Optimal control strategies for the premium policy of an insurance firm with jump diffusion assets and stochastic interest rate

Guerdouh, Dalila; Khelfallah, Nabil; Vives, Josep Article Optimal control strategies for the premium policy of an insurance firm with jump diffusion assets and stochastic interest rate Journal of Risk and Financial Management Provided in Cooperation with: MDPI – Multidisciplinary Digital Publishing Institute, Basel Suggested Citation: Guerdouh, Dalila; Khelfallah, Nabil; Vives, Josep (2022) : Optimal control strategies for the premium policy of an insurance firm with jump diffusion assets and stochastic interest rate, Journal of Risk and Financial Management, ISSN 1911-8074, MDPI, Basel, Vol. 15, Iss. 3, pp. 1-19, https://doi.org/10.3390/jrfm15030143 This Version is available at: https://hdl.handle.net/10419/258866 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/   Citation: Guerdouh, Dalila, Nabil Khelfallah, and Josep Vives. 2022. Optimal Control Strategies for the Premium Policy of an Insurance Firm with Jump Diffusion Assets and Stochastic Interest Rate. Journal of Risk and Financial Management 15: 143. https://doi.org/10.3390/jrfm15030143 Academic Editor: Svetlozar Rachev Received: 17 November 2021 Accepted: 14 March 2022 Published: 17 March 2022 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Copyright: © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). Journal of Risk and Financial Management Article Optimal Control Strategies for the Premium Policy of an Insurance Firm with Jump Diffusion Assets and Stochastic Interest Rate Dalila Guerdouh 1, Nabil Khelfallah 1,* and Josep Vives 2 1Laboratoire de Mathématiques Appliquées, Université de Biskra, PB 145, Biskra 07000, Algeria; [email protected] 2 Departament de Matemàtiques i Informàtica, Universitat de Barcelona, Gran Via 585, 08007 Barcelona, Spain; [email protected] *Correspondence: [email protected] Abstract: In this paper, we present a stochastic optimal control model to optimize an insurance firm problem in the case where its cash-balance process is assumed to be described by a stochastic differential equation driven by Teugels martingales. Noticing that the insurance firm is able to control its cash-balance dynamics by regulating the underlying premium rate, the aim of the policy maker is to select an appropriate premium in order to minimize the total deviation of the state process to some pre-set target level. As a part of stochastic maximum principle approach, a verification theorem is used to fulfill this achievement. Keywords: forward-backward stochastic differential equations; teugels martingales; lévy processes; optimal premium policies 1. Introduction An insurance is a contract, represented by a policy, used as a method of protection against losses, whether big or small. This means that the insured can receive or reimburse some financial amount to offset his or her losses from an insurance company. Due the fact that an insurance premium is the amount of money that a person or a company ought to pay for an insurance policy, one can perceive it in two different ways. In one side, it can be considered as an income by the insurance company. On the other side, it can also be considered as a liability in that the insurer must provide coverage for claims being made against the policy. In the present paper, we focus on the first case, where the policy maker looks forward to maximize the terminal wealth of its firm’s cash-balance under a demand law. The main problem in optimal control theory is to characterize an optimal control process. There are two main approaches, the Pontryagin’s maximum principle and the Bellman’s dynamic programming. We note that optimal control theory has been used for