The Application Of Bootstrap In Moving Average Analysis Strategy

One of the major differences between the modern capital market theory andpractice of financial invest is the contradiction between the efficient markethypothesis and technical analysis. But, since the1990s, a lot of researches provideevidence that technical analysis is efficient. One of the reasons is that the earlierstudies usually consider price changes in linear mode, but the recent researchesconsider the dynamic process of the real rate of return. So how do you portray andcapture yield is a critical issue. Bootstrap methodology is developed in recent yearsand widely used statistical methods that the most superiority is that you don notneed to assume the distribution and do not need to get the analytic formula inadvance. For the Chinese market, the validity and applicability of the Bootstraptechnology is essential. And that is the precondition to use in stock. VAR is a goodobject to test. Improve the prediction accuracy of the VAR lies in the ability toeffectively capture the distribution of the left tail characteristics. But traditionalmethod is proved to be having a problem because of fat tail of the returns. If we canprove that result based on the Bootstrap technology is better than traditionaltechnology, we can prove that Bootstrap technology captures the characteristics ofthe unknown distribution tail better. That also strongly supports the features ofBootstrap technology in practice.In addition, the moving average of technical analysis is used widely by investor.However, traditional T test is unreasonable in returns, because this method does nottake the characteristics of "fat tail" in account. So, the result from that isquestionable. At the same time, many researches point technical analysis is useful.We compare the result of VAR from Bootstrap methodology and traditionalmethod in three markets, which are bull market, bear market, and neither of them.We get the conclusion that Bootstrap methodology achieves the better result thantraditional method. And then we combine Bootstrap methodology and movingaverage to search the trading portfolio which can get the most returns in differentholding period for three Short-term trading strategies. At last we compute these trading portfolios according to real data from1Nov2012to19Feb2013. Comparedwith the real result, we find the trading portfolio getting the most returns based onBootstrap technical is consistent with the test sample. And now, we can say we canuse Bootstrap technical in moving average to search the trading portfolio to helpinvestor to make a strategic decision.