We develop a constrained indirect inference approach based on a wrapped skewed-t auxiliary model for the estimation of the wrapped stable distribution. To improve the finite-sample properties of the estimators, we devise a bootstrap-based estimate of the weighting matrix employed in the indirect inference program. The simulation study suggests that, in terms of root-mean-squared-error, the indirect inference estimator of the skewness parameter is better than the corresponding maximum likelihood estimator, whereas maximum likelihood is mostly preferable for the index of stability.

Keywords. Indirect inference; circular statistics; stable distribution; weighting matrix.