The Method And Application Of Weighted Median Realized Volatility In High Frequency Financial Data

Financial market information continuously affects the price change process ofsecurities market, however the discrete degree of data acquisition will lead to the loss ofinformation in different level. With the rapid development of storage and computingtechnology, high frequency data with real-time transaction acquisition market become true,the high frequency data measurement and application has become an important trend in thefinancial field. Effective identification of financial asset return volatility directlydetermines the quality of the risk management. VaR is the common methods of riskmeasurement. Various risk measurement models has been put forward by scholars at homeand abroad. The purpose of this paper is to find a better risk measurement model throughtheoretical comparison, study whether there are differences between the two markets in ourcountry. The work of this paper is organized as follows:Based on the study among “realized volatility (RV), two bi-power realized variation(BV), tri-power variation (TV), the median realized volatility (MedRV)” this paperconstructed the weighted median realized volatility (WMedRV).,Which contains themedian realized volatility (MedRV) and the weight realized volatility (RRV). Furthermore,proved that weighted median realized volatility (WMedRV) have good properties: such asminimum variance and unbiased.In addition, in order to observe the effectiveness of this method, this paper designedthe application of the classical VaR method in our country’s negotiable securities market.