1: \begin{abstract}

2: \abstract{We present a systematic study of moment evolution in

3: multidimensional stochastic difference systems, focusing on

4: characterizing systems whose low-order moments diverge in the

5: neighborhood of a stable fixed point. We consider systems with a

6: simple, dominant eigenvalue and stationary, white noise. When the

7: noise is small, we obtain general expressions for the approximate

8: asymptotic distribution and moment Lyapunov exponents. In the case

9: of larger noise, the second moment is calculated using a different

10: approach, which gives an exact result for some types of noise.  We

11: analyze the dependence of the moments on the system's dimension,

12: relevant system properties, the form of the noise, and the

13: magnitude of the noise. We determine a critical value for noise

14: strength, as a function of the unperturbed system's convergence

15: rate, above which the second moment diverges and large

16: fluctuations are likely. Analytical results are validated by

17: numerical simulations. We show that our results cannot be extended

18: to the continuous time limit except in certain special cases.

19: %\end{abstract}

20: