本文首先进行传统技术与Bootstrap技术在三种类型市场(牛市、熊市、非牛市非熊市)的VaR度量上的比较分析。其中,传统技术为对收益率做标准正态分布和t分布的两种假设情况,从实证角度证实了Bootstrap在VaR度量上优于基于传统统计分布的检验技术。在此基础上,主要针对3种短期移动平均交易策略分别在不同持有期下结合Bootstrap技术,以比较各种交易策略的收益情况,从而搜索出收益最高的交易组合。最后,根据近期的真实数据结果在上述交易策略下做检验,并证实由基于Bootstrap技术得到的交易策略与真实数据计算结果所得到的交易策略一致。从而证实了结合Bootstrap技术的移动平均交易策略,能够搜索出收益最高的交易策略。关键词:VaR;尖峰厚尾;Bootstrap技术;技术分析;移动平均; 论文类型:应用研究 Abstract One of themajor differences between themodem capital market theory and practice off'mancial invest isthe contradiction between theefficientmarket hypothesis and technical ,since the 990s,a lotofresearchesprovide evidence thattechnical analysis ofthe reasons isthattheearlier studies usually consider pricechanges inlinearmode,but the recent researches consider thedynamic process ofthereal how doyou portray and capture yield is methodology isdeveloped in recent years andwidely used statisticalmethods thatthemost superiority isthatyou don notneed toassume thedistribution and do not need toget theanalytic formula inadvance. Forthe Chinese markeL thevalidity andapplicability oftheBootstrap technology is istheprecondition tO use is agoodobject totest. Improve theprediction accuracy ofthe VaR liesintheability toeffectively capture thedistribution ofthe traditionalmethod isproved tobe having aproblem because that resultbased ontheBootstrap technology isbeaer than traditionaltechnology,we canprove that Bootstrap technology captures thecharacteristics ofthe unknown distribution tail alsostrongly supports the features ofBootstrap technology inpractice. Inaddition,themoving average oftechnical analysis isusedwidely by investor. However,traditional T testisunreasonable inreturns,because thismethod does not take the characteristics of”fat tail”in result from that is the sanle time,many researches point technical analy