example in Hipp and Vogt (2003) in order to determine an optimal dynamic unlimited excess of loss reinsurance strategy to minimize the infinite time ruin probability, see also Højgaard and Taksar (1998), where the authors applied a proportional reinsurance policy for diffusion models in order to find a policy that maximizes a given return function before the time of ruin. In Cairns (2000), Cairns studied the optimization problem of stochastic pension fund models in continuous time. The mean-variance portfolio management for an insurance company was studied by Josa-Fombellida and Rincón-Zapatero (2008) by using dynamic programming techniques and also by Xie et al. (2008) using the general stochastic linear quadratic control technique. In Moore and Young (2006), Moore and Young obtained optimal dynamic consumption, investment, and insurance strategies, using the dynamic programming principle and a Markov chain approximation method. Ngwira and Gerrad J. Risk Financial Manag. 2022,15, 143. https://doi.org/10.3390/jrfm15030143 https://www.mdpi.com/journal/jrfm J. Risk Financial Manag. 2022,15, 143 2 of 19 in Ngwira and Gerrard (2007) showed that the optimal contribution and asset allocation policies have similar forms as in the pure diffusion case, but with a modification due to the effect of jumps. In Huang et al. (2010), Huang et al. explicitly derived the insurance company’s optimal premium strategy and the associated optimal cost function. We refer the reader to a list of recent papers and the references therein for more details in this subject, Asmussen et al. (2019); Supian and Mamat (2021); Li and Young (2021); Lin et al. (2020). Motivated by the above results, in this paper, we solve an optimal premium policy problem of an insurance firm. The main tool used in proving our main results is the Pontryagin maximum principle. More precisely, the sufficient condition of optimality. Noting that, in pretty much all of the previous papers, the authors dealt with the problem of optimal insurance in continuous-time models, we impose here to work with a quite general semi-martingale framework assuming that the liability process is driven by both a Brownian motion and a family of pairwise orthogonal martingales associated with a Lévy process. This kind of models comes naturally from the fact that in many real cases, the continuity of the trajectories condition cannot be satisfied. Indeed, the empirical distribution of the cash balance process tends to deviate from normal distributions, either due to inspected dusters or huge profits, many successive incidents or even because of the lack of continuity in the real world of applications. The paper is organized as follows. In Section 2, we introduce the preliminaries and formulate our control problem. In Section 3, we derive a sufficient condition of optimality for stochastic differential equations driven by a Brownian motion and a family of Teugels martingales. In Section 4, we apply our result to study problems of optimal premium rate of an insurance firm with jump diffusion assets with stochastic interest rate. 2. Preliminaries and Problem Statement 2.1. Preliminaries Let T> 0 and (Ω,F,F,P) be a complete filtered probability space supporting a onedimensional standard Brownian motion W and a one-dimensional Lévy process L with triplet (γ , σ2 , ν) , defined on [ 0, T] , independent each other. Recall that γ∈R , σ2> 0 and ν are Lévy measures. We denote by E the expectation with respect to P . Moreover, F:={Ft , t∈[ 0, T]} denotes the complete right-continuous natural filtration generated by processes Wand L. Recall briefly the L2 theory of Lévy processes as it is presented in Nualart and Schoutens (2000). Assume there exists α>0 such that for every ε>0, Z(−ε,ε)ceα|z|ν(dz)<∞. Note that this guarantee that νhas moments of all orders. Denote by ∆L(t):=L(t)−L(t−) the jumps of the Lévy process and define the power jump processes L(1)(t):=L(t)and L(i)(t):=∑ 0<s≤t (∆L(s))i,i≥2. If we define Y(i)(t):=L(i)(t)−EhL(i)(t)i,i≥1, then the family of Teugels martingales {H(i)(t)}∞ i=1is defined by H(i)(t):= i ∑ j=1 aijY(j)(t) where coefficients aij correspond to the orthonormalization of polynomials 1, x , x2 , . . . with respect to measure µ(dx)=x2ν(dx)+σ2δ0(dx) . Recall that {H(i)(t)}∞ i=1 is a family J. Risk Financial Manag. 