Optimal Stopping in the <i>L</i> log <i>L</i> -Inequality of Hardy and Littlewood (Q4354262)

Let \(B=(B_t)_{t\geq 0}\) be standard Brownian motion started at zero. We prove: \[ E\Bigl( \max_{0\leq t\leq\tau}|B_t|\Bigr)\leq 1+\frac{1}{e^c(c-1)}+ cE(|B_\tau|\log^+|B_\tau|) \] for all \(c>1\) and all stopping times \(\tau\) for \(B\) satisfying \(E(\tau^r)<\infty\) for some \(r>1/2\). This inequality is sharp, and the equality is attained at the stopping time \[ \tau_*= \inf\{t>0\mid S_t\geq u_0,\;X_t=1\vee\alpha S_t\} \] where \(\alpha=(c-1)/c\) for \(c>1\), \(X_t=|B_t|\) and \(S_t= \max_{0\leq r\leq t}|B_r|\). Likewise, we prove \[ E\Bigl( \max_{0\leq t\leq\tau}|B_t|\Bigr)\leq \frac{1}{e^c(c-1)}+ cE(|B_\tau|\log^+|B_\tau|) \] for all \(c>1\) and all stopping times \(\tau\) for \(B\) satisfying \(E(\tau^r)<\infty\) for some \(r>1/2\). This inequality is sharp, and the equality is attained at the stopping time \[ \sigma_*=\inf \{t>0\mid S_t\geq v_0,\;X_t=\alpha S_t\} \] where \(\alpha=(c-1)/c\) for \(c>1\). These results contain and refine the results on the \(L\log L\)-inequality of \textit{D. Gilat} [Proc. Am. Math. Soc. 97, 429-436 (1986; Zbl 0654.60033)] which are obtained by analytic methods. The method of proof given here is probabilistic and is based upon solving the optimal stopping problem with the payoff \(V=\sup_\tau E(S_\tau- cF(X_\tau))\) where \(F(x)\) equals either \(x\log^+x\) or \(x\log x\). This optimal stopping problem has some new interesting features, but in essence is solved by applying the principle of smooth fit and the maximality principle. The results extend to the case when \(B\) starts at any given point (as well as to all non-negative submartingales).