本文分析台灣障礙型權證的價格行為，並估計券商發行障礙型權證時所採行的波動性加成幅度。實證結果顯示實務上常用的Rubinstein和Reiner(1991)障礙選擇權定價公式解對台灣障礙型權證有顯著的評價誤差；相對而言，較具彈性的馬可夫鏈選擇權評價演算法得以使節點恰與關卡價格重合，確實呈現較佳的評價績效。然而馬可夫鏈選擇權評價演算法仍然顯著低估障礙型權證的市價，除了權證的價內外程度、距到期日時間與權證市場的流動性會影響演算法的準確性之外，權證發行商將歷史波動性加成計算，藉以提高障礙型權證價格，也是原因之一。本文進一步以馬可夫鏈選擇權評價演算法所估計的理論價格為基準，推估出台灣障礙型權證的波動性加成幅度約在10%至25%之間。
This study analyzes the price behavior of barrier warrants, and also estimates the level of volatility markup while the firm issuing the barrier warrants. Both the up-and-out call warrants and the down-and-out put warrants are selected to be our research samples. The results show that the analytical barrier option pricing method of Rubinstein and Reiner (1991) has significant pricing errors. With its flexibility of combining the notes and the trigger price, Markov chain option pricing algorithm presents better pricing performance. However, Markov chain option pricing algorithm still significantly underprices the barrier warrants, and the explanatory factors include monyness, maturity, liquidity of the warrant, and also the volatility markup. Therefore, the study takes the theoretic price induced from Markov chain option pricing algorithm as a benchmark, and estimates that the volatility markup is around 10% to 25%.
