Empirical Research On The Number Of Stock Transactions And Closing Prices Based On VAR Model And INAR Model

As Chinese stock market meeting its rapid development,financial market risk is becoming clearer to regulators and investors of the market.For better avoiding risk in the stock market and relevant operations,we have to observe the changes of indexes in stock exchange market,and to timely carry out macro-control and moderate intervene for the stock market.Knowing the fluctuation relation between number of transactions and closing price will help regulators to take effective risk management to prevent the stock market from fluctuating violently,as well as help investors to predict the market,reduce investment risk and increase their investment return.As a key concept in the stock market,transaction volume means a general term for stock number traded and trading amount.Stock number traded is the sum of shares exchanged in a day for one certain stock,acting as an element basis for judging the trend of stock.Its fluctuation should be followed closely by investors.Number of transaction means how many times the exchange happens,and stock number traded in one transaction is uncertain.The closing price means the volume weighted average price(VWAP)of all exchanges within the minute before the last trade of the day,acting as most intuitive data in the market.Usually in empirical research on stock only stock yield rate and transaction volume are involved.Innovation point of this paper is that a brand new index is introduced,namely the number of transaction,whose causal link with closing price is researched to help us understand its effect on stock market.First of all,we build a Vector Autoregressive Model(VAM)to select three representative stocks with complete historical data listed from Shanghai Stock Exchange,namely,Juhua,LANHUA SCI-TECH VENTURE,and Fujian Expressway.Stationarity test,impulse response analysis,variance decomposition and granger causality test are carried out for their daily number of transaction and closing price for the past five years.For further investigation of the effect of number of transaction on thedevelopment trend of stock market,we build an integer numerical autoregressive model(INAR)that can fit better the integer data of number of transaction.Its comparison with VAR model shows that INAR has a better fitting result for integer numerical sequence.The paper will help us understand the concepts of number of transaction and closing price of stock,and then to understand better the financial market risk.