Filtration consistent nonlinear expectations and evaluations of contingent claims (Q705074): Difference between revisions

Let \(X_t\), \(t\in [0,T]\), be an \(\mathbb R^d\)-valued process defined on a time interval \(t\in [0,T]\). Let \(Y\) be a random value depending on the trajectory of \(X\). Assume that, at each fixed time \(t\leq T\), the information available to an agent is the trajectory of \(X\) before \(t\). Thus at time \(T\), the random value of \(Y(\omega)\) will become known to this agent. The question is: how will this agent evaluate \(Y\) at the time \(t\)? If this \(Y\) is traded in a financial market, it is called a contingent claim, i.e., a contract whose outcome depends on the evolution of the underlying process \(X\). If \(Y\) is negative, it represents the measure of the risk of \(Y\) by this agent. Over the past 30 years, the well-known Black \(\&\) Scholes has been a very important model of evaluation by a financial market of the contingent claims. In this paper, the evaluation of \(Y\) is treated from a more general viewpoint. An evaluation operator \(\mathcal E_t[Y]\) is introduced to define the value of \(Y\) given by this agent at time \(t\). This operator \(\mathcal E_t[\cdot]\) assigns an \((X_s)_{0\leq s\leq T}\)-dependent random variable \(Y\) to an \((X_s)_{0\leq s\leq t}\)-dependent random variable \(\mathcal E_t[Y]\). Although this value \(\mathcal E_t[Y]\) is very complicated and is different from one agent to another, the author introduces some axiomatic conditions to describe the mechanics of this operator. Some typical situations with incomplete statistical data of the model of \(X\) are considered. The so called super evaluation and the mechanism of evaluations dominated by these evaluations are introduced. In many situations, the evaluation operators are filtration consistent nonlinear expectations. In some situations, this evaluation coincides with \(g\)-expectation which is the solution of a backward stochastic differential equation (BSDE). A more general fully nonlinear situation is also considered. The author mainly treats the situation in which the process \(X\) is a solution of a SDE \[ dX_s=b(X_s,{\overline\theta}_s)\,ds+ \sigma(X_s,{\overline\theta}_s)\,dB_s, \quad s\geq0,\quad X_0=0,\quad{\overline\theta}\in\Theta, \] in which the drift coefficient \(b\) and diffusion coefficient \(\sigma\) contain unknown parameter \(\theta \). The super evaluation, when the agent is a seller of the asset \(Y\), is considered.