e−(λ+λ¯)t(∫0aj(x)dje−δsds+∫aj(x)tdj−1e−δsds)dt,I2,j−1(x)=∫aj(x)aj−1(x)e−(λ+λ¯)t(λ∫0bj−1−dj−1(t−aj(x))e−δt×v(bj−1−dj−1(t−aj(x))−y)dFY(y)+λ¯∫0∞e−δtv(bj−1−dj−1(t−aj(x))+y)dFY¯(y))dt,…I1,1(x)=∫a2(x)a1(x)(λ+λ¯)e−(λ+λ¯)t(∫0aj(x)dje−δsds+∫aj(x)aj−1(x)dj−1e−δsds+⋯+∫a3(x)a2(x)d2e−δsds+∫a2(x)td1e−δsds)dt,I2,1(x)=∫a2(x)a1(x)e−(λ+λ¯)t(λ∫0b1−d1(t−a2(x))e−δt×v(b1−d1(t−a2(x))−y)dFY(y)+λ¯∫0∞e−δtv(b1−d1(t−a2(x))+y)dFY¯(y))dt,I1,0(x)=∫a1(x)∞(λ+λ¯)e−(λ+λ¯)t(∫0aj(x)dje−δsds+∫aj(x)aj−1(x)dj−1e−δsds+⋯+∫a3(x)a2(x)d2e−δsds+∫a2(x)a1(x)d1e−δsds)dt,
and the functions a1(x), a2(x),…,aj(x) are defined in the proof of Theorem 1.
Note that the terms I1,i(x) and I2,i(x), 1≤i≤j, in (10) correspond to the case where (Xtb(x))t≥0 is in the ith layer when its first jump occurs, and the term I1,0(x) corresponds to the case where there are no jumps of (Xtb(x))t≥0 up to the time a1(x). The terms I1,i(x), 0≤i≤j, are equal to the discounted dividend payments until the first jump of (Xtb(x))t≥0 provided that the process is in the ith layer, whereas the terms I2,i(x), 1≤i≤j, are equal to the corresponding expected discounted dividend payments after that time.
Next, we set
I1,∗(x)=I1,j(x)+I1,j−1(x)+⋯+I1,1(x)+I1,0(x),x∈[bj−1,bj].
Thus, I1,∗(x) describes the expected discounted dividend payments until the first jump of (Xtb(x))t≥0.
Rearranging terms in the expression for I1,∗(x) gives
I1,∗(x)=(λ+λ¯)(∫0aj(x)e−(λ+λ¯)t(∫0tdje−δsds)dt+∫aj(x)aj−1(x)e−(λ+λ¯)t(∫aj(x)tdj−1e−δsds)dt+⋯+∫a2(x)a1(x)e−(λ+λ¯)t(∫a2(x)td1e−δsds)dt+∫aj(x)∞e−(λ+λ¯)t(∫0aj(x)dje−δsds)dt+∫aj−1(x)∞e−(λ+λ¯)t(∫aj(x)aj−1(x)dj−1e−δsds)dt+⋯+∫a1(x)∞e−(λ+λ¯)t(∫a2(x)a1(x)d1e−δsds)dt).
Taking all the integrals on the right-hand side of (12) and simplifying the resulting expression we get
I1,∗(x)=djλ+λ¯+δ(1−e−(λ+λ¯+δ)aj(x))+dj−1λ+λ¯+δ(e−(λ+λ¯+δ)aj(x)−e−(λ+λ¯+δ)aj−1(x))+⋯+d1λ+λ¯+δ(e−(λ+λ¯+δ)a2(x)−e−(λ+λ¯+δ)a1(x)).
Changing the variable x−djt=s in the outer integral in the expression for I2,j(x) gives
I2,j(x)=1dje−(λ+λ¯+δ)x/dj∫bj−1xe(λ+λ¯+δ)s/dj×(λ∫0sv(s−y)dFY(y)+λ¯∫0∞v(s+y)dFY¯(y))ds.
Likewise, changing the variable bj−1−dj−1(t−aj(x))=s in the outer integral in the expression for I2,j−1(x) yields
I2,j−1(x)=1dj−1e−(λ+λ¯+δ)(aj(x)+bj−1/dj−1)∫bj−2bj−1e(λ+λ¯+δ)s/dj−1×(λ∫0sv(s−y)dFY(y)+λ¯∫0∞v(s+y)dFY¯(y))ds.
Next, in the same manner we change variables in all those outer integrals on the right-hand side of (10) that are not included in the sum (11). Eventually, changing the variable b1−d1(t−a2(x))=s in the outer integral in the expression for I2,1(x) we obtain
I2,1(x)=1d1e−(λ+λ¯+δ)(a2(x)+b1/d1)∫b0b1e(λ+λ¯+δ)s/d1×(λ∫0sv(s−y)dFY(y)+λ¯∫0∞v(s+y)dFY¯(y))ds.
From (13) it follows immediately that I1,∗′(x) is differentiable on [bj−1,bj], and for all x∈[bj−1,bj], we get
I1,∗′(x)=−λ+λ¯+δdj(I1,∗(x)−djλ+λ¯+δ).
Next, from the above and equality (10) we conclude that v(x) is continuous on [bj−1,bj], and hence, on R+. Hence, (14) implies that I2,j(x) is differentiable on [bj−1,bj], and for all x∈[bj−1,bj], we have
I2,j′(x)=−λ+λ¯+δdjI2,j(x)+1dj(λ∫0xv(x−y)dFY(y)+λ¯∫0∞v(x+y)dFY¯(y)).
