Title: Estimation of jump diffusion models by Jarque-Bera normality test
Authors: Yuma Uehara - The Institute of Statistical Mathematics (Japan) [presenting]
Hiroki Masuda - Kyushu University (Japan)
Abstract: The estimation problem of jump diffusion models based on high-frequency samples is considered. A well-known estimation approach for the models is to use a jump detection threshold. However, in practical use, the choice of the threshold still remains as an annoying problem. To avoid such a difficulty, we propose a new estimation method based on Jarque-Bera normality test. We will present some numerical experiments and the theoretical properties of our method.