本研究利用多元迴歸共變異數模型分析標的股票初次發行認股權證對股價波動性的影響，並配合與標的股票同產業與公司規模、變異數相近似的股票「控制樣本」，來比較認股權證發行與股票控制樣本之間對標的股的不同影響，進行分析認股權證的發行是否會影響股價。

研究取樣 1986~1996 年香港聯交所初次發行認股權證之恆生指數成分股標的上市公司股票與 1994~1998 年境外台股認股權證與境內台灣證交所初次發行上市掛牌之認購權證股票為研究標的。香港是目前全球認股權證交易罪活絡的市場，因此，香港及靜內外台灣認股權證的資料可供實證研究，探討有關衍生性商品發行後對標的股的波動性有無增減，並提供投資人與政府政策決定的重要參考依據。本研究之結論主要如下：

1.台灣與香港，初次發行認購權證標的股票與控制樣本股票總變異比率及其可解釋變數之間兩者的不同關係，僅在較低的信賴水準下之發行初期有所不同，亦即以樣本全部時期來看，權證發行長期對股票波動性並無顯著的改變。

2.境外台灣認股權證的發行增加了短期波動性，可能是在權證發行初期，權證市場私人交易部份外溢到現貨市場，以及國內投資人對權證的發行過渡反應，使得市場雜訊增加，因而造成短期波動增加。

3.認股（售）權證的發行對於股票變異數的影響較一般以往所想像來的弱效果。僅接受發行初期股票變異數在香港樣本與台灣權證上市後境期樣本有降低股票變異數的情形。以樣本涵蓋的全部時期而言，權證初次發行對於股票波動性並無顯著的增加或減少。

This research will analyze the impact on the price volatility of the underlying stocks of the first initial issuance of warrants by means of muliple regression analysis. The analysis will contruct control samples of stocks that share the similar characteristics of the underlying stocks in terms of thier industry, size and variance(or volatility). By analyzing the impact of the warrant issuance on the underlying stocks and comparing it with the price volatility of the control samples, whether or not there is change in volatility due to the warrant issuance can then be determined.

Our research will be conducted from a warrant data set collected from the Hong-Kong Stock Exchange and the recently issued warrants on Taiwan stocks as the underlying assets. The H.K. warrant market is the most active and liquid warrant market in the world. Therefore, the analysis based on the H.K. warrant as well as Taiwan warrants is deserved to carry out, and the empirical evidence of this study will have a policy implication regarding the price volatility of the underlying stocks after the introduction of warrants.

We find that the behavior of Taiwan and Hong Kong Warranted stocks' variances is not affected by Warrant introdution, except for the earliest introductions at a low level of confidence. We find a significant impact on residual variances, but only for the earliest introductions.

There finding differ both from those of previus studies, which find warrant introduction to be associated with decreases in stock variances, as well as from other more recent studies that find no effect.

Two factors account for our different conclusions. One is that the effects have changed over time; that is the conclusion reached depends in part upon the period tested. The second is that we control directly for changes in marketwide variances and for maean reversion in stock variances. We account indirectly for other systematic factors that may affect stock variances by incorporation a carefully matched control sample of stock.