Abstract

Realized range and bipower variance are two important improvements for increasingly popular high-frequency realized estimators in financial markets. This paper verifies a new type of estimator named realized range-based bipower variance and carries out its empirical research with high frequency data from Shanghai Composite Index (SHCI) and Shenzhen Synthesis Index (SZSI). The results show that 1) this estimator combines merits of bothrealized range and realized bipower variance. It is as efficient as realized range estimator and at the same timeremains a consistent estimation of integrated variance of Chinese financial markets’ fluctuation; 2) After standardized by realized range based bipower variance, the distributions of SHCI and SZSI’s daily returns are neither skew nor with high kurtos is anymore. The fact tail and high peak of daily returns are basically eliminated by this high-frequency variance estimator, and the standardized distributions of these returns are nearly normal. 3) Comparative studies show that among four types of realized volatility estimators, the range based bipower variance is the best one to rebuild normality in Chinese financial markets. These findings mean when measuring volatility or fluctuations of financial assets, the usage of this new estimator will increase the performance of many financial practices like pricing or risk management. One feasible way to extend this paper is to consider co-estimators of related assets and detect their impacts to the dynamics of volatilities.

Highlights

  • Introduction and Literature ReviewLiteratures of using range as the proxy of volatility are prolific

  • Beckers (1983) incorporates the information of close prices into the range estimator and adjusted the estimator by Parkinson (1980), his empirical study of 208 kinds of stocks and options proves that the performance of range estimator is much superior to the traditional return estimator

  • Wiggins (1991) shows that compared with the return estimator, the range estimator have a problem of downward bias

Read more

Summary

Introduction and Literature Review

Literatures of using range as the proxy of volatility are prolific. According to the definition of Chou (2005), range is the distance between the highest and the lowest prices of assets in some fixed sampling period. Mandelbrot (1963) uses range to test the long-run dependence characteristic of asset prices. Parkinson (1980) argues the natural logarithm of stock prices roughly follow Random Walk process with a constant diffusion parameter which equals the variation of returns. Parkinson (1980) argues the natural logarithm of stock prices roughly follow Random Walk process with a constant diffusion parameter which equals the variation of returns. He compared variations of the range estimator with variations of the traditional return estimator and found that the range estimator is five times more efficient than the return estimator. Beckers (1983) incorporates the information of close prices into the range estimator and adjusted the estimator by Parkinson (1980), his empirical study of 208 kinds of stocks and options proves that the performance of range estimator is much superior to the traditional return estimator. Andersen and Bollerslev (2001) find that range estimator provides a higher R-square value than the traditional return estimator in Mincer-Zarnowitz (MZ) regression when using realized volatility as the proxy of true underlying volatility

Theory with High Frequency Data in Financial Markets
Realized Range Based Volatility Estimator and Its Consistency
Results
Discussions
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call