金融衍生证券的人工神经网络定价方法研究进展评述  

Review on Artificial Neural Network Valuation Methods for Pricing Financial Derivative Securities

作　　者：刘凤琴[1] 马俊海[2] 

机构地区：[1]浙江财经学院信息学院,浙江杭州310018 [2]浙江财经学院金融学院,浙江杭州310018

出　　处：《财经论丛》2008年第3期47-53,共7页Collected Essays on Finance and Economics

基　　金：国家自然科学基金资助项目(70571068)

摘　　要：近些年来,金融衍生证券的人工神经网络定价方法已经得到学术研究领域的高度关注和实际问题中的广泛应用。本文主要对国内外在这一研究领域所开展的主要工作及成果进行分析与评述;在此基础上,提出该研究领域的进一步研究方向。结论认为,非参数化的神经网络方法将成为解决金融衍生证券定价问题的重要途径;充分融合参数化定价方法的有用信息,将成为未来该领域研究的重要思路与方式。In recent years, as an important method for pricing financial derivative securities, artificial neural network methods have drawn more attention in financial academic field and wide application in practice. In the paper, first we make a systematic review on main research work in the field domestic and abroad. On the basis of the review, we propose the direction for further research. The conclusion is that nonparametric artificial neural networks will become an important way for pricing financial derivative securities, which will be an inevitable trend to make full use of information from parametric valuation models.

关 键 词：金融衍生证券 非参数化定价模型 人工神经网络 隐合波动率 

分 类 号：F830.91[经济管理—金融学]