A0492
Title: Adaptive change point estimation: Interval time series analysis for GBM models
Authors: Li-Hsien Sun - National Central University (Taiwan) [presenting]
Chi-Yang Chiu - University of Tennessee Health Science Center (United States)
Abstract: A method for detecting structural shifts is proposed within time series data, and the change-point estimation is obtained. In the field of finance, most models are developed for the daily closing price. Nevertheless, based on the intra-daily information from the financial market, maximum and minimum prices can also be observed. Hence, instead of a one-dimensional time series, an interval time series model is proposed that includes the daily maximum, minimum, and closing prices based on the geometric Brownian motion (GBM) model. The likelihood function and the corresponding maximum likelihood estimates (MLEs) are obtained using the Girsanov theorem and the Newton-Raphson (NR) algorithm. The proposed approach is evaluated through simulations. In empirical studies, the performance relies on real stock return data (S\&P 500 index) in two distinct periods: the 2008 financial crisis and the COVID-19 pandemic in 2020.
