Font Size: a A A

Optimal Statistical Arbitrage Model

Posted on:2016-05-28Degree:MasterType:Thesis
Country:ChinaCandidate:X LiuFull Text:PDF
GTID:2309330476453578Subject:Applied Statistics
Abstract/Summary:PDF Full Text Request
Review of China’s security market, around the year of 2007, it has already experienced the largest wave of boom and slump. Shanghai Stock Index has been rising from 998.23 to 6124.04 point, then quickly fell to 1664.93 point. Since 2010, Short-Mechanism including securities margin trading and stock index futures started to launch, which provide investors a two-way trading function. These short-tools effectively promote the application of statistical arbitrage strategy in China’s security market. Currently, at the end of 2014, the performance of security market becomes the most popular topic. Just in November 28 th, the record-breaking volume in SH&SZ has reached 710.4 billion, which is the highest around the world. Quantitative investments including statistical arbitrage strategy are involving in this market more than ever.This paper first summarized statistical arbitrage related concepts and basic theories of economics and introduced pairs trading strategy based on cointegration method. Then we focus on the study of optimal timing of arbitrage, with 2 models adopted for analysis, including revenue function maximized model based on OU process and trading signal updating model. Meanwhile, this paper presents a new model, trading signal real-time update model, which actually is an extension of the cointegration model.In the empirical analysis part, we select CSI 300 index ETF and CSI 300 index future for pairs trading, which are of high liquidity and high correlation. Select 1100 trading days’ data from June 3rd, 2010 to December 16 th, 2014 as sample. Apply the OU model on the whole sample to maximize revenue function and obtain optimal trading timing. By using first 100 data to verify that the model earns more in short time period. During the analysis, separately set up the continuous and discrete rules. The outcome shows that the discrete rule receives higher earnings. At last, apply the trading signal updating model and compared with the OU model. The characteristics of the two models are listed, which offers appropriate choice to investors.
Keywords/Search Tags:Statistical Arbitrage, Pairs Trading, O-U process, Trading Signal Updating Model, Optimal Trade Timing
PDF Full Text Request
Related items