Research On Application Of Quantitative Trading Strategy Model

With the electronic trading platform in the worldwide stock market graduallymature, quantitative trading becomes prevalent in worldwidestock market.Quantitative trading is widely used in operation of investment institutions,abroad.HFT,rapid rise in recent years but controversial,belongs to emerging ofquantitative trading way. Because of it’s characters, such asprogramming trading,huge volume and yield of extreme position time, and low profitable in one trade buthigh profitable in total revenue, HTF has already become the focus in the field ofinvestment researchdomestically and internationally. The mushrooming of securitiesmarket in China,however, isrelatively late. The electronic trading platformsimprovegradually, and quantitative trading platform is still in the development stage.Meanwhile, T+0transaction in China will restart shortly, will be to our country inthis play a major role in the promotion of trade, which would dramatically affect thepopularization of quantitative trading.Therefore, the quantitative trading modelresearch on Chinese securities market has significant practical implications as well asprovides further research directions for marketing investment research. The object ofthis paper is trading strategy based on quantitative model. The data is chosen from sixsample stocks and one predictable stock, which involves in transaction details andquantitative transaction detailsthat including third-hand before clinch a deal the price,volume, sales amount, transaction time, quantitative trading volume, and other stockmarket data. The purpose of the paper is to understand the quantitative tradingstrategy and to make trading decision by using strategy model. In accordance withprevious research on stock investment strategy, this article first analyzes therelationship between the fluctuation of stock price and the large investment strategy.The article finds some dependent relationship between quantitative trading strategyand fluctuation of price. Secondly, to further explore the relationship between thedependent model, the article uses the cumulative probit model to fit the sample data, and use the fitted model to forecast stock strategy analysis.By doing this procedure,Ifind that1)Spot variables have more significant affect on investment decision thanprophase variables2) quantitative trading strategy prefer to make synthetic deal withmarket price changes,3) quantitative trading strategies prefer to purchase stock underactive stock market with high volume and sell stock under stock market with lowvolume.4) buying are cancelled and selling are cancelled affect quantitative tradingstrategy alternation, the current high volume sold deity tend to make quantitativetrading strategy to sell trading makes the next phase of quantitative trading strategy,tends to buying, quantitative trading strategy is affected by the changes of buy andsell pending.5) the implementation of quantitative trading strategy for the share pricehas strict requirements. I predict the predictable set and observe the rate of error. Then,to explore different ways to model the impact on the strategy decision the article usesBayesian network and support vector machine (SVM) model to predict the predictionset and start the model training. By compared with cumulative probit model, I findthat each of these three model has their own advantages. For example, SVM is betterunder prediction precisely criterion. Under explanation criterion, cumulative probitmodel has better performance. Bayesian network is located between two models.Finally, I find strong relationship between creating strategy and share price byanalysis above. The paper tries to fit the stock price and make prediction bymultivariable time series model, and find that VARX model and state space modelhave great prediction performance for stock price. Finally, combine with thecharacteristics of quantitative trading, Isupply several reasonable suggestions aboutmonitoring the new trading way in Chinese securities market.