本研究探討其流動性是否會影響其ETF定價誤差的問題，透過Karolyi, Lee, and van Dijk (2012)所提出對於流動性指標的估計，計算出台灣50與中型100 ETF兩檔指數股票型基金在2008年到2016年的流動性，並將流動性導入其VECM-GARCH與TVECM-GARCH模型中，探討當兩檔ETF受到其流動性影響而產生定價誤差時，其市價與淨值對於前期價格的誤差調整的方向以及幅度。
在迴歸模型中可以發現，流動性確實影響其ETF定價誤差之現象，並由VECM-GARCH以及TVECM-GARCH模型中發現，若不考慮流動性較佳或是不足的情況下，當ETF產生其定價誤差時，台灣50 ETF與中型100 ETF在透過市價或淨值調整其偏離價格的方向與幅度呈現不一致的情況。若考慮其流動性門檻值時，在流動性較好的情況下或流動性不足的情況下，台灣50 ETF與中型100 ETF市價與淨值在對其偏離價格做調整時，亦呈現不同的調整情況，代表當流動性處於不同水準的情況下，其指數股票型基金市價與淨值分別調整的方向以及幅度會呈現不同的結果。
In this thesis, we investigate the liquidity of underlying stock whether affected the pricing efficiency of ETF. We use indicator of liquidity which proposed from Karolyi, Lee, and van Dijk (2012) to calculate two kinds of ETF at 2008 to 2016, respectively Taiwan 50 index and Mid Cap 100. And driven liquidity into the model of VECM-GARCH and TVECM-GARCH to discuss when the liquidity of two kinds of ETF affects the pricing error, the market price and net asset value will adjust the direction and range for previous price.
We can finding the liquidity of underlying stock indeed affected the pricing efficiency of ETF in regression model. On the other hand, we also finding the model of VECM-GARCH and TVECM-GARCH that we do not consider the liquidity its better or worse. Taiwan 50 ETF and Mid Cap 100 ETF will adjust the price deviation differently by net price and market price. Otherwise, if consider the threshold of liquidity, Taiwan 50 ETF and Mid Cap 100 ETF also adjust the price deviation differently, its consistent with
our anticipation that there are exist different liquidity level, the market price and net price of ETF adjust the price deviation will have different direction and range.
