Abstract: This study investigates the price discovery function in Shanghai 50 index markets: the index futures, ETF and ETF options markets. Put-call parity is applied to calculate the implied spot price of ETF options. Using 5 minutes high frequency matched synchronous data, Granger causality tests, generalized impulse response, Johannsen cointegration tests and vector error correction model are employed to analyse the lead-lag relationship of spots and derivatives. The results indicate that the three price series are a cointegrated system with one long-run stochastic trend, and stock index futures price leads other market at least 5 minutes. When the contribution of price discovery is calculated by using the generalized information sharing model and permanent transitory model, it is found that the futures market serves the dominant price discovery function in the long term. However, the ETF market serves the dominant during the boom period, and the ETF options market serves the dominant during the crash period.

Key words: price discovery, put call parity, vector error correction model, generalized information share model

摘要： 比较基于上证50指数的股指期货、ETF期权与现货ETF市场的价格发现能力,选取5分钟高频数据进行实证分析,并将暴涨暴跌行情与全样本区间进行了对比分析。首先,采用买权卖权等价理论反推期权价格隐含的现货价格;其次,运用向量误差修正模型,结合广义脉冲响应函数等分析方法研究市场间价格的领先滞后关系;最后,运用广义信息共享模型量化各个市场的价格发现贡献度。结果表明:在不同区间中,期货市场均领先其他市场至少5分钟;从长期来看,期货在价格发现中的贡献度最大,期权次之;在暴涨区间中,ETF的价格发现贡献度最大,期货次之;在暴跌区间中,期权的价格发现贡献度最大,期货次之。

关键词: 价格发现, 买权卖权等价理论, 向量误差修正模型, 广义信息共享模型