Milstein schemes for delay McKean equations and interacting particle systems
In this paper, we derive fully implementable first order time-stepping schemes for point delay McKean stochastic differential equations (McKean SDEs), possibly with a drift term exhibiting super-linear growth in the state component. Specifically, we propose different tamed Milstein schemes for a time-discretised interacting particle system associated with the McKean equation and prove strong convergence of order 1 and moment stability, making use of techniques from calculus on the space of probability measures with finite second order moments. In addition, we introduce a truncated tamed Milstein scheme based on an antithetic multi-level Monte Carlo approach, which leads to optimal complexity estimators for expected functionals without the need to simulate Lévy areas.
READ FULL TEXT