1: \begin{abstract}

2: In this paper we undertake the error analysis of the

3: time discretization of systems of Forward-Backward Stochastic Differential

4: Equations (FBSDEs) with drivers having polynomial

5: growth and that are also monotone in the state variable. 

6: 

7: We show with a counter-example that the natural explicit Euler

8: scheme may diverge, unlike in the canonical Lipschitz driver case. 

9: This is due to the lack of a certain stability property of the Euler  scheme

10: which is essential to obtain convergence. 

11: However, a thorough analysis of the family of $\theta$-schemes reveals  that

12: this required stability property can be recovered if the scheme is

13: sufficiently implicit. As a by-product of our analysis  we shed some light on

14: higher order approximation schemes for FBSDEs under non-Lipschitz condition.

15: We then return to fully explicit schemes and show that an appropriately tamed

16: version of the explicit Euler scheme enjoys the required stability property

17: and as a consequence converges. 

18: 

19: In order to establish convergence of the several discretizations we  extend

20: the canonical path- and first order variational regularity results 

21: to FBSDEs with polynomial growth drivers which are also monotone. These 

22: results are of independent interest for the theory of FBSDEs. 

23: \end{abstract}