本文使用向量自我回歸 (vector autoregressive, VAR)模型架構下的一般化預測誤差變異數分解 (forecast error variance decomposition), 來進行十二個已開發國家和八個新興國家之間, 於2001年至2015年間股市流動性關聯程度的靜態與動態分析。其中, 此方法並不會因為變數不同的排列而產生不同的結果。本文所使用的流動性資料是由各國股市的每日價格指數和每日交易金額所建構, 並使用修改過後的 Amihud不流動性指標計算而來。結果顯示, 當發生經濟衰退時, 股市流動性的關聯性似乎會變得更加緊密。而且, 已開發國家股市中的外溢效果和關聯程度, 比起新興國家都還要來得更大、更強。同時我們也發現一國的地理特徵與經濟發展是決定與他國關聯程度大小的重要因素。值得注意的是此篇文章的幾個實證結果在某種程度上與一些資產關聯性的研究結果一致。;Based on the generalized vector autoregressive framework in which forecast error variance decompositions are invariant to variable ordering, we provide both static (full-sample) and dynamic (rolling-sample) analyses for the liquidity connectedness in 12 developed and 8 emerging stock markets during 2001-2015 period. The liquidity data we use in this paper is constructed from daily price index and turnover and it is calculated by the revised Amihud illiquidity ratio. Our empirical results indicate that the liquidity connectedness seems to be more intensive when economic downturns happen. What’s more, stock markets in developed countries generate higher spillovers and connectedness than stock markets in emerging countries. We also prove that the geographical features and economic developments of a country are crucial factors deciding its level of connectedness to others. It is noteworthy that some empirical outcomes in this paper is consistent with some equity connectedness studies at a certain extent.
