A Study On Measuring Information Risk Of Stocks

Measuring information risk of stocks accurately has important significance on asset pricing, risk management and measuring the market performance. If information risk of stocks is higher, or the extent of information asymmetry is more serious in one country, then the justice and efficiency of market in the country will be damaged. Every country's supervisors of stock market try to reduce information risk or the extent of information asymmetry in stock market in order to maintain the justice and efficiency of market. However, there is still different extent of information risk or information asymmetry in every county's stock market.Then, is information risk of stocks a systematic risk? If it is, investors who hold stocks with higher information risk should require obtaining higher risk premium. When we want to test whether information risk is a systematic risk empirically, we firstly need to measure information risk accurately. Easley, Kiefer, O'Hara and Paperman (1996) first put forward the PIN model used to directly measure information risk of stocks, hereafter this model became one of the most popular models used to meaure information risk. However, the negative correlation between buy and sell orders implied in the PIN model is not accordant with the positive correlation between buy and sell orders in real data. Furthermore, the variance of buy and sell orders implied in the PIN model can't match the larger variance of buy and sell orders in real data either.This paper put forward a corrected PIN model by adding trading motive based on the classical PIN model put forward by Easley, Kiefer, O'Hara and Paperman (1996). Based on the trading data tick by tick of Chinese stocks, this paper applied the corrected PIN model to measure information risk of stocks empirically, and found that the correlation between buy and sell orders and the variance of buy and sell orders implied in the corrected PIN model can better match the real data.This paper empirically measured information risk of stocks with different trading activity using the classical PIN model and the corrected PIN model respectively, and found that information risk of stocks is negative with trading activity. The classical PIN model tended to overestimate information risk of stocks due to ignoring the trading motive aroused by market order flow shock.This paper constructed the time series of information risk of stocks using the method of rolling, and did an empirical study on the relation between information risk of stocks and stock returns, found that information risk had not significant impact on stock returns, while the probability of market order flow shock which is related to liquidity had persistent and significant impact on stock returns.