2022,15, 143 3 of 19 of strongly orthogonal martingales such that for any i and j , DH(i),H(j)Et=δijt and hH(i),H(j)i−DH(i),H(j)Eis a martingale. Throughout this article, we will use the following spaces: •l2(R): the Hilbert space of real-valued sequences x=(xn)n≥0with norm ∞ ∑ i=1 x2 i!1 2 <∞. •P2(R): the Hilbert space of l2(R)-valued functions fii≥0such that ∞ ∑ i=1ZT 0|fi(t)|2dt!1 2 <∞. •l2 F(0, T;R): the Hilbert space of P2(R) -valued F -predictable processes fi(t)i≥0 such that EZT 0 ∞ ∑ i=1 |fi(t)|2dt!1 2 <∞. •L2 F(0, T;R): the Hilbert space of real-valued F-adapted processes f(t)such that EZT 0|f(t)|2dt1 2 <∞. •S2 F(0, T;R): the Banach space of real-valued, F -adapted and càdlàg processes f(t) such that Esup 0≤t≤T |f(t)|2!1 2 <∞. •L2(Ω,F,P,R): the Hilbert space of real-valued square integrable random variables defined on (Ω,F,P). 2.2. Description of the Model In this paper, we handle a cash management problem of an insurance firm (also called cash-balance problem). The insurer aims to manage the firm’s operating cash in order to meet demand. We consider that the insurance firm can constantly modify its premium policy while investing in certain financial accounts. Its concern is to pay due benefits, but at the same time to stabilize the insurance schemes by avoiding a large deviation on the cash balance. We address an optimal premium policy problem of an insurance firm under stochastic interest rate when the liability process, also called the payment function, B(t) , is modeled by the stochastic differential equation: −dB(t)=(b(t)+v(t))dt +a(t)dW(t)+ ∞ ∑ i=1 πi(t)dH(i)(t),B(0) = 0, where b is a locally bounded function that denotes the liability rate, that is, the expected liability per unit time due to premium loading, v is the premium rate (premium policy) and a and πi∞ i=1 are locally bounded functions that denote the volatility rates measuring the liability risks belonging respectively to the Brownian and Teugels martingale components. For simplicity, from now on, we will write π(t):=πi(t)∞ i=1 , H(t):=nH(i)(t)o∞ i=1 and J. Risk Financial Manag. 2022,15, 143 4 of 19 π(t)dH(t):= ∞ ∑ i=1 πi(t)dH(i)(t). Assume moreover that the cash balance process of the insurer X(·) is described by the formula X(t)=e∆(t)X(0)−Zt 0e−∆(s)dB(s), (1) where X(0)=x≥ 0 represents the initial reserve and ∆(·) represents the stochastic interest rate, that follows the dynamics d∆(t)=δ(t)dt +α(t)dW(t),t∈[0, T],∆(0)=0, (2) with δand αlocally bounded functions. Note that X(t) is the difference between the initial capital and the net expenses up to time t. Now, the Itô formula applied to the process X(·) leads to the following controlled Stochastic Differential Equation (SDE): dX(t)=f1(t,X(t),v(t))dt +σ1(t,X(t),v(t))dW(t)+π(t)dH(t), X(0)=x,(3) where f1(t,X(t),v(t)) =δ(t)+1 2α2(t)X(t)+α(t)a(t)+b(t)+v(t), and σ1(t,X(t),v(t)) =X(t)α(t)+a(t). Herein, the process v(·) stands for the control variable. We require that process v(·) is adapted, with càdlàg trajectories and taking values in U , a non-empty convex subset of R , such that the fourth-power condition EZT 0|v(t)|4dt <∞ holds. Furthermore, we assume there exists a positive constant c0 such that the SDE (3) has a unique