1: \begin{abstract}

2: It is a well-established fact in the scientific literature that Picard iterations of backward stochastic

3: differential equations with globally Lipschitz continuous nonlinearity converge at least exponentially fast to the solution.

4: In this paper we prove that this convergence is in fact

5: at least square-root factorially fast.

6: We show for one example that no higher convergence speed is possible in general.

7: Moreover, if the nonlinearity is $z$-independent,

8: then the convergence is even factorially fast.

9: Thus we reveal a phase transition in the speed of convergence

10: of Picard iterations of backward stochastic differential equations.

11: \end{abstract}