The study of volatility is of considerable interest because of its importance in portfolio allocation, risk management, derivatives pricing, futures hedging, trading strategies and asset pricing. Several studies have highlighted the importance of volatility estimators that utilise the opening, high, low and closing prices. Among them, the Yang & Zhang (YZ) estimator proposed by Yang & Zhang (2000) is unbiased regardless of the drift parameter, and incorporates opening jump in estimation. In this paper, we make use of the variance estimator that utilises high, low and closing prices proposed by Yang & Zhang (2000) in the context of detecting sudden changes in volatility. In particular, we investigate the performance of Inclan and Tiao's (1994) Iterated Cumulative Sum of Squares (IT-ICSS) algorithm based on the YZ estimator and compare it with the performance of the demeaned squared returns as alternative estimators of the unconditional variance of a time series by means of Monte Carlo simulation experiments.
This study compares the size and power properties of IT-ICSS test with respect to both the volatility proxies for various data generating processes like the independently and identically distributed (i.i.d.) random numbers from the Gaussian, Student’s t, double exponential, gamma-mixture and generalised error distributions (GED), the generalised autoregressive conditional heteroskedasticity (GARCH) model, the stochastic volatility (SV) model and the fractionally integrated GARCH (FIGARCH) model. For the GARCH, the SV and the FIGARCH models, the innovations have been taken from the normal, the Student’s t and the GED distributions.
The findings of this study indicate that the YZ estimator exhibits more desirable size and power characteristics when applied with IT-ICSS algorithm than the demeaned squared returns. Hence, this study proposes the use of the YZ estimator with IT-ICSS test to detect sudden changes in volatility. On the application side, this study detects sudden breaks in the YZ estimator and the demeaned squared returns of three major exchange rates (USD/Euro, USD/Japanese Yen and USD/GBP).