solution satisfying the terminal constraint E[X(T)] =c0. (4) This last equality means that the insurance firm is looking for some regulatory requirement c0 described by the average value of its cash balance process at the terminal time T. Herein, a control variable is said to be an admissible control if and only if it satisfies all the above conditions. Let us point out that the aim of the policy maker is to simultaneously minimize the deviation between the firm’s cash-balance process and its dynamic benchmark over the set of all admissible controls, which will be denoted by U , the cost of the premium policy over the whole time interval [0, T] , and the terminal variance of the cash-balance process under some given constraint. Therefore, it is quite natural that the cost functional takes the following form: J(v(·)) =EZT 0e−βtg1(t,X(t),v(t))dt +e−βTϕ1(X(T)), (5) with J. Risk Financial Manag. 2022,15, 143 5 of 19 g1(t,x,v)=1 2R(t)(x−A(t))2+N(t)v2(t), (6) and ϕ1(x)=1 2M(x−c0)2. (7) Here, β≥ 0 is a discounting factor, A(t) is some dynamic pre-set target, representing the dynamic benchmark of X , processes R(·) , N(·) and the positive constant M are the weighting factors which make the cost functional (5) more general and flexible to control the preference of the policy-maker. Furthermore, we suppose that process A converges to c0as tgoes to T. Now, we can formulate the firm’s optimal premium problem as Problem A: To find ˆ v∈ U such that ˆ v minimizes the cost function (5) subject to (3) and the state constraint (4). To deal with this problem, we need to impose the following assumption on the previous coefficients: Assumption 1. Functions R(·)≥ 0, N(·)> 0, N−1(·) and A(·) are all deterministic and bounded on the time interval [0, T]. Let us now reformulate the above control problem (3)–(5) as a generalized stochastic recursive optimal control problem with state constraint by introducing the following backward stochastic differential equation (BSDE), in term of processes Y(·) , Z(·) and K(·) = {Ki(·)}∞ i=1: −dY(t)=[g1(t,X(t),v(t)) −βY(t)]dt −Z(t)dW(t)−K(t)dH(t), Y(T)=ϕ1(X(T)).(8) Notice that (3) together with (8) form a semi-coupled Forward-Backward Stochastic Differential Equation (FBSDE) driven by both the Teugels martingales and an independent Brownian motion. Lemma 2.1 in Meng and Tang (2009) shows that under (H1), (3) admits a unique solution X(t)for each v(·)∈ U. Moreover, we have sup 0≤t≤T EX4(t)<+∞. The proof of this estimate is placed in Lemma 1 Section 3. As a consequence, the terminal condition of (8) is square-integrable. Then, for the foregoing v(t) and X(t) , thanks to Theorem 3.1 in Bahlali et al. (2003), the Backward Stochastic Differential Equation (BSDE) (8) admits a unique solution (Y(·),Z(·),K(·)) under (H1). That is, for any v(·)∈ U , the semi-coupled FBSDE consisting of (3) and (8) admits a unique solution: (X(·),Y(·),Z(·),K(·)). Obviously, by using the dual technique to the BSDE (8) one can get Y(0)=J(v(·)) . Then, we can reformulate Problem A in the following way: Problem B: To find ˆ v∈ U such that J(ˆ v(·)) =E(Y(0)) (9) subject to (3),(4)and (8). J. Risk Financial Manag. 2022,15, 143 6 of 19 In the next section, we are going to prove a sufficient condition of optimality for the above problem, but in a more general form by assuming that coefficients are not necessarily linear with respect to the state variables. 