Furthermore, it follows immediately, e.g. from (15) and (16), that all the functions I2,j−1(x),…,I2,1(x) are differentiable on [bj−1,bj], and for all x∈[bj−1,bj], we obtain
I2,j−1′(x)=−λ+λ¯+δdjI2,j−1(x),…,I2,1′(x)=−λ+λ¯+δdjI2,1(x).
By (10), we conclude that v(x) is also differentiable on [bj−1,bj]. Differentiating (10) and taking into account expressions for I1,∗′(x), I2,j′(x), I2,j−1′(x),…,I2,1′(x) we get
v′(x)=−λ+λ¯+δdjv(x)+1+1dj(λ∫0xv(x−y)dFY(y)+λ¯∫0∞v(x+y)dFY¯(y)),x∈[bj−1,bj],
which immediately yields equation (9). □
To solve equation (9), we obtain the first k−1 boundary conditions from the equality vj(bj)=vj+1(bj) or, equivalently, limx↑bjv(x)=limx↓bjv(x) for all 1≤j≤k−1. Moreover, if the net profit condition holds, applying arguments similar to those in [34, p. 70] we can show that limx→∞v(x)=dk/δ. Lastly, it is easily seen that v(0)=0. The uniqueness of the required solution to (9) should be justified in each case. Explicit expressions for v(x) in the case where claim and premium sizes are exponentially distributed are given in Section 5.
If FY(y) has positive points of discontinuity, then v(x) may be not differentiable at some interior points of the intervals [bj−1,bj], 1≤j≤k. Furthermore, from (9) we deduce that v(x) is not differentiable at x=bj, 1≤j≤k−1.
Exponentially distributed claim and premium sizes
In this section, we concentrate on the case where claim and premium sizes are exponentially distributed, i.e.
fY(y)=1μe−y/μandfY¯(y)=1μ¯e−y/μ¯,y≥0.
Explicit formulas for the ruin probability
Let now w(·,·)≡1 and δ0=0. Taking into account (17), equation (4) for the ruin probability can be written as
djψ′(x)+(λ+λ¯)ψ(x)=λμe−x/μ∫0xψ(u)eu/μdu+λe−x/μ+λ¯μ¯ex/μ¯∫x∞ψ(u)e−u/μ¯du
for all x∈[bj−1,bj] and 1≤j≤k.
We now reduce piecewise integro-differential equation (18) to a piecewise linear differential equation with constant coefficients.
Let the surplus process(Xtb(x))t≥0be defined by (2) under the above assumptions, and let claim and premium sizes be exponentially distributed with means μ andμ¯, respectively. Thenψ(x)is a solution to the piecewise differential equationdjμμ¯ψ‴(x)+(dj(μ¯−μ)+μμ¯(λ+λ¯))ψ″(x)+(λ¯μ¯−λμ−dj)ψ′(x)=0for allx∈[bj−1,bj]and1≤j≤k.
It is easily seen that the right-hand side of (18) is differentiable on [bj−1,bj]. Therefore, ψ(x) is twice differentiable on [bj−1,bj]. Differentiating (18) gives
djψ″(x)+(λ+λ¯)ψ′(x)=−1μ(λμe−x/μ∫0xψ(u)eu/μdu+λe−x/μ)+λ¯μ¯2ex/μ¯∫x∞ψ(u)e−u/μ¯du+(λμ−λ¯μ¯)ψ(x),x∈[bj−1,bj].
Multiplying (20) by μ and adding (18) we get
djμψ″(x)+(dj+μ(λ+λ¯))ψ′(x)+λ¯(1+μμ¯)ψ(x)=λ¯μ¯(1+μμ¯)ex/μ¯∫x∞ψ(u)e−u/μ¯du,x∈[bj−1,bj].
From (21) it follows that ψ(x) has the third derivative on x∈[bj−1,bj]. Differentiating (21) yields
djμψ‴(x)+(dj+μ(λ+λ¯))ψ″(x)+λ¯(1+μμ¯)ψ′(x)=λ¯μ¯2(1+μμ¯)ex/μ¯∫x∞ψ(u)e−u/μ¯du−λ¯μ¯(1+μμ¯)ψ(x),x∈[bj−1,bj].
Finally, multiplying (22) by (−μ¯) and adding (21) we obtain (19). □
For 1≤j≤k, we now define the following constants, which are used in the assertion of Theorem 3 below:
Dj=(dj(μ+μ¯)+μμ¯(λ−λ¯))2+4λλ¯μ2μ¯2,z1,j=−(dj(μ¯−μ)+μμ¯(λ+λ¯))+Dj2djμμ¯
and
z2,j=−(dj(μ¯−μ)+μμ¯(λ+λ¯))−Dj2djμμ¯.
