On the moments of certain stochastic integrals (Q689502)

Let \(W(t)\) be a Wiener process and \(B(t)=W(t)-tW(1)\) the Brownian bridge associated to \(W(t)\). The authors consider the functional \(F=\int^ 1_ 0 W(t)dB(t)\) and compute its expectation and variance in two different ways. Since \(W\) and \(B\) are dependent, the well-known formulas cannot directly be applied. In the first approach the integral is approximated by finite sums and then the limit is taken. The second way uses the Malliavin calculus by explicitly computing \({\mathcal L}F\) and \({\mathcal L}(F^ 2)\) for the above \(F\) and using the fact \(E\bigl( {\mathcal L}\Phi(F) \bigr)=0\) for smooth \(\Phi\).