3. Sufficient Condition of Optimality 3.1. Problem Formulation Motivated by the above optimal premium policy of an insurance firm, we are now going to focus on the following control problem where the state process is described by the following controlled stochastic differential equation driven by both a Brownian motion and a family of Teugels martingales,        dX(t)=f(t,X(t),v(t))dt +σ(t,X(t),v(t))dW(t) + ∞ ∑ i=1 πi(t,X(t−),v(t))dH(i)(t), X(0)=x, (10) and the general stochastic differential utility is given by the BSDE    −dY(t)=g(t,X(t),v(t))dt −Z(t)dW(t)−∞ ∑ i=1 Ki(t)dH(i)(t), Y(T)=ϕ(X(T)). (11) where X(t):=(X(t),Y(t),Z(t),K(t)), and f:[0, T]×Ω×R×U→R, σ:[0, T]×Ω×R×U→R, π:[0, T]×Ω×R×U→ P2(R), g:[0, T]×Ω×R×R×R× P2(R)×U→R, ϕ:[0, T]×Ω×R→R, are progressively measurable functions. In the sequel, for notational simplicity, we shall use the shorthand notation π(t,X(t−),v(t))dH(t)and K(t)dH(t) instead of ∞ ∑ i=1 πi(t,X(t−),v(t))dH(i)(t)and ∞ ∑ i=1 Ki(t)dH(i)(t) respectively, where K(t)=Ki(t)∞ i=1and π(t)=πi(t)∞ i=1. Let us now introduce the following basic assumption on coefficients that will be needed in the sequel: Assumption 2. (i) The function g is F− progressively measurable for all (y,z,k)∈R×R× P2(R) and for any v(·)∈ U, g(t,X(t),v(t)) =G(t,Y(t),Z(t),K(t)) +R(t)(X(t)−A(t))2+N(t)v2(t), with E  T Z 0 |g(s, 0, 0, 0, 0, v(t))|2ds <∞. (ii) For every x ∈R,ϕ∈L2,f,σand πare F−progressively measurable and for any v(·)∈ U, J. Risk Financial Manag. 2022,15, 143 7 of 19 E  T Z 0|f(s, 0, v(t))|4+|σ(s, 0, v(t))|4+kπ(s, 0, v(t))k4 P2(R)ds <∞. (iii) Functions G , f , σ and π are continuously differentiable with bounded derivatives with respect to x,y,z,k and v. Note that Lemma 2.1 in Meng and Tang (2009) shows that under the Assumption (H2) the SDE (10) admits a unique solution that belongs to S2 F(0, T,R) . On the other hand, since the function g is uniformly Lipschitz with respect to y , z and k , by using the assumptions (H2), one can easily check that the BSDE (11) satisfies all the conditions in Theorem 3.1 in Bahlali et al. (2003), and hence, it has a unique solution that belongs to S2 F(0, T,R)× L2 F(0, T,R)×l2 F(0, T,R). The following key lemma gives an L4 estimate of the solution of the state equation under Assumption (H2). Since Assumption (H1) implies Assumption (H2), the result is still valid under Assumption (H1). Lemma 1. Assume that X(·) is a solution of the state Equation (10) .Then, under Assumption (H2), the following estimate holds true: sup 0≤t≤T EX4(t)<+∞. (12) Proof. First, we introduce the stopping times Rk=inf{t:|X(t)|≥k}, assuming k great enough to include X( 0 ) = x in [−k , k] . Then, applying Itô’s formula to |X(t∧Rk)|4we obtain X4(t∧Rk)=X4(0) +4Zt∧Rk 0X3(s)f(s,X(s),v(s))ds +4Zt∧Rk 0X3(s)σ(s,X(s),v(s))dW(s) +4Zt∧Rk 0X3(s−)π(s,X(s−),v(s−))dH(s) +6Zt∧Rk 0X2(s)σ2(s,X(s),v(s))ds +6Zt∧Rk 0X(s)2||π(s,X(s),v(s))||2ds +6 ∞ ∑ i,j=1Zt∧Rk 0X2(s−)(πiπj)(s,X(s−),v(s−))dMi,j(t) +I(t∧Rk) where I(t):=∑ 0<s≤tnX4(s)−X4(s−)−4X3(s−)∆X(s)−12X2(s−)(∆X(s))2o. The stochastic integrals with respect to W , H and Mi,j in the right hand side of the above equality are in fact martingales with zero expectation. Then, taking expectations and later absolute values, one finds J. Risk Financial Manag. 2022,15, 143 8 of 19 E|X(t∧Rk)|4≤E|X(0)|4+4ERt∧Rk 0|X(s)|3|f(s,X(s),v(s))|ds +6ERt∧Rk 0|X(s)|2|σ(s,X(s),v(s))|2ds +6ERt∧Rk 0|X(s)|2||π(s,X(s),v(s))||2ds+E(|I(t∧Rk)|). In view of Young’s inequality, one can easily check that E|X(t∧Rk)|4≤E|X(0)|4+9ERt∧Rk 0|X(s)|4ds +E(|I(t∧Rk)|) +ERt∧Rk 0|f(s,X(s),v(s))|4ds +3ERt∧Rk 0|σ(s,X(s),v(s))|4ds +3ERt∧Rk 0||π(s,X(s),v(s))||4ds. From now on, Cwill denote a constant that changes from line to line. Using Lipschitz condition in (H2), we have E|X(t∧Rk)|4≤E|X(0)|4+E|I(t∧Rk)|+CERt∧Rk 0|X(s)|4ds +CERt∧Rk 0|f(s, 0, v(s))|4+|σ(s, 0, v(s))|4+||π(s, 0, v(s))||4ds. We infer, using Gronwall’s inequality, that E|X(t∧Rk)|4≤eC(t∧Rk)E|X(0)|4+E|I(t∧Rk)|+CERt∧Rk 0|f(s, 0, v(s))|4+ |σ(s, 0, v(s))|4+||π(s, 0, v(s))||4ds. This implies, using Fatou’s lemma and taking the supremum over the interval [0, T] , that sup 0≤t≤T E|X(t)|4≤CE|X(0)|4+E|I(T)|+ERT 0|f(t, 0, v(t))|4+ |σ(t, 