Integrated Variance Estimation for Assets Traded in Multiple Venues
利用同一资产在多个交易场所的价格序列，通过向量误差修正模型估计有效价格，进而计算日内收益平方和得到无偏且高效的综合方差估计量，适用于期货与现货、交叉上市股票等场景。
In this paper, we show that the availability of multiple price series for the same asset can be exploited to estimate its integrated variance. We use a vector error correction model for those prices and its common trend representation to estimate the efficient price of the asset. Because the estimated efficient price satisfies the martingale property, we can then compute the sum of its squared intra-day returns to obtain an unbiased and efficient realized variance estimator. We derive the asymptotic distribution of the proposed estimator and provide extensive simulation experiments. An application to spot and futures prices as well as cross-listed stocks shows that this approach outperforms univariate estimators if at least two venues contribute substantially to the price discovery process.