Font Size: a A A

Studies On Futures Statistical Arbitrage

Posted on:2015-02-10Degree:MasterType:Thesis
Country:ChinaCandidate:H B SongFull Text:PDF
GTID:2309330431978867Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
China futures market has entered the fast track of development in recentyears.Trading volume and turnover has increased significantly.The number of futuresvarieties are beyond40. The futures market is a high-risk and high-yield market, itbecame a natural arbitrage trading platform. The world’s only glass futures listed onDecember3,2012.From January to July in2013, both trading volume and turnover wasat the top of Zhengzhou Commodity Exchange which has seen some unusual levels ofactivity and it attracts money to invest in it for profit.Statistical arbitrage is a kind of arbitrage method using asset price historical rule,which is based on the statistical method for arbitrage opportunities.It identifies theinternal unreasonable pricing relationship through the establishment of financial assetprice time series model,and seeks for the existing arbitrage opportunities between twoor more assets.It executes the instruction to obtain the profits according to certaintrading strategy.This paper mainly use the knowledge of statistical arbitrage to glass futures.Thispaper firstly introduces the research status of domestic and foreign statistical arbitrage,then introduces the definition, classification of statistical arbitrage and related theorymodel used in need, and then use these arbitrage method to carry on the empirical studyon glass futures.This paper uses the constant proportion and different proportion arbitrage in theempirical study.The constant proportion arbitrage include common cointegrationarbitrage and cointegration arbitrage based on GARCH. The different proportionarbitrage uses Calman filter model arbitrage. They have superiority over the generalspread arbitrage and ratio arbitrage and achieved good results in the historical data.Inorder to test whether the profitability of model is persistent, this paper uses threedifferent extrapolation method respectively.Finally, this paper evaluates the performanceof three models by the scientific method.The yield of three models in the extrapolateddata has decline sharply.It also shows that the models have a common weakness: poorpredictability.The whole paper contain academic research and practice.It gives the majority of investors a set of futures investment approach and has a strong reference and practicalsignificance.
Keywords/Search Tags:Statistical Arbitrage, Cointegration Arbitrage, GARCH Model, Calman Filter
PDF Full Text Request
Related items