Testing For Linear And Nonlinear Granger Causality Between Stock Price And Warrant Price In China

On the basis of reviewing the history and analyzing mechanism of warrant market in China, theoretical prices of all 29 expired warrants are calculated and then comparison with their actual prices using Black-Scholes pricing model, Binominal-Tree pricing model and historical volatility. We find all call warrants' actual prices are very close to their theoretical prices except Baosteel JTP1, which is the first warrant issuing in China warrant market.Then, Linear and nonlinear Granger causality tests are used to examine the dynamic relation between daily warrants returns and their underlying stocks returns. Because warrant price is reprehensive as a nonlinear function of high-order of stock price, the probability that there is nonlinear causality between warrants and stocks returns is larger than the probability that there is just linear causality between warrants and stocks returns, so this paper focus on the Baek and Brock nonlinear Granger causality test which is used nonparametric method and modified by Hiemstra and Jones(1994). We find significant nonlinear causality between two returns. After controlling for ARCH effect in returns, we continue to find causality from most underlying stocks returns to warrants returns. This results show the evidence that stock and warrant seem not to be segmentation market and the information flow between two markets is efficient. But we find there are only one third warrants have causality from warrant to stock returns. Most warrants haven't played the full role of prices stability and risk hedge because of restricts on their selves. Government should take those effects into account when developing police, as well as traders when investing. This paper proposes a policy suggestion of developing a real stock option market in the end.