An Empirical Study On The Structural Change Of Stock Volume And Price Level Based On PELT Algorithm

The emergence and development of the stock market has greatly promoted the formation and improvement of the market-oriented resource allocation mechanism.China's stock market is a emerging market,with a higher proportion of retail investors than that of foreign countries.In most cases,investors do not know the timing of the introduction of major policies and the strength of policies.At the same time,investors are sensitive to various policies,such as monetary,fiscal policy,trade policy,etc.and various information,such as IPO announcements,asset restructuring,financial crisis,that are very sensitive factor lead people to catch up when increasing and withdraw when falling,causes abnormal volatility in stock returns,resulting in jump behavior.At the same time,we are also concerned that stock prices and trading volume are also changing,and showing structural jump behavior.Change-point analysis is a hot topic in the study of statistics,and it is a new field in statistics.By analyzing the characteristics of structural change-points,it is of great significance to understand the changing regulation of things and to formulate relevant countermeasures.We have noticed that the occurrence of structural change points of stock volume and price will bring huge impact to the stock market.Studying the rule behind the structural change points of stock volume and price plays a positive role in guarding against investor investment risk and stock market risk.This paper selects the CSI 300 index component stocks as the research object,and selects 165 sustained stable corporate stocks from January 1 2008 to December 31 2018 as the research sample,takes the stock daily return,closing price,trading volume,turnover rate,daily amplitude as the research variables.Selects the structure of stock change-points by Pruned Exact Linear Time(PELT)algorithm,and then selects overlapping points of change-points and jumping points to make comparative analysis.Use the event study method to estimate the cumulative abnormal return rate,average standardized abnormal volume,average standardized abnormal turnover rate and average standardized abnormal daily amplitude to compare and analyze the differentproperties of structural change points,jump points and overlapping points.To study the structural changes of stock price level and analyze the trading characteristics of domestic stock market.Then through the empirical results of the stock price data from January 1 2019 to September 30,it can be found out the solutions when facing the following five points,such as price and trading volume increase at the same time,price increase but trading volume remains unchanged,slightly decreasing both price and trading volume,price decreasing but trading volume increasing,trading volume increasing but price remains unchanged.Thus this study makes great contribution to help investors to make the right decision when investing in stock market,and also offers theoretical support for further study.