Quantitive Timing Strategy Based On Relationship Of Price And Volume After Wavelet Packet Noise Reduction

The variables of price and volume,in theory of technical analysis,are the most basic elements in describing the characteristics of income and risk of financial assets,and meanwhile they have the advantages of easy access and monitoring.We assume that "market behavior covers everything",then we can build a timing strategy based on the analysis of the relationship between price and the trading volume of the stock market.However,price and volume are influenced by contingency factors as time series data.Usually they have the features of low SNR,unstability and nonlinearity,which have a negative influence on further analysis.Therefore,the first task is to minimize the negative impact data before analyzing the relationship and constructing the timing strategy.But traditional methods' performances of noise reduction are not satisfactory.By contrast,wavelet transform has a trait of muti-resolution,which makes it very suitable for non-stationary signal processing.As a result,wavelet is introduced to economic and financial field by many researchers,to analyze and quantify phenomenon in economic and financial fields.In order to find a better choice of denoising and better solution of investment strategy,this paper was guided by the logical thought of the problem finding-problem analysis-problem solving.First of all,taking the relationship between price and volume in Chinese A share market as research object,this paper calculates the year-on-year ratio in order to study the relationship between them from a cyclical perspective.Secondly,in consideration of the limits of traditional methods in time series noise reduction,this paper choose to use wavelet packet nonlinear thresholding denoising to reduce noise impact.Through wavelet packet decomposition,optimal tree calculating,threshold setting and wavelet packet reconstruction,bad influence of noises on study is downsized on the premise of retaining original characteristics of the signal.At last,we build a timing strategy based on the relationship between volume and price that have been studied in previous chapter of this paper and carry out a simulation transaction.We proved the superiority of the strategy by comparing the performance of the strategy before and after the noise reduction.All analysis in this paper leads to several conclusions which will be told as follows: after wavelet packet denoising,price and trading volume has a long-term equilibriumrelationship on monthly basis;and volume is the Granger reason of price changes,but price is not Granger reason of volume changes;when the year-on-year ratio of volume gets bigger and is greater than zero possibility of price rising is relatively lager,namely64.23%;Compared with the strategy before noise reduction,strategy after noise reduction apparently has a better performance in SHARP ratio,the maximum drawdown and the annual yield,especially the maximum drawdown.So far it also proves the effectiveness of quantified investment strategy based on volume and price relationship that has been built in this paper and the availability of wavelet in financial data noise reduction.