Let the surplus process(Xtb(x))t≥0be defined by (2) under the above assumptions, and let claim and premium sizes be exponentially distributed with means μ andμ¯, respectively. If the net profit condition (3) holds, then we haveψj(x)=C1,jez1,jx+C2,jez2,jx+C3,jfor allx∈[bj−1,bj]and1≤j≤k, whereC3,k=0and all the other constantsC1,j,C2,jandC3,jare determined from the system of linear equations (24)–(27):λe−bj−1/μ∑l=1j−1(∑i=12Ci,lμzi,l+1(e(zi,l+1/μ)bl−e(zi,l+1/μ)bl−1)+(ebl/μ−ebl−1/μ)C3,l)+∑i=12(λ¯ebj−1/μ¯μ¯zi,j−1(e(zi,j−1/μ¯)bj−e(zi,j−1/μ¯)bj−1)−(djzi,j+λ+λ¯)ezi,jbj−1)Ci,j−(λ¯e(bj−1−bj)/μ¯+λ)C3,j+λ¯ebj−1/μ¯∑l=j+1k(∑i=12Ci,lμ¯zi,l−1(e(zi,l−1/μ¯)bl−e(zi,l−1/μ¯)bl−1)−(e−bl/μ¯−e−bl−1/μ¯)C3,l)=−λe−bj−1/μ,1≤j≤k,C1,1+C2,1+C3,1=1,dj∑i=12zi,jezi,jbjCi,j−dj+1∑i=12zi,j+1ezi,j+1bjCi,j+1=0,1≤j≤k−1,and∑i=12ezi,jbjCi,j+C3,j−∑i=12ezi,j+1bjCi,j+1−C3,j+1=0,1≤j≤k−1,provided that its determinant is not equal to 0.
Taking into account the notation introduced in Section 2 and applying Lemma 1 we conclude that the function ψj(x) is a solution to (19) on x∈[bj−1,bj] for each 1≤j≤k. The corresponding characteristic equation has the form
djμμ¯z3+(dj(μ¯−μ)+μμ¯(λ+λ¯))z2+(λ¯μ¯−λμ−dj)z=0
for all 1≤j≤k. We first show that the equation
djμμ¯z2+(dj(μ¯−μ)+μμ¯(λ+λ¯))z+(λ¯μ¯−λμ−dj)=0
has two negative roots. Indeed, its discriminant is equal to
(dj(μ¯−μ)+μμ¯(λ+λ¯))2−4djμμ¯(λ¯μ¯−λμ−dj)=dj2(μ¯−μ)2+μ2μ¯2(λ+λ¯)2+2djμμ¯(λ+λ¯)(μ¯−μ)+4djμμ¯(dj+λμ−λ¯μ¯)=dj2(μ+μ¯)2+μ2μ¯2(λ+λ¯)2+2djμμ¯(λ−λ¯)(μ+μ¯)=(dj(μ+μ¯)+μμ¯(λ−λ¯))2+4λλ¯μ2μ¯2.
Hence, it is positive and coincides with the constant Dj defined above. Consequently, z1,j and z2,j defined before the assertion of the theorem are two real roots of equation (29). By the net profit condition (3), we have
λ¯μ¯−λμ−dj>00\]]]>
and
dj(μ¯−μ)+μμ¯(λ+λ¯)=μ(λ¯μ¯−λμ−dj)+λμ2+λμμ¯+djμ¯>00\]]]>
for all 1≤j≤k, which implies that z1,j<0 and z2,j<0.
Therefore, z1,j<0, z2,j<0 and z3,j=0 are roots of equation (28), and we get (23) with some constants C1,j, C2,j and C3,j. Moreover, since condition (3) holds, using standard considerations (see, e.g., [28, 30, 33]) we can easily show that limx→∞ψ(x)=0, which yields C3,k=0.
To determine all the other constants C1,j, C2,j and C3,j, we use the following boundary conditions. The first k conditions are obtained by letting x=bj−1 in (18) for 1≤j≤k:
djψ′(bj−1)+(λ+λ¯)ψ(bj−1)=λμe−bj−1/μ∫0bj−1ψ(u)eu/μdu+λe−bj−1/μ+λ¯μ¯ebj−1/μ¯∫bj−1∞ψ(u)e−u/μ¯du.
One more condition is obtained from the equality ψ(0)=1. Finally, the last 2(k−1) conditions are derived from the equalities djψj′(bj)=dj+1ψj+1′(bj) and ψj(bj)=ψj+1(bj) for 1≤j≤k−1. Note that the first equality easily follows from (18).
Taking into account (23), for all 1≤j≤k, we get:
ψj′(x)=C1,jz1,jez1,jx+C2,jz2,jez2,jx,x∈[bj−1,bj],∫0bj−1ψ(u)eu/μdu=∑l=1j−1∫bl−1blψl(u)eu/μdu=∑l=1j−1(∑i=12Ci,lzi,l+1/μ×(e(zi,l+1/μ)bl−e(zi,l+1/μ)bl−1)+C3,lμ(ebl/μ−ebl−1/μ))
and
∫bj−1∞ψ(u)e−u/μ¯du=∑l=jk∫bl−1blψl(u)e−u/μ¯du=∑l=jk(∑i=12Ci,lzi,l−1/μ¯×(e(zi,l−1/μ¯)bl−e(zi,l−1/μ¯)bl−1)−C3,lμ¯(e−bl/μ¯−e−bl−1/μ¯)).
Substituting (23), (31), (32) and (33) into (30) and doing some simplifications yield (24). Next, from the equality ψ(0)=1 we get (25). Lastly, substituting (31) into djψj′(bj)=dj+1ψj+1′(bj) and (23) into ψj(bj)=ψj+1(bj) for 1≤j≤k−1 immediately yields (26) and (27), respectively.
