TESTS FOR CONDITIONAL HETEROSCEDASTICITY WITH FUNCTIONAL DATA AND GOODNESS-OF-FIT TESTS FOR FGARCH MODELS

浙江大学数学科学学院九十周年院庆系列活动之六十五

TESTS FOR CONDITIONAL HETEROSCEDASTICITY WITH FUNCTIONAL DATA ANDGOODNESS-OF-FIT TESTS FOR FGARCH MODELS

时间 / Date andTime:          2018/08/06 (Mon), 10:00 - 11:00

地点 / Venue:                    浙江大学玉泉校区逸夫工商楼200-9楼报告厅

报告人 / Speaker:               Professor Tony Wirjanto

                                     University of Waterloo

                                      https://uwaterloo.ca/statistics-and-actuarial-science/people-profiles/tony-wirjantol

摘要 / Abstract:

Functionaldata objects that are derived from high-frequency financial data often exhibitvolatility clustering characteristic of conditionally heteroscedastic timeseries. Versions of functional generalized autoregressive conditionallyheteroscedastic (FGARCH) models have recently been proposed to describe suchdata, but so far there are no diagnostic tools available for these models. Wepropose two portmanteau type tests to measure conditional heteroscedasticity inthe squares of return curves. A complete asymptotic theory is provided for eachtest, and we further show how these methods can be applied to model residualsin order to evaluate the adequacy and aid in order selection of FGARCH models.Simulation studies show that both tests have good size and power to detectconditional heteroscedasticity and model mis-specification in finite samples.In an application, the proposed methods suggest that intra-day asset returncurves exhibit conditional heteroscedasticity. Moreover, we find that themagnitude of inter-daily returns alone is not sufficient to capture thisconditional heteroscedasticity, but it is adequately modeled by an FGARCH(1,1)model.

欢迎各位师生参加

*** ALL ARE WELCOME ***

联系人 / Enquiries:

张  奕 / Yi Zhang

Tel: (86) 13588118020   Email: zhangyi63@zju.edu.cn