1: \begin{abstract}

2: \noindent Theoretical results for importance sampling rely on the existence of certain moments

3: of the importance weights, which are the ratios between the proposal and target densities.

4: In particular, a finite variance ensures square root convergence and asymptotic normality of the importance sampling estimate,

5: and can be important for the reliability of the method in practice.

6: We derive conditions for the existence of any required moments of the weights for Gaussian proposals

7: and show that these conditions are almost necessary and sufficient for a wide range of models with latent Gaussian components.

8: Important examples are time series and panel data models with measurement densities which belong to the exponential family.

9: We introduce practical and simple methods for checking and imposing the conditions

10: for the existence of the desired moments.

11: We develop a two component mixture proposal that allows us to flexibly adapt

12: a given proposal density into a robust importance density.

13: These methods are illustrated on a wide range of models

14: including generalized linear mixed models,

15: non-Gaussian nonlinear state space models and panel data models with autoregressive random effects.

16: %\footnotesize{

17: \bigskip

18: 

19: \noindent {\sc Keywords}:  MCMC, simulated maximum likelihood, state space models, robustness

20: 

21: %\medskip

22: %\noindent {\sc JEL}: C32, C51, E43.

23: %}

24: \end{abstract}