Thus, we get the system of 3k−1 linear equations (24)–(27) to determine 3k−1 unknown constants. That system has a unique solution provided that its determinant is not equal to 0. Hence, piecewise differential equation (19) has a unique solution satisfying certain conditions and that solution is given by (23). Since we have derived (19) from (18) without any additional assumptions concerning the differentiability of ψ(x), we conclude that the functions ψj(x), 1≤j≤k, given by (23) are unique solutions to (18) on the intervals [bj−1,bj] satisfying certain conditions. This guaranties that the functions ψj(x) we have found coincide with the ruin probability on [bj−1,bj], which completes the proof. □
In particular, if k=2, then C3,2=0 and the constants C1,1, C2,1, C3,1, C1,2 and C2,2 are determined from the system of linear equations (34)–(38):
∑i=12(λ¯μ¯zi,1−1(1−e(zi,1−1/μ¯)b1)+d1zi,1+λ+λ¯)Ci,1+(λ¯e−b1/μ¯+λ)C3,1+∑i=12λ¯e(zi,2−1/μ¯)b1μ¯zi,2−1Ci,2=λ,∑i=12λe−b1/μμzi,1+1(1−e(zi,1+1/μ)b1)Ci,1+λ(e−b1/μ−1)C3,1+∑i=12(λ¯ezi,2b1μ¯zi,2−1+(d2zi,2+λ+λ¯)ezi,2b1)Ci,2=λe−b1/μ,C1,1+C2,1+C3,1=1,d1∑i=12zi,1ezi,1b1Ci,1−d2∑i=12zi,2ezi,2b1Ci,2=0
and
∑i=12ezi,1b1Ci,1+C3,1−∑i=12ezi,2b1Ci,2=0
provided that its determinant is not equal to 0.
The proposition below enables us to check whether the system of equations (34)–(38) has a unique solution. Let
Δ=1λ¯(μ+μ¯)2×(d1μ¯(z1,1−z2,1)(eb1/μ¯−e−b1/μ)((λ¯μ¯−λμ−d1)eb1/μ¯−(λ¯μ¯−λμ−d2))−d1μ¯eb1/μ¯(z1,1−z2,1)(λ¯μ¯−λμ−d1)(eb1/μ¯−e−b1/μ)+d1μ(1−1/μ¯)(ez1,1b1−ez2,1b1)((λ¯μ¯−λμ−d1)eb1/μ¯−(λ¯μ¯−λμ−d2))+(d2−d1)(ez1,1b1−ez2,1b1)(λ¯μ¯−λμ−d1)(eb1/μ¯−e−b1/μ)+d12μeb1/μ¯(μ¯−1)(z2,1ez1,1b1−z1,1ez2,1b1)+d1μ¯(d2−d1)(eb1/μ¯−e−b1/μ)(z2,1ez1,1b1−z1,1ez2,1b1)).
The system of linear equations (34)–(38) has a unique solution if and only ifΔ≠0.
From (36) we have C3,1=1−C1,1−C2,1. Substituting that into (34), (35) and (38) yields
∑i=12(λ¯μ¯zi,1−1(1−e(zi,1−1/μ¯)b1)+d1zi,1+λ¯−λ¯e−b1/μ¯)Ci,1+∑i=12λ¯e(zi,2−1/μ¯)b1μ¯zi,2−1Ci,2=−λ¯e−b1/μ¯,∑i=12(λe−b1/μμzi,1+1(1−e(zi,1+1/μ)b1)+λ(1−e−b1/μ))Ci,1+∑i=12(λ¯ezi,2b1μ¯zi,2−1+(d2zi,2+λ+λ¯)ezi,2b1)Ci,2=λ
and
∑i=12(ezi,1b1−1)Ci,1−∑i=12ezi,2b1Ci,2=−1.
Thus, the system of equations (34)–(38) has a unique solution if and only if the system of equations (39), (40), (37) and (41) does.
Multiplying (41) by (−d2z1,2), adding (37) and rearranging the terms we obtain
ez2,2b1C2,2=d2z1,2−∑i=12(d1zi,1ezi,1b1+d2z1,2(1−ezi,1b1))Ci,1d2(z1,2−z2,2).
Similarly, multiplying (41) by (−d2z2,2), adding (37) and rearranging the terms we obtain
ez1,2b1C1,2=d2z2,2−∑i=12(d1zi,1ezi,1b1+d2z2,2(1−ezi,1b1))Ci,1d2(z2,2−z1,2).
Substituting (42) and (43) into (39) and doing some simplifications yield
∑i=12(μ¯zi,1eb1/μ¯−ezi,1b1μ¯zi,1−1+d1zi,1eb1/μ¯−1+−d1μ¯zi,1ezi,1b1+d2(1−μ¯z1,2−μ¯z2,2)(1−ezi,1b1)d2(μ¯z1,2−1)(μ¯z2,2−1))Ci,1=−μ¯2z1,2z2,2(μ¯z1,2−1)(μ¯z2,2−1).
By Vieta’s theorem applied to (29) for j=2, we have
z1,2z2,2=λ¯μ¯−λμ−d2d2μμ¯,1−μ¯z1,2−μ¯z2,2=d2μ¯+μμ¯(λ+λ¯)d2μ
and
(μ¯z1,2−1)(μ¯z2,2−1)=λ¯μ¯(μ+μ¯)d2μ.
Substituting these equalities into (44) gives
∑i=12(μ¯zi,1eb1/μ¯−ezi,1b1μ¯zi,1−1+d1zi,1eb1/μ¯−1+−d1μzi,1ezi,1b1+(d2+μ(λ+λ¯))(1−ezi,1b1)λ¯(μ+μ¯))Ci,1=d2+λμ−λ¯μ¯λ¯(μ+μ¯).
Next, multiplying (41) by (λ+λ¯) and adding (37) and (40) we get
∑i=12(λe−b1/μμzi,1+1(1−e(zi,1+1/μ)b1)+λ(1−e−b1/μ)+d1zi,1ezi,1b1+(λ+λ¯)(ezi,1b1−1))Ci,1+∑i=12λ¯ezi,2b1μ¯zi,2−1Ci,2=−λ¯.
