Continuous \(\Theta\)-methods for the stochastic pantograph equation (Q1595879): Difference between revisions

The one-dimensional stochastic Itô pantograph equation \[ dX(t) = \{ a X(t) + b X(qt) \} dt + \{\sigma_1 +\sigma_2 X(t) +\sigma_3 X(qt)\} dW(t), \quad X(0) = X_0, \] for \(t\in [0, T]\) (\(T\) fixed and finite), driven by a Wiener process \(W\) and with deterministic delay parameter \(0<q<1\) is studied, both analytically and numerically. The authors discuss existence and uniqueness of strong solutions of this bilinear Itô stochastic delay differential equation. Finally, they study the mean square accuracy of approximations to the solution obtained by a continuous extension of the \(\Theta\)-Euler scheme with deterministic implicitness \(\Theta\in [0, 1]\). They prove the rate \(\gamma = 0.5\) of mean square convergence of these approximations using a bounded mesh of uniform step size \(h\), rising in the case of additive noise to \(\gamma = 1.0\). Some illustrative numerical experiments supplement this paper.