CMStatistics 2017: Start Registration
View Submission - CFE
Title: Indirect inference for locally stationary models Authors:  Bonsoo Koo - Monash University (Australia) [presenting]
David Frazier - Monash University (Australia)
Abstract: An indirect inference method is proposed for locally stationary processes. We develop a local indirect inference algorithm and establish the asymptotic properties of the proposed estimator. The convergence rate of the proposed estimator follows the usual semiparametric one due to the employment of the semiparametric kernel estimation procedure. We also validate our methodology via simulation studies. Our local indirect inference method is applied to locally stationary moving average models and locally stationary multiplicative stochastic volatility models.