Multiplying (39) by (−eb1/μ¯) and adding (46) we obtain
∑i=12(−λμzi,1e−b1/μ−λezi,1b1μzi,1+1+−λ¯eb1/μ¯+λ¯ezi,1b1μ¯zi,1−1−(d1zi,1+λ¯)eb1/μ¯+(d1zi,1+λ+λ¯)ezi,1b1)Ci,1=λ−λ¯.
Thus, if the system of equations (45) and (47) has a unique solution, then C1,2 and C2,2 can be found from (43) and (42), respectively. Consequently, the system of equations (39), (40), (37) and (41) has a unique solution if and only if the system of equations (45) and (47) does.
A standard computation shows that the determinant of the system of equations (45) and (47) is equal to Δ defined above. In particular, here we use Vieta’s theorem applied to (29) for j=1. Therefore, the system of equations (34)–(38) has a unique solution if and only if Δ≠0. □
Explicit formulas for the expected discounted dividend payments until ruin
By (17), equation (9) for the expected discounted dividend payments can be written as
djv′(x)+(λ+λ¯+δ)v(x)=λμe−x/μ∫0xv(u)eu/μdu+λ¯μ¯ex/μ¯∫x∞v(u)e−u/μ¯du+dj
for all x∈[bj−1,bj] and 1≤j≤k.
The piecewise integro-differential equation (48) can also be reduced to a piecewise linear differential equation with constant coefficients.
Let the surplus process(Xtb(x))t≥0be defined by (2) under the above assumptions, and let claim and premium sizes be exponentially distributed with means μ andμ¯, respectively. Thenv(x)is a solution to the piecewise differential equationdjμμ¯v‴(x)+(dj(μ¯−μ)+μμ¯(λ+λ¯+δ))v″(x)+(μ¯(λ¯+δ)−μ(λ+δ)−dj)v′(x)−δv(x)=−djfor allx∈[bj−1,bj]and1≤j≤k.
The proof of the lemma is similar to the proof of Lemma 1.
For 1≤j≤k, let
D˜j=−18δdjμμ¯(dj(μ¯−μ)+μμ¯(λ+λ¯+δ))(μ¯(λ¯+δ)−μ(λ+δ)−dj)+4δ(dj(μ¯−μ)+μμ¯(λ+λ¯+δ))3+(dj(μ¯−μ)+μμ¯(λ+λ¯+δ))2(μ¯(λ¯+δ)−μ(λ+δ)−dj)2−4djμμ¯(μ¯(λ¯+δ)−μ(λ+δ)−dj)3−27(δdjμμ¯)2.
Let the surplus process(Xtb(x))t≥0be defined by (2) under the above assumptions, and let claim and premium sizes be exponentially distributed with means μ andμ¯, respectively. If the net profit condition (3) holds andmin1≤j≤kD˜j>00$]]>, then we havevj(x)=C˜1,jez˜1,jx+C˜2,jez˜2,jx+C˜3,jez˜3,jx+dj/δfor allx∈[bj−1,bj]and1≤j≤k, wherez˜1,j,z˜2,jandz˜3,jare distinct real roots of the cubic equationdjμμ¯z3+(dj(μ¯−μ)+μμ¯(λ+λ¯+δ))z2+(μ¯(λ¯+δ)−μ(λ+δ)−dj)z−δ=0,C˜3,k=0and all the other constantsC˜1,j,C˜2,jandC˜3,jare determined from the system of linear equations (52)–(55):λe−bj−1/μ∑l=1j−1(∑i=13C˜i,lμz˜i,l+1(e(z˜i,l+1/μ)bl−e(z˜i,l+1/μ)bl−1))+∑i=13(λ¯ebj−1/μ¯μ¯z˜i,j−1(e(z˜i,j−1/μ¯)bj−e(z˜i,j−1/μ¯)bj−1)−(djz˜i,j+λ+λ¯+δ)ez˜i,jbj−1)C˜i,j+λ¯ebj−1/μ¯∑l=j+1k(∑i=13C˜i,lμ¯z˜i,l−1(e(z˜i,l−1/μ¯)bl−e(z˜i,l−1/μ¯)bl−1))=dj(λ+λ¯)δ−λe−bj−1/μδ∑l=1j−1dl(ebl/μ−ebl−1/μ)+λ¯ebj−1/μ¯δ∑l=jkdl(e−bl/μ¯−e−bl−1/μ¯),1≤j≤k,C˜1,1+C˜2,1+C˜3,1=−d1/δ,dj∑i=13z˜i,jez˜i,jbjC˜i,j−dj+1∑i=13z˜i,j+1ez˜i,j+1bjC˜i,j+1=dj−dj+1,1≤j≤k−1,and∑i=13ez˜i,jbjC˜i,j−∑i=13ez˜i,j+1bjC˜i,j+1=dj+1−djδ,1≤j≤k−1,provided that its determinant is not equal to 0.
By Lemma 2 and the notation introduced in Section 2, we deduce that the function vj(x) is a solution to (49) on x∈[bj−1,bj] for each 1≤j≤k. In addition, it is easily seen that (51) is the corresponding characteristic equation and its discriminant coincides with the constant D˜j introduced above. The assumption min1≤j≤kD˜j>00$]]> guarantees that cubic equation (51) has three distinct real roots z˜1,j, z˜2,j and z˜3,j. Hence, the general solution to (49) is given by (50) with some constants C˜1,j, C˜2,j and C˜3,j.