0, v(t))|4+||π(t, 0, v(t))||4ds, (13) Now, we turn out to estimate E(|I(T)|) . For each k , it is clear that the stopped process ˜ X=X 1 [0,Rk) is bounded by k , and it is a semi-martingale as a product of two semimartingales. Note that the estimate ∑ 0<s≤tn˜ X4(s)−˜ X4(s−)−4˜ X3(s)∆˜ X(s)−12 ˜ X2(s)∆˜ X(s)2o ≤C˜ X,˜ Xt holds true P−a . s . This is a consequence of the fact that ˜ X(·) takes values in the compact interval [−k,k]and the fact that for h(x)=x4, it is easy to see that h(x)−h(y)−(y−x)h0(x)−(y−x)2h00 (x)≤C(y−x)2. Thus, ∑ 0<s≤T |˜ X4(s)−˜ X4(s−)−4˜ X3(s)∆X(s)−12 ˜ X2(s)∆˜ X(s)2| ≤C∑ 0<s≤T∆˜ Xs2≤C˜ X,˜ XT<∞,P−a.s. (14) Since the last inequality is valid for ˜ X , for each k , it also remains valid for X . Combining (14)with (13), we get the desired result and this completes the proof. 3.2. Verification Theorem In this subsection, we study Problem B with a more general state process. We establish a sufficient stochastic maximum principle for stochastic control of forward-backward SDEs J. Risk Financial Manag. 2022,15, 143 15 of 19 and          dY(t)=1 2c1(t)X2(t)+c2(t)X(t)+c3(t)v2(t)+c4(t)−βY(t)dt −Z(t)dW(t)−K(t)dH(t) Y(T)=1 2(X(T)−a)2 (37) where a is real constant and a1 , a2 , a3 , b1 , b26= 0, b3 , c1 , c2 , c3 , c4 are deterministic functions satisfying some properties. Note that particular choices of the coefficients of (36) and (37) generate the SDE (10)and the BSDE (11). Here, the Hamiltonian (17)gets the form H=1 2c1(t)X2(t)+c2(t)X(t)+c3(t)v2(t)+c4(t)−βY(t)λ(t) +(a1(t)X(t)+a2(t)v(t)+a3(t))p(t) +(b1(t)X(t)+b2(t)v(t)+b3(t))q(t)+π(t)ρ(t), λ(t)=e−βtand (p(·),q(·),ρ(·)) satisfies the adjoint backward equation        dp(t)=−c1(t)X(t)λ(t)+a1(t)p(t)+b1(t)q(t)+1 2c2(t)λ(t)dt +q(t)dW(t)+ρ(t)dH(t), p(T)=θ+λ(t)ϕ0(X(T)). (38) By using the sufficient condition (19)given in Section 3, we get λ(t)c3(t)v(t)+a2(t)p(t)+b2(t)q(t)=0. (39) To solve (38), we try a process pof the form p(t)=φ(t)X(t)+ψ(t), (40) where φand ψare derivable functions with continuous derivatives. Applying Itô’s formula to (40)and using (36), we get dp(t)=[(φ(t)a1(t)+φ0(t))X(t)+φ(t)a2(t)v(t)+φ(t)a3(t)+ψ0(t)]dt +φ(t)[b1(t)X(t)+b2(t)v(t)+b3(t)]dW(t)+φ(t)π(t)dH(t)(41) Comparing with (38), we obtain ρ(t)=φ(t)π(t), q(t)=φ(t)[b1(t)X(t)+b2(t)v(t)+b3(t)] (42) and −c1(t)X(t)λ(t)+a1(t)p(t)+b1(t)q(t)+1 2c2(t)λ(t) =(φ(t)a1(t)+φ0(t))X(t)+φ(t)a2(t)v(t)+φ(t)a3(t)+ψ0(t).(43) Substituting (42)into (39), we obtain ˆ v(t)=−(a2(t)+b1(t)b2(t))φ(t)ˆ X(t)+a2(t)ψ(t)+b2(t)b3(t)φ(t) λ(t)c3(t)+b2 2(t)φ(t). (44) J. Risk Financial Manag. 2022,15, 143 16 of 19 On the other hand, from (43)we have ˆ v(t)=−c1(t)λ(t)+2a1(t)φ(t)+b2 1(t)φ(t)+φ0(t)ˆ X(t) (a2(t)+b1(t)b2(t))φ(t)(45) −a1(t)ψ(t)+b1(t)b3(t)φ(t)+1 2c2(t)λ(t)+a3(t)φ(t)+ψ0(t) (a2(t)+b1(t)b2(t))φ(t). Combining (44)and (45), we obtain φ2(t)h(a2(t)+b1(t)b2(t))2−b2 2(t)2a1(t)+b2 1(t)i−φ(t)φ0(t)b2 2(t) −c3(t)2a1(t)+b2 1(t)+c1(t)b2 2(t)φ(t)λ(t) −c3(t)φ0(t)λ(t)=c1(t)c3(t)λ2(t),φ(T)=λ(T). (46) and ψ(t)φ(t)a2(t)(a2(t)+b1(t)b2(t)) −a1(t)b2 2(t)−c3(t)a1(t)λ(t) −ψ0(t)c3(t)λ(t)+φ(t)b2 2(t) =−φ2(t)(a2(t)+b1(t)b2(t))b2(t)b3(t)−b2 2(t)(b1(t)b3(t)+a3(t)) +c3(t)(b1(t)b3(t)+a3(t))φ(t)λ(t) +1 2c2(t)b2 2(t)φ(t)+c3(t)c2(t)λ(t)λ(t),ψ(T)=θ−λ(T)c0. (47) Let us emphasize that, under the boundedness of its coefficients, the first order differential Equation (46) admits a unique solution. Moreover, a simple computation shows the (47)has an explicit solution, which is given by ψ(t)=(θ−λ(T)c0)exp(−Λt(T)) −exp(Λt(T)) RT te−Λt(s)Gt(s)ds, where Λt(T)=RT t φ(s)(a2(s)+b1(s)b2(s))a2(s)−a1(s)b2 2(s) c3(s)λ(s)+b2 2(s)φ(s)ds −RT t c3(s)a1(s)λ(s) c3(s)λ(s)+φ(s)b2 2(s)ds, and Gt(s)=−RT t φ2(s)b2(s)b3(s)(a2(s)+b1(s)b2(s)) c3(s)λ(s)+φ(s)b2 2(s)ds +RT t φ2(s)b2 2(s)(b1(s)b3(s)+a3(s)) c3(s)λ(s)+φ(s)b2 2(s)ds +RT t c3(s)(b1(s)b3(s)+a3(s))λ(s)φ(s) c3(s)λ(s)+φ(s)b2 2(s)ds. +RT t 1 2c2(s)b2 2(s)φ(s)+c3(s)c2(s)λ(s)λ(s) c3(s)λ(s)+φ(s)b2 2(s)ds Thus, we have the following theorem: Theorem 3. Let Xv(t) be the cash balance satisfying (36) .Then, the optimal premium policy of Problem A is ˆ v(t)=−(a2(t)+b1(t)b2(t))φ(t)ˆ X(t)+a2(t)ψ(t)+b2(t)b3(t)φ(t) c3(t)λ(t)+b2 2(t)φ(t), where λ(t) = e−βtand φand ψare the solutions of (46)and (47), respectively. J. Risk Financial Manag. 