Pricing a contingent claim liability with transaction costs using asymptotic analysis for optimal investment (Q457188)

Using asymptotic analysis for optimal investment, the author investigates pricing a contingent claim liability with transaction costs. He considers the problem of an investment on a finite interval \([0,T]\). Let \(\{Z_{t}\}_{0\leq t\leq T}\) be a standard Brownian motion. The market model evolutions for the wealth invested in bonds, the number of shares of a stock, and the stock price are given by \(dB_{s}=rB_{s}ds-(1+\varepsilon)S_{s}d L_{s}+ (1-\varepsilon)S_{s}dM_{s}\), \(B_{-t}=B\), \(dy_{s}=dL_{s}-dM_{s}\), \(y_{-t}=y\), \(dS_{s}=S_{s}(\mu ds+\sigma dZ_{s})\), \(S_{t}=S\), where \(r,\mu\) and \(\sigma\) are nonnegative constants, \(\mu>r\), \(\varepsilon\in(0,1)\) is the proportional transaction costs, and \(L,M\) represent the cumulative numbers of shares bought or sold respectively at time \(s\). Let us denote by \(c(y,S)=(1-\varepsilon\,\text{sign}\,y)yS\) the cash value of a number of shares of stock \(S\). The author considers two problems of optimal investment: the problem without a claim (\(j=1)\), and the problem with a claim with smooth payoff \(g(S_{T})\) (\(j=w\)). The wealth at the final time \(T\) is given by \(\Phi^{(1)}(T,B_{T},y_{T},S_{T})=B_{T}+c(y_{T},S_{T})\), \(\Phi^{(w)}(T,B_{T},y_{T},S_{T})=B_{T}+c(y_{T}-g'(S_{T}),S_{T})-(g(S_{T})-g'(S_{T})S_{T})\). The value functions are defined as \(V^{(j)}(t,B,y,S)=\sup E_{t}^{B,y,S}[U(\Phi^{(j)}(T,B_{T},y_{T},S_{T}))]\), \(j=1,w\), where the supremum is taken over all admissible strategies. An asymptotic expansion of the value function in powers of \(\varepsilon^{1/3}\) is presented. Also it is obtained a ``nearly optimal'' trading policy that, if followed, produces an expected utility of final wealth that asymptotically matches the value function at the order \(\varepsilon^{2/3}\). As an application of the main results the author considers pricing securities with a mollified call option payoff and a mollified barrier option payoff.