The Study, Based On The System Of Financial Time Series Average Portfolio Trading Strategies

This paper studies stock index system trading strategy based on financial time series average line combination, through completeness search to the average length of the parameter space in the different market environment, try to find repetitive and stable profitable trading system strategy.In this paper, we chose two kinds of technical indicators as a reserch object,one is moving average (MA) to reflect the change of price,and the other one is volume ratio(VR) to reflect che change of dealing volume.Based on these indicators,Four kinds of trading strategies were studied and discussed,which is dual-line cross strategy, dual-line neutral zone strategy, triple-line cross strategy and volume ratio strategy.Research shows that in the use historical data carries on the parameter optimal process to four kind of trading strategies,and these four kind of trading strategy can obtain a higher income compared to buy-and-hold strategy,but in the use of optimized parameters carries on the forecasting test, dual-line cross strategy and dual-line neutral zone strategy shown it's stability and get a satisfactory result in all circumstances ,but other strategies are unable to provide valuable instruction. At the same time, dual-line moving average strategy also has a good stability and convergence,these are very useful characteristics for using index transaction instruct stock transaction.In the statistical test, the dual-line strategy also significantly reject the null hypothesis. and shown it can help investors to obtain the higher income than buy-hold strategy.the other strategies is unable to shown that its strategy has remarkable difference with buy-and-hold strategy.Based on technical indicators system trading strategy, could help the investor seeks to has the statistical significance part, sends out the explicit trading signal, help the investor obtains the long-term stability the incomes in highly stochastic or the non-stochastic fluctuant market.but on the other hand, these kinds of strategy also has some flaws,such as the low rate of success and difficult to determine the optimal parameters.This research work has three remarkable innovations: First: In the time selection, using different time windows for the whole sample simulation tests,include of the complete bull-bear market, the bull market and the bear market. inspects the different strategy's profit making, the predictability and the convergence of parameters.Second: In the space, no longer limit to the standard commercial indicators default parameter, establishes the development data test platform voluntarily, carries on the mesh-point to the entire parameter space to exhaustive search, inspects its parameter performance.Third: In the sample selection, the use of 27 kinds of sector index data to carry on the testing , the index data has good stability, but also the most difficult being manipulated,and this article also take into account the index's other properties.