2022,15, 143 17 of 19 This is clearly the unique solution under Assumption (H1). Notice that we can get explicit solutions in some particular cases. For example, if g1= 0 we can see the problem as a mean-variance one, where our objective is to find v(t)such that it minimizes V(X(T)) =Eh(X(T)−E(X(T)))2i, under the terminal constraint condition (4). By the Lagrangian multiplier method, the problem can be reduced to minimize the following equivalent problem: J(v(·)) =E1 2(Xv(T)−c0)2+θ(Xv(T)−c0). (48) In this case, the Hamiltonian (17)gets the form H=(a1(t)X(t)+a2(t)v(t)+a3(t))p(t) +(b1(t)X(t)+b2(t)v(t)+b3(t))q(t) +π(t)ρ(t). Hence, the adjoint equation takes the following form dp(t)=−(a1(t)p(t)+b1(t)q(t))dt +q(t)dW(t)+ρ(t)dH(t) p(T)=θ+(X(T)−c0). Note that in this case λ(t)=1, ∀t∈[0, T]. In the following corollary, we solve Problem (48)assuming that Xv(t)satisfies (36). Corollary 1. Let Xv(t) be the cash balance satisfying (36) .Then, the optimal premium policy of Problem (48)is ˆ v(t)=−(a2(t)+b1(t)b2(t))φ(t)ˆ X(t)+a2(t)ψ(t)+b2(t)b3(t)φ(t) b2 2(t)φ(t), where φand ψare given, respectively, by φ(t)=exp{− RT t (a2(s)+b1(s)b2(s))2−(2a1(s)+b2 1(s))b2 2(s) b2 2(s)ds} and ψ(t)=(θ−c0)exp(Λt(T)) −exp(Λt(T)) RT texp(−Λt(s))φ(s)(a3(s)b2(s)−a2(s)b3(s))b2(s)ds, with Λt(T)=−ZT t (a2(s)(a2(s)+b1(s)b2(s))) −a1(s)b2 2(s) b2 2(s)ds. 4.2.2. Second Case In this case, we assume the liability of the surplus process and the interest rate are given by two different and independent Brownian motions. Let ¯ W:=(¯ W(t))t≥0 be another Brownian motion defined in (Ω,F,P)and independent of Wand L. Assume Ft=σ(W(s),¯ W(s),L(s), 0 ≤s≤t)∨ N . We consider that the interest rate is given by the stochastic differential equation d∆(t)=δ(t)dt +α(t)d¯ W(t),t∈[0, T],∆(0)=0. (49) J. Risk Financial Manag. 2022,15, 143 18 of 19 Itô formula applied to the process X(·)described by (1)in Section 2, leads to        dX(t)=X(t)δ(t)+1 2α2(t)+b(t)+v(t)dt +X(t)α(t)d¯ W(t) +a(t)dW(t)+π(t)dH(t), X(0)=x. (50) Define a process Vsuch that X(t)α(t)d¯ W(t)+a(t)dW(t)=X2(t)α2(t)+σ2(t)1 2dV(t), that is, V(t)=Zt 0X2(s)α2(s)+a2(s)−1 2(X(s)α(s)d¯ W(s)+a(s)dW(s)). Clearly, the process V is a continuous martingale with quadratic variation hVit=t , and so, it must be a standard Brownian motion, see for example Theorem 6.1 in Chung and Williams (1990). Thus, we can write Equation (50)by          dX(t)=X(t)δ(t)+1 2α2(t)+b(t)+v(t)dt +X2(t)α2(t)+a2(t)1 2dV(t)+π(t)dH(t), X(0)=x. (51) We shall derive the solution of Problem A. Here, the Hamiltonian (17)gets the form H=X(t)δ(t)+1 2α2(t)+b(t)+v(t)p+X2(t)α2(t)+a2(t)1 2q +π(t)ρ+1 2e−βthR(t)(X(t)−A(t))2+N(t)v2(t)i. The adjoint process (p(·),q(·),ρ(·)) satisfies the BSDE    −dp(t)=a4(t)p(t)+a5(t)+R(t)(X(t)−A(t))e−βtdt −q(t)dV(t)−ρ(t)dH(t), p(T)=θ+Me−βT(X(T)−c0). where a4(t)=δ(t)+1 2α2(t) and a5(t)=α2(t)X(t)q(t) pX2(t)α2(t)+a2(t) . Notice that in this case, λ(t)=1, ∀t∈[0, T]. By the same technique used in Section 4.1, we infer that the optimal premium policy is given by ˆ v(t)=−N−1(t)eβt(φ(t)X(t)+ψ(t)), where X(·)satisfies (51)and φand ψare, respectively, the solutions of the equations φ0(t)+2δ(t)+α2(t)φ(t)−N−1(t)eβtφ2(t)+R(t)e−βt=0 φ(T)=Me−βT, and    ψ0(t)+δ(t)+1 2α2(t)−N−1(t)eβtφ(t)ψ(t)+b(t)φ(t)−A(t)R(t)e−βt=0, ψ(T)=θ−c0Me−βT. Moreover, the optimal cost functional is given by J. Risk Financial Manag. 