| Since the beginning of the last decade of 20th century, in less than 20 years'period, China's capital markets have made a successful progress which striked the attention of the world. China's capital markets are mainly based on stock markets, which successfully optimized the allocation of resources. But there are two questions which blocked the further development of China's stock market:the one is equity division problem, the other is lack of short sale mechanism. The former problem was solved by Share-merger reform, the latter one was solved by margin mechanism and stock index future of Hushen-300 in the first half year of 2010. The short sale mechanism may bring forth a deep reform on China's stock markets, the profit pattern which have lasted for nearly 20 years will take a change ever since.The building up of stock index futures will bring the quickly application of quantitative investment and statistical arbitrage in China's stock markets. Quantitative investment is based on historical data to find some patterns of asset return which can be used to invest. Statistical arbitrage is one of quantitative investment method which take hedging into account. The difference between statistical arbitrage and riskless arbitrage is, there would have risk when using statistical arbitrage while using riskless arbitrage to invest there was no risk at all. In reality, riskless arbitrage opportunities are seldom, while statistical arbitrage opportunities may be found more easily relatively, especially in the immature markets of China.This paper studied a statistical arbitrage strategy-of-multi-factor model based on excess return. First I make a review of past asset pricing theory——the CAPM model, the APT model, the multi-factor model and efficient market hypothesis, then introduced the related theories of statistical arbitrage, and latter I constructed a excess return model which based on return-risk investment utility function. I also did an empirical work on the excess return model which I constructed before, used the constituent stocks'data of Hushen-300 index from 2005 to the first half year of 2010. Through the former work, I find an optimum forecasting excess asset return multi-factor model under the framework of this paper, and did a robust test on the model which added the time effect in. latter I use the optimum excess return forecasting model to filter stocks and structured stock groups which were sort by the forecasted excess return, at the same time I sell appropriate amount of Hushen-300 stock index future, find that the strategy of buying group 1 (of which the forecast excess return was the highest)and sell stock index future can significantly bring a return about 1.4% in May 2010 and 3.1% in Jun 2010 when using the real trade data(not take the trade cost into account), while in the whole period which I studied in this paper, the excess return is about 2.2 percent. So the conclusion is quite simple:the strategy I studied in this paper is useful. And this paper also find the value effect,CAPM-beta effect,positive effective of size,reverse effective and negative effective of operating leverage.In the empirical work section, I used more than 20 independent variables which reflected almost all respect of financial index of a company, such as debt paying ability,solidity,profitability,risk level,profitability ratio of shareholder development capability and other variables including CAPM-beta,size,reverse effect and liquidity. The process of logic analysis and model construction was quite pithily but strictly. There are a lot of pictures and tables which made this paper very lively and simplicity. There are many small innovations through this paper, but the most primary innovation is the using of deviation of variables, through this way the empirical work section can be more effective, and what's more, this method constructed a bridge which connected the excess return model and real investment work. But been limited by the level and structure of expertise, this paper also have some deficiency which will be discussed in the text. However, this paper have it's practice and academic value as talked in the former paragraph.The main target of this paper is not to find a strategy which was expected to gain profit forever, but to research whether statistical arbitrage are effective or not in China market, and then inspire the interesting of other researchers on this field. In the years to come, statistical arbitrage strategy may be broadly used in China market, so the research of statistical arbitrage of China market is worthiness. At last, I have to declare that this paper was just an academic research, so it can't be used to do any investment directly. |
PDF Full Text Request