By Vieta’s theorem, we conclude that (51) has either two or no negative roots for each 1≤j≤k. Since the net profit condition (3) holds, applying arguments similar to those in [34, p. 70] shows that limx→∞vk(x)=dk/δ. Consequently, if equation (51) for j=k had no negative roots, the function vk(x) would be constant, which is impossible. Therefore, equation (51) for j=k has two negative roots. We denote those negative roots by z˜1,k and z˜2,k, and let z˜3,k be the third root. Since z˜3,k>00$]]>, we conclude that C˜3,k=0.
To determine all the other constants C˜1,j, C˜2,j and C˜3,j, we need 3k−1 boundary conditions. The first k conditions are obtained by letting x=bj−1 in (48) for 1≤j≤k:
djv′(bj−1)+(λ+λ¯+δ)v(bj−1)=λμe−bj−1/μ∫0bj−1v(u)eu/μdu+λ¯μ¯ebj−1/μ¯∫bj−1∞v(u)e−u/μ¯du+dj.
One more condition is obtained from the equality v(0)=0. The last 2(k−1) conditions are derived from the equalities djvj′(bj)−dj=dj+1vj+1′(bj)−dj+1, which easily follow from (48), and vj(bj)=vj+1(bj) for 1≤j≤k−1.
Substituting (50) into (56) as well as into the equalities v(0)=0, djvj′(bj)−dj=dj+1vj+1′(bj)−dj+1 and vj(bj)=vj+1(bj) for 1≤j≤k−1 and doing some simplifications yield the system of linear equations (52)–(55), which has a unique solution provided that its determinant is not equal to 0. Thus, piecewise differential equation (49) has a unique solution satisfying certain conditions, and that solution is given by (50). Applying arguments similar to those in the proof of Theorem 3 guaranties that the functions vj(x) we have found coincide with the expected discounted dividend payments until ruin on [bj−1,bj], which completes the proof. □
In particular, if k=2, then C˜3,2=0 and the constants C˜1,1, C˜2,1, C˜3,1, C˜1,2 and C˜2,2 are determined from the system of linear equations (57)–(61):
∑i=13(λ¯μ¯z˜i,1−1(1−e(z˜i,1−1/μ¯)b1)+d1z˜i,1+λ+λ¯+δ)C˜i,1+∑i=12λ¯e(z˜i,2−1/μ¯)b1μ¯z˜i,2−1C˜i,2=−d1λδ−λ¯(d1−d2)e−b1/μ¯δ,λe−b1/μ∑i=13C˜i,1μz˜i,1+1(1−e(z˜i,1+1/μ)b1)+∑i=12(λ¯ez˜i,2b1μ¯z˜i,2−1+(d2z˜i,2+λ+λ¯+δ)ez˜i,2b1)C˜i,2=λ(d1−d2)δ−d1λe−b1/μδ,C˜1,1+C˜2,1+C˜3,1=−d1/δ,d1∑i=13z˜i,1ez˜i,1b1C˜i,1−d2∑i=12z˜i,2ez˜i,2b1C˜i,2=d1−d2
and
∑i=13ez˜i,1b1C˜i,1−∑i=12ez˜i,2b1C˜i,2=d2−d1δ
provided that its determinant is not equal to 0.
Numerical illustrations
To present numerical examples for the results obtained in Section 5, we set λ=0.1, λ¯=2.3, μ=3, μ¯=0.2, b=5 and δ=0.01.
In addition, we denote by ψ∗(x) the ruin probability in the corresponding model without dividend payments. It is given by
ψ∗(x)=λ(μ+μ¯)μ¯(λ+λ¯)exp(−(λ¯μ¯−λμ)xμμ¯(λ+λ¯)),x∈[0,∞),
(see [5, 28]). For the parameters chosen above, ψ∗(x)≈0.666667e−0.111111x.
Moreover, let now d1=0.05 and d2=0.1. Applying Theorems 3 and 4 as well as Remarks 6 and 7 we can calculate the ruin probability ψ(x) and the expected discounted dividend payments until ruin v(x):
ψ1(x)≈0.530821e−0.084781x+0.179668e−43.248552x+0.289512,x∈[0,5],ψ2(x)≈0.826718e−0.051863x−7.043723·1038e−19.28147x,x∈[5,∞);v1(x)≈5−2.992137e−43.470279x−4.421273e−0.124597x+2.41341e0.061543x,x∈[0,5],v2(x)≈10−2.198169·1040e−19.405407x−6.97712e−0.107684x,x∈[5,∞).
Table 1 presents the results of calculations for some values of x.
The ruin probabilities without and with dividend payments and the expected discounted dividend payments, d1=0.05 and d2=0.1
x
ψ∗(x)
ψ(x)
v(x)
0
0.666667
1
0
1
0.596560
0.777184
3.663273
2
0.533825
0.737542
4.283457
5
0.382502
0.636926
5.911685
7
0.306284
0.575029
6.716708
10
0.219462
0.492173
7.623108
15
0.125917
0.379750
8.612682
20
0.072245
0.293007
9.190265
50
0.002577
0.061825
9.967986
70
0.000279
0.021912
9.996285
Next, for d1=0.1 and d2=0.05, we get
ψ1(x)≈1.204304e−0.051863x+0.218067e−19.28147x−0.42237,x∈[0,5],ψ2(x)≈0.772527e−0.084781x+1.012903·1091e−43.248552x,x∈[5,∞);v1(x)≈10−2.219094e−19.405407x−5.609737e−0.107684x−2.171169e0.079758x,x∈[0,5],v2(x)≈5+5.716149·1092e−43.470279x−2.857069e−0.124597x,x∈[5,∞).
