To model the assets and liabilities, we consider a probability space
\(({\varOmega },{\mathscr {H}},{\mathbb {Q}})\), with
\({\mathscr {H}}={\mathscr {F}}\vee {\mathscr {G}}\) and
\({\mathscr {F}}\) and
\({\mathscr {G}}\) independent. We further assume that
\({\mathbb {F}} {=} ({\mathscr {F}}_{t \in [0,T]})\) is a filtration on
\({\mathscr {F}}\), generated by a Brownian motion
W satisfying the usual conditions and
\(T>0\) denotes the time horizon for the investor.
W is used to model the risky asset and hedgeable liability risks and
\({\mathscr {G}}\)-measurable random variables are used to model the unhedgeable risks. We introduce the risk-free asset
\(P_0(t)\) satisfying
$$\begin{aligned} dP_0(t) = P_0(t) r(t) dt, \end{aligned}$$
with
r(
t) being a non-negative, absolutely-integrable, deterministic function on [0,
T]. We also assume that the investor can invest in a risky asset
P following the dynamics
$$\begin{aligned} dP(t)=P(t)\left( \mu (t) dt + \sigma (t) dW(t)\right) , \end{aligned}$$
(1)
where
\(\mu (t)\) is an absolutely-integrable, and
\(\sigma (t)\) is a square-integrable, strictly positive, deterministic function on [0,
T]. Depending on the application, we may also restrict
\(\mu (t),\sigma (t)\) and
r(
t) to be constant.
\({\tilde{Z}}(t)\) denotes the pricing kernel in this complete market given by
$$\begin{aligned} {\tilde{Z}}(t) := e^{-\int _0^t r(s)+\frac{1}{2}\gamma ^2(s)ds - \int _0^t\gamma (s)dW(s)}, \end{aligned}$$
with
\(\gamma (t) := (\mu (t)-r(t))/\sigma (t)\) being the market price of risk. We assume that the pricing kernel fulfills
\({\mathbb {E}}[{\tilde{Z}}(T)] < \infty \) and we write
$$\begin{aligned} {\tilde{Z}}(t,T):=\frac{{\tilde{Z}}(T)}{{\tilde{Z}}(t)}. \end{aligned}$$
We consider
\({\mathbb {F}}\)-progressively measurable, self-financing investment strategies with
\(\pi (t)\) being the fraction of wealth invested in the risky asset at time
\(t\in [0,T]\). The remaining portion
\(1- \pi (t)\) is invested in the money market account. The wealth process is denoted by
V(
t) and the initial wealth by
\(V(0)=v_0>0\). The dynamics of
V(
t) are given by
$$\begin{aligned} dV(t) = V(t)\left( (r(t) + \pi (t)(\mu (t)-r(t)))dt + \pi (t)\sigma (t)dW(t)\right) . \end{aligned}$$
(2)
We assume
$$\begin{aligned} \int _0^T \pi ^2(s)V^2(s)ds < \infty , \quad {\mathbb {Q}}\text {-a.s.} \end{aligned}$$
To set up the portfolio optimization problem, we first introduce the liabilities in a way which is inspired by the illiquid asset in [
4]. We assume that the value of the liabilities at time
T can be modeled by random variables
\(L_1(T)\) and
\(L_2(T)\) and that we cannot invest in the liabilities directly.
\(L_1(T)\) will be used to model a performance-participating part and
\(L_2(T)\) will be used to model a part which is not directly depending on the wealth.
\(L_2(T)\) can be interpreted, e.g., as an index-linked part. For both,
\(L_1\) and
\(L_2\), we want to allow for hedgeable and non-hedgeable components. The hedgeable risks can be interpreted as, e.g., interest-rate risks and the unhedgeable risks as, e.g., inflation risks (income growth of the policy holder), mortality risk or operational risk inherent in the liabilities. The hedgeable components are modeled by a stochastic process
X, which satisfies the following assumption.
We call the left part of the sum
\(vL_1(T,X(T),{\mathscr {U}}_1)\) performance-linked and the right part
\(L_2(T,X(T),{\mathscr {U}}_2)\) index-linked. To simplify the notation, we will write
L(
T) or
L(
T,
v) instead of
\(L(T,v,X(T),{\mathscr {U}}_1,{\mathscr {U}}_2)\) and
\(L_i(T)\) instead of
\(L_i(T,X(T),{\mathscr {U}}_i)\),
\(i=1,2\) sometimes. For each
\(\omega \in {\varOmega }\), we further assume the existence of the worst case scenario with respect to the unhedgeable risks defined by
$$\begin{aligned} {\hat{\omega }}_i:=\arg \sup _{{\hat{\omega }}\in {\varOmega }}L_i(T,X(T,\omega ),{\mathscr {U}}_i({\hat{\omega }})),\,i=1,2. \end{aligned}$$
We want to maximize the expected utility of the surplus of the assets over the liabilities at time
T and thus define the surplus as introduced in [
20] by
$$\begin{aligned} S(T):=V(T)-\psi _LL(T,V(T)), \end{aligned}$$
where the factor
\(\psi _L\in (0,1]\) is constant as in [
20] and allows us to consider a flexible portion of the liabilities. In
T, the liabilities are covered even for the worst outcomes of the unhedgeable risk components, if
$$\begin{aligned} V(T,\omega )\ge \psi _LV(T,\omega )L_1(T,X(T,\omega ),{\mathscr {U}}_1({\hat{\omega }}_1))+\psi _LL_2(T,X(T,\omega ),{\mathscr {U}}_2(\hat{\omega _2})), \end{aligned}$$
(4)
i.e. if
\(V(T,\omega )\ge {\hat{v}}_0(\omega )\) with
$$\begin{aligned} {\hat{v}}_0(\omega ):=\frac{\psi _LL_2(T,X(T,\omega ),{\mathscr {U}}_2({\hat{\omega }}_2))}{1-\psi _LL_1(T,X(T,\omega ),{\mathscr {U}}_1({\hat{\omega }}_1))}. \end{aligned}$$
(5)
In particular, this means that the investor has enough initial capital to hedge the worst outcomes of the unhedgeable risk components associated with
\({\mathscr {U}}_i\),
\(i=1,2\). For a pension plan or insurance company with liabilities consisting of the discounted cashflows of the future payments, an upper bound could be, e.g., the sum of all (non-discounted) payments or a cap in benefits to the policy holders (see [
6]). For a discussion of optimal investment strategies in a liability-driven investment context with underfunding, which is not covered here, see e.g. [
1,
10] or [
3]. A self-financing strategy
\(\pi (t)\) is called admissible if
$$\begin{aligned} V(t)-{\mathbb {E}}\left[ {\tilde{Z}}(t,T){\hat{v}}_0(\omega )|{\mathscr {F}}_t\right] \ge 0 \; {\mathbb {Q}}\text{-a.s. } \text{ for } \text{ all }\, t \in [0,T] \end{aligned}$$
(6)
and
\({\mathbb {E}}\left[ U^-(S(T))\right] <\infty \). We denote the set of all admissible strategies corresponding to initial wealth
\(v_0\) by
\({\varLambda }(v_0)\).