No. 1039

Title:

Emprirical Analysis for Estimating Timevaring Volatility with Mixtured System Noise Model

Author(s):

Yamashita, Takashi (Department of Statistical Science The graduate University for Advanced Studies 4-6-7 Minami-Azabu, Minato-ku, Tokyo 106-8569, Japan);
Ozaki, Tohru (Department of Statistical Science The graduate University for Advanced Studies 4-6-7 Minami-Azabu, Minato-ku, Tokyo 106-8569, Japan)

Key words:

Stock price; GARCH; timevarying volatility; jump detection; innovation approach; kalman filter

Abstract:

    The purpose of my work is researching the negative impact of price jumps to estimate timevaring volatility using the GARCH model introduced by Nelson and Foster, then report it by empirical analysis. It was found that the distribution of the return does not show the normal distribution that is assumed in practical realm. The returns volatility is not constant and exist price jumps. Conventional GARCH model driven by Gaussian system noise overestimates the volatility because it uses its estimation error as the system noise of the volatility. We can improve fitness performance for timevarying volatility using mixtured Poisson noise system model with jump detection technique.