The values of ψ∗(x), ψ(x) and v(x) for some x are given in Table 2.
The ruin probabilities without and with dividend payments and the expected discounted dividend payments, d1=0.1 and d2=0.05
x
ψ∗(x)
ψ(x)
v(x)
0
0.666667
1
0
1
0.596560
0.721066
2.611525
2
0.533825
0.663275
2.930525
5
0.382502
0.506845
3.490686
7
0.306284
0.426750
3.805635
10
0.219462
0.330912
4.178134
15
0.125917
0.216577
4.559200
20
0.072245
0.141747
4.763582
50
0.002577
0.011141
4.994372
70
0.000279
0.002044
4.999534
The results presented in Tables 1 and 2 indicate that dividend payments substantially increase the ruin probability. The first strategy is much more profitable, although the corresponding ruin probability is larger in that case.
Acknowledgement
The author is deeply grateful to the anonymous referees for careful reading and valuable comments and suggestions, which helped to improve the earlier version of the paper.
ReferencesAlbrecher, H., Hartinger, J.: A risk model with multilayer dividend strategy. 11, 43–64 (2007). MR2380719. https://doi.org/10.1080/10920277.2007.10597447Asmussen, S., Albrecher, H.: . World Scientific, Singapore (2010). MR2766220. https://doi.org/10.1142/9789814282536Badescu, A., Landriault, D.: Recursive calculation of the dividend moments in a multi-threshold risk model. 12, 74–88 (2008). MR2485710. https://doi.org/10.1080/10920277.2008.10597501Badescu, A., Drekic, S., Landriault, D.: On the analysis of a multi-threshold Markovian risk model. 2007, 248–260 (2007). MR2416548. https://doi.org/10.1080/03461230701554080Boikov, A.V.: The Cramér–Lundberg model with stochastic premium process. 47, 489–493 (2003). MR1975908. https://doi.org/10.1137/S0040585X9797987Boudreault, M., Cossette, H., Landriault, D., Marceau, E.: On a risk model with dependence between interclaim arrivals and claim sizes. 2006, 265–285 (2006). MR2328676. https://doi.org/10.1080/03461230600992266Chadjiconstantinidis, S., Vrontos, S.: On a renewal risk process with dependence under a Farlie–Gumbel–Morgenstern copula. 2014, 125–158 (2014). MR3177095. https://doi.org/10.1080/03461238.2012.663730Cheng, Y., Tang, Q.: Moments of the surplus before ruin and the deficit of ruin in the Erlang(2) risk process. 7, 1–12 (2003). MR1986883. https://doi.org/10.1080/10920277.2003.10596073Chi, Y., Lin, X.S.: On the threshold dividend strategy for a generalized jump-diffusion risk model. 48, 326–337 (2011). MR2820045. https://doi.org/10.1016/j.insmatheco.2010.11.006Cossette, H., Marceau, E., Marri, F.: On the compound Poisson risk model with dependence based on a generalized Farlie–Gumbel–Morgenstern copula. 43, 444–455 (2008). MR2479592. https://doi.org/10.1016/j.insmatheco.2008.08.009Cossette, H., Marceau, E., Marri, F.: Analysis of ruin measures for the classical compound Poisson risk model with dependence. 2010, 221–245 (2010). MR2732285. https://doi.org/10.1080/03461230903211992Cossette, H., Marceau, E., Marri, F.: Constant dividend barrier in a risk model with a generalized Farlie–Gumbel–Morgenstern copula. 13, 487–510 (2011). MR2822392. https://doi.org/10.1007/s11009-010-9168-9Cossette, H., Marceau, E., Marri, F.: On a compound Poisson risk model with dependence and in the presence of a constant dividend barrier. 30, 82–98 (2014). MR3191344. https://doi.org/10.1002/asmb.1928De Finetti, B.: Su un’impostazione alternativa dell teoria colletiva del rischio. 2, 433–443 (1957)Deng, C., Zhou, J., Deng, Y.: The Gerber–Shiu discounted penalty function in a delayed renewal risk model with multi-layer dividend strategy. 82, 1648–1656 (2012). MR2951000. https://doi.org/10.1016/j.spl.2012.05.002Gerber, H.U., Shiu, E.S.W.: On the time value of ruin. 2, 48–72 (1998). MR1988433. https://doi.org/10.1080/10920277.1998.10595671Gerber, H.U., Shiu, E.S.W.: The time value of ruin in a Sparre Andersen model. 9, 49–69 (2005). MR2160108. https://doi.org/10.1080/10920277.2005.10596197Heilpern, S.: Ruin measures for a compound Poisson risk model with dependence based on the Spearman copula and the exponential claim sizes. 59, 251–257 (2014). MR3283226. https://doi.org/10.1016/j.insmatheco.2014.10.006Jiang, W., Yang, Z., Li, X.: The discounted penalty function with multi-layer dividend strategy in the phase-type risk model. 82, 1358–1366 (2012). MR2929787. https://doi.org/10.1016/j.spl.2012.03.012Landriault, D.: Constant dividend barrier in a risk model with interclaim-dependent claim sizes. 42, 31–38 (2008). MR2392066. https://doi.org/10.1016/j.insmatheco.2006.12.002Li, B., Ma, S.: Markov-dependent risk model with multi-layer dividend strategy and investment interest under absolute ruin. 6, 260–268 (2016)Li, B., Wu, R., Song, M.: A renewal jump-diffusion process with threshold dividend strategy. 