資訊管理學報
陳安斌;張志良;
頁: 63-80
日期: 2001/01
摘要: 類神經網路具有學習與高速計算之能力，再加上非線性處理與容錯之特性，使其在行為預測上表現相當優異，雖過去曾有多篇文獻使用類神經網路對選擇權進行評價，但至今尚未見其在避險上之運用，本研究運用基因演算法自動演化之類神經網路，掌握特定認購權證之時間價值與避險比例行為，以進行價格預測與避險模擬。實證結果顯示，以類神經網路為基礎之方法，針對台灣已到期之十五檔認購權證，不論在評價上之解釋能力與誤差程度，或在避險上之風險暴露與獲利均優於BS模型，即表示在台灣認購權證市場中，基因演算法自動演化之類神經網路能提供一個比BS模型更能精準評價，以及更有效率避險的模型。
關鍵字: 選擇權;基因演算法;類神經網路;評價;避險;
Abstract: Neural networks have the ability of learning and performing high-speed calculations, also with nonlinear processing and tolerance of faults, its prediction faculty becomes quite outstanding. Although most literature is available on options pricing via neutral networks, little attention has been paid to hedging. This study applies the genetic adaptive neural network to the pricing and hedging of warrants via utilizing the pattern of specific warrants time value and 'Delta' behavior. The empirical results indicate that the method based on neural networks excels the BS model in interpretive capability and error degrees on pricing, risk exposure and profits on hedging. It means that in the Taiwanese warrant market, the proposed model can provide a more accurate pricing and efficient hedging model than the BS model.
Keywords: Option;Genetic algorithm;Neural network;Pricing;Hedging;
瀏覽次數: 8133 下載次數: 35
引用 導入Endnote
頁: 63-80
日期: 2001/01
摘要: 類神經網路具有學習與高速計算之能力，再加上非線性處理與容錯之特性，使其在行為預測上表現相當優異，雖過去曾有多篇文獻使用類神經網路對選擇權進行評價，但至今尚未見其在避險上之運用，本研究運用基因演算法自動演化之類神經網路，掌握特定認購權證之時間價值與避險比例行為，以進行價格預測與避險模擬。實證結果顯示，以類神經網路為基礎之方法，針對台灣已到期之十五檔認購權證，不論在評價上之解釋能力與誤差程度，或在避險上之風險暴露與獲利均優於BS模型，即表示在台灣認購權證市場中，基因演算法自動演化之類神經網路能提供一個比BS模型更能精準評價，以及更有效率避險的模型。
關鍵字: 選擇權;基因演算法;類神經網路;評價;避險;
A Genetic Adaptive Neural Network Approach to Options Pricing and Hedging: Analysis and Evidence
Abstract: Neural networks have the ability of learning and performing high-speed calculations, also with nonlinear processing and tolerance of faults, its prediction faculty becomes quite outstanding. Although most literature is available on options pricing via neutral networks, little attention has been paid to hedging. This study applies the genetic adaptive neural network to the pricing and hedging of warrants via utilizing the pattern of specific warrants time value and 'Delta' behavior. The empirical results indicate that the method based on neural networks excels the BS model in interpretive capability and error degrees on pricing, risk exposure and profits on hedging. It means that in the Taiwanese warrant market, the proposed model can provide a more accurate pricing and efficient hedging model than the BS model.
Keywords: Option;Genetic algorithm;Neural network;Pricing;Hedging;
瀏覽次數: 8133 下載次數: 35
引用 導入Endnote