2022,15, 143 19 of 19 J(ˆ v(·)) =1 2(RT 0e−βtR(t)A2(t)dt +Me−βTc2 0) + 1 2φ(0)x2+ψ(0)x−c0θ +1 2RT 0φ(t)a2(t)+φ(t)α2(t)+ψ(t)2b(t)−N−1(t)eβtψ(t)dt. Author Contributions: All authors have equally contributed to the paper in all aspects. All authors have read and agreed to the published version of the manuscript. Funding: Research of Josep Vives is partially financed by the Spanish MICIN grant PID2020- 118339GB-100 (2021-2024). Data Availability Statement: Not applicable. Conflicts of Interest: The authors declare no conflict of interest. References Asmussen, Søren, Bent Jesper Christensen, and Julie Thøgersen. 2019. Nash equilibrium premium strategies for push–pull competition in a frictional non-life insurance market. Insurance: Mathematics and Economics 87: 92–100. [CrossRef] Baghery, Fouzia, Nabil Khelfallah, Brahim Mezerdi, and Isabelle Turpin. 2014. Fully coupled forward backward stochastic differential equations driven by Lévy processes and application to differential games. Random Operators and Stochastic Equations 22: 151–61. [CrossRef] Bahlali, Khaled, M’hamed Eddahbi, and El Hassan Essaky. 2003. BSDEs associated with Lévy processes and application to PDIE. Journal of Applied Mathematics and Stochastic Analysis 16: 1–17. [CrossRef] Cairns, Andrew. 2000. Some notes on the dynamics and optimal control of stochastic pension fund models in continuous time. Astin Bulletin 30: 19–55. [CrossRef] Chung, Kai Lai, and Ruth J. Williams. 1990. Introduction to Stochastic Integration, 2nd ed. Basel: Birkhaüser. Hipp, Christian, and Michael Vogt. 2003. Optimal dynamic XL-reinsurance. Astin Bulletin 33: 193–207. [CrossRef] Højgaard, Bjarne, and Michael Taksar. 1998. Optimal proportional reinsurance policies for diffusion models. Scandinavian Actuarial Journal 2: 166–80. [CrossRef] Huang, Jianhui, Guangchen Wang, and Zhen Wu. 2010. Optimal premium policy of an insurance firm: Full and partial information. Insurance: Mathematics and Economics 47: 208–15. [CrossRef] Josa-Fombellida, Ricardo, and Juan Pablo Rincón-Zapatero. 2008. Mean-variance portfolio and contribution selection in stochastic pension funding. European Journal of Operational Research 187: 120–37. [CrossRef] Kalfin, Sukono Firman, Sudrajat Supian, and Mustafa Mamat. 2021. Mitigation and models for determining premiums for natural disaster insurance due to excessive rainfall. Journal of Physics: Conference Series 1722: 012058. [CrossRef] Li, Danping, and Virginia R. Young. 2021. Bowley solution of a mean–variance game in insurance. Insurance: Mathematics and Economics 98: 35–43. [CrossRef] Lin, Jiaxin, Juliang Zhang, and Tai Chiu Edwin Cheng. 2020. Optimal pricing and return policy and the value of freight insurance for a retailer facing heterogeneous consumers with uncertain product values. International Journal of Production Economics 229: 107767. [CrossRef] Meng, QingXin, and MaoNing Tang. 2009. Necessary and sufficient conditions for optimal control of stochastic systems associated with Lévy processes. Science in China Series F: Information Sciences 52: 1982–92. [CrossRef] Moore, Kristen S., and Virginia R. Young. 2006. Optimal insurance in a continuous-time model. Insurance: Mathematics and Economics 39: 47–68. [CrossRef] Ngwira, Bernard, and Russell Gerrard. 2007. Stochastic pension fund control in the presence of Poisson jumps. Insurance: Mathematics and Economics 40: 283–92. [CrossRef] Nualart, David, and Wim Schoutens. 2000. Chaotic and predictable representations for Lévy processes. Stochastic Processes and their Applications 90: 109–22. [CrossRef] Xie, Shuxiang, Zhongfei Li, and Shouyang Wang. 2008. Continuous-time portfolio selection with liability: Mean–variance model and stochastic LQ approach. Insurance: Mathematics and Economics 42: 943–53. [CrossRef]