A Study On Returns Of Listed Companies In China

Since 1980s, the Capital Asset Pricing Theory based on Efficient Market Hypothesis has been challenged by much empirical evidence. The so called anomalies on the security market including Size Effect, Value Premium, Reversal Behavior and so on, has brought a serial of revolutions to the field of financial asset pricing theory. On one hand, the researchers seek for improving the theoretical model, on the other hand, they try to have empirical research to test the model. Traditional methods of modelling returns and testing the Capital Asset Pricing Model (CAPM) do so at the mean of the conditional distribution. The different empirical results bring a lot of debate.In this paper, on the basis of the classic theory, we test how the risk factor influence the stock return holds at other points of the distribution by utilizing the technique of quantile regression (Koenker and Bassett 1978). This method allows us to model the performance of firms or portfolios that underperform or overperform in the sense that the conditional mean under-or overpredicts the firm's return. From the empirical analysis of the panel data of 150 listed companies in Shanghai and Shenzhen from 2005,1-2007,12, we find that the market price of beta risk is significant and negative over the conditional distribution of returns, and this relationship is more obvious in high quantile. the firms that overperform have a strong negative relationship between beta risk and return, this is opposite to the CAPM. The empirical evidence also show that underperforming firms exhibit a negative relationship between size and returns but positive for firms that overperform. This paper also analyzes the relationship between the firm-specific risk factors and the stock market, including book to market value, net income to price, liquidity, there is some evidence that the relationship is changing over the distribution. The quantile regression methodology enables an entirely fresh perspective on the standard types of relationships studied in this literature. Indeed, we uncover several results that are at odds with past empirical studies that focused exclusively on the mean, and appropriate way of analyzing the returns. One of the implications of the failure of the interquantile tests to find parameter homogeneity across quartiles is that the multifactor conditional CAPM considered here is rejected, since the prices of risk factors, including beta risk, systematically change across the conditional distribution of returns. A more profound implication is that we open the door for further theoretical exploration of why such risk factors should be priced differently, depending on the firm received good, bad or indifferent news during the sample period.