Full metadata record
DC FieldValueLanguage
dc.contributor.author陳怡和en_US
dc.contributor.authorChen, Peteren_US
dc.contributor.author陳安斌en_US
dc.contributor.authorAn-Pin Chenen_US
dc.date.accessioned2014-12-12T02:17:21Z-
dc.date.available2014-12-12T02:17:21Z-
dc.date.issued1996en_US
dc.identifier.urihttp://140.113.39.130/cdrfb3/record/nctu/#NT850396021en_US
dc.identifier.urihttp://hdl.handle.net/11536/61852-
dc.description.abstract選擇權是財務金融界中最為活躍的金融工具之一。在70年代以前,人 們沒有一種適切的方法來評估選擇權價值。直到1973年,Fisher Black和 Myron Scholes兩位學者導出了Black- Scholes評價模式,提供了一種便 捷的評價方式,選擇權市場也因此而日漸蓬勃。但是這些都是建立在許多 假設下的統計模式,在面臨實際應用上,出現評價模式價格與實際價格不 一的現象。 然而在近年來多人在人工智慧的研究上已發現,類神經網 路具有學習和高速計算之能力,更由於類神經網路之非線性處理和容錯的 能力,使其預測能力相當優異。故本研究將嘗試運用類神經網路所掌握到 的某特定選擇權之定價行為﹐進行價格預測與分析之行為比對。 本研 究針對芝加哥商業交易所之德國馬克外匯選擇權市場進行實證研究,嘗試 利用基因演算法自動演化之類神經網路來架構一新的選擇權評價模式,與 傳統之Black-Scholes評價模式相比較,並評估這兩種方法對市場價格的 誤差程度與解釋能力。研究期間涵蓋1990年至1992年,共計3年。 研 究結果顯示,類神經網路評價模式不論在誤差程度、變動程度或解釋能力 上都優於Black-Scholes評價模式。表示在德國馬克外匯選擇權市場中, 基因演算法自動演化之類神經網路能夠提供一個比Black-Scholes評價模 式更接近市價且更穩定的選擇權價格之評價模式。 Options is one of the most active financial tools of the financial management fields. However, there is not a properly evaluation method for options prices. Until 1973, Fisher Black and Myron Scholes derive the Black-Scholes pricing models and provide an adequate evaluation method for the options. The options markets therefore develop more and more rapid. But, their model was established under many assumption and deduce the difference between the practical price and theory price. In these years, many researches in AI areas found that Neural Networks pertain excellent learning and computing capabilities. Furthermore, the nonlinear process and fault-tolerance characteristic make the Neural Network method contain excellent forecast capabilities. This research applies the Neural Network on the pricing of specific options and then compares the result with the traditional method. Here, Deutsche mark options traded in CME are chosen as the foreign currency options contracts. The te The result displays whether in error degrees, variant degrees and interpret capability, the Neural Network method is better than the Black-Scholes pricing model. Meanwhile, in Deutsch mark currency options market, a new pricing model using the Neural Network method is a better and more stable than the traditional Black-Scholes pricing model.zh_TW
dc.language.isozh_TWen_US
dc.subject外匯選擇權zh_TW
dc.subject實證研究zh_TW
dc.subjectBlack-Scholes評價模式zh_TW
dc.subject基因演算法zh_TW
dc.subject類神經網路zh_TW
dc.subjectCurrency Optionen_US
dc.subjectEmpirical Studyen_US
dc.subjectBlack-Scholes Pricing Modelen_US
dc.subjectGenetic Algorithmen_US
dc.subjectNeural Networken_US
dc.title運用類神經網路在外匯選擇權評價模式之實證研究zh_TW
dc.titleAn Empirical Study of the Foreign Currency Options Pricing Model by Using Artificial Neural Networksen_US
dc.typeThesisen_US
dc.contributor.department資訊管理研究所zh_TW
Appears in Collections:Thesis