“Formulae of differentiation” and their use for solving stochastic equations

New tools for the solution of stochastic differential equations are presented. It is shown that there exist simple formulae for splitting correlations, which arise when dealing with widely used models of random processes such as Poisson, gaussian and Markov jump processes with an exponential form of the covariance. The use of the formulae is demonstrated.