228, 41–55 (2009). MR2517685. https://doi.org/10.1016/j.cam.2008.08.046Li, S., Garrido, J.: On a class of renewal risk models with a constant dividend barrier. 35, 691–701 (2004). MR2106143. https://doi.org/10.1016/j.insmatheco.2004.08.004Lin, X.S., Pavlova, K.P.: The compound Poisson risk model with a threshold dividend strategy. 38, 57–80 (2006). MR2197303. https://doi.org/10.1016/j.insmatheco.2005.08.001Lin, X.S., Sendova, K.P.: The compound Poisson risk model with multiple thresholds. 42, 617–627 (2008). MR2404318. https://doi.org/10.1016/j.insmatheco.2007.06.008Lin, X.S., Willmot, G.E., Drekic, S.: The classical risk model with a constant dividend barrier: analysis of the Gerber-Shiu discounted penalty function. 33, 551–566 (2003). MR2021233. https://doi.org/10.1016/j.insmatheco.2003.08.004Liu, D., Liu, Z., Peng, D.: The Gerber–Shiu expected penalty function for the risk model with dependence and a constant dividend barrier. 2014, 730174, 1–7 (2014). MR3246356. https://doi.org/10.1155/2014/730174Mishura, Y., Ragulina, O.: . ISTE Press – Elsevier, London (2016). MR3643478Mishura, Y., Ragulina, O., Stroev, O.: Practical approaches to the estimation of the ruin probability in a risk model with additional funds. 1, 167–180 (2014). MR3316485. https://doi.org/10.15559/15-VMSTA18Mishura, Y.S., Ragulina, O.Y., Stroev, O.M.: Analytic property of infinite-horizon survival probability in a risk model with additional funds. 91, 131–143 (2015)Mitric, I.-R., Sendova, K.P., Tsai, C.C.-L.: On a multi-threshold compound Poisson process perturbed by diffusion. 80, 366–375 (2010). MR2593575. https://doi.org/10.1016/j.spl.2009.11.012Ragulina, O.: The risk model with stochastic premiums, dependence and a threshold dividend strategy. 4, 315–351 (2017). MR3739013. https://doi.org/10.15559/17-vmsta89Rolski, T., Schmidli, H., Schmidt, V., Teugels, J.: . John Wiley & Sons, Chichester (1999). MR1680267. https://doi.org/10.1002/9780470317044Schmidli, H.: . Springer, London (2008). MR2371646Shi, Y., Liu, P., Zhang, C.: On the compound Poisson risk model with dependence and a threshold dividend strategy. 83, 1998–2006 (2013). MR3079035. https://doi.org/10.1016/j.spl.2013.05.008Sun, L.-J.: The expected discounted penalty at ruin in the Erlang(2) risk process. 72, 205–217 (2005). MR2137163. https://doi.org/10.1016/j.spl.2004.12.015Wang, W.: The perturbed Sparre Andersen model with interest and a threshold dividend strategy. 17, 251–283 (2015). MR3343407. https://doi.org/10.1007/s11009-013-9332-0Xie, J.-H., Zou, W.: On the expected discounted penalty function for a risk model with dependence under a multi-layer dividend strategy. 46, 1898–1915 (2017). MR3574735. https://doi.org/10.1080/03610926.2015.1030424Xing, Y., Wu, R.: Moments of the time of ruin, surplus before ruin and the deficit at ruin in the Erlang(N) risk process. 22, 599–606 (2006). MR2248523. https://doi.org/10.1007/s10255-006-0333-4Yang, H., Zhang, Z.: Gerber–Shiu discounted penalty function in a Sparre Andersen model with multi-layer dividend strategy. 42, 984–991 (2008). MR2435369. https://doi.org/10.1016/j.insmatheco.2007.11.004Yang, H., Zhang, Z.: On a perturbed Sparre Andersen risk model with multi-layer dividend strategy. 232, 612–624 (2009). MR2555426. https://doi.org/10.1016/j.cam.2009.06.032Yang, H., Zhang, Z.: The perturbed compound Poisson risk model with multi-layer dividend strategy. 79, 70–78 (2009). MR2655109. https://doi.org/10.1016/j.spl.2008.07.017Yang, H., Zhang, Z., Lan, C.: On the time value of absolute ruin for a multi-layer compound Poisson model under interest force. 78, 1835–1845 (2008). MR2453921. https://doi.org/10.1016/j.spl.2008.01.038Yin, J.: On a dual model with multi-layer dividend strategy under stochastic interest. 422, 775–778 (2012)Yong, W., Xiang, H.: Differential equations for ruin probability in a special risk model with FGM copula for the claim size and the inter-claim time. 2012, 156, 1–13 (2012). MR3085740. https://doi.org/10.1186/1029-242X-2012-156Zhang, Z., Yang, H.: The compound Poisson risk model with dependence under a multi-layer dividend strategy. 26, 1–13 (2011). MR2776860. https://doi.org/10.1007/s11766-011-2279-4Zhang, Z., Yang, H.: Gerber–Shiu analysis in a perturbed risk model with dependence between claim sizes and interclaim times. 235, 1189–1204 (2011). MR2728058. https://doi.org/10.1016/j.cam.2010.08.003Zhou, Z., Xiao, H., Deng, Y.: Markov-dependent risk model with multi-layer dividend strategy. 252, 273–286 (2015). MR3305106. https://doi.org/10.1016/j.amc.2014.12.016