| After Financial crisis, some investors followed the trend entered into the stock market. But unfortunately, their stocks were stuck for a long time. Hoping that the bull market in2009could be continued then their stock could be sold on no losses. But still unfortunately, another bear market was coming. Being deeply held-up in the stock market was spending investor’s confidence on the market. Therefore, there have been many accounts no transactions for a long time in securities companies, which is so called "invalid account".Under such background, the idea of this paper comes from my thinking during my internship in a securities company. Although anticipation said that bear market is coming, the help-up investor still doesn’t want to liquidate their position with some loss. Then their positions will face a great deal of market risk. In theory, this investor can lock their positions through sell short the stock index futures. But if the correlation of stock and stock index futures is too low, it is impossible to lock this stock position successfully. Therefore these kinds of held-up stocks have to face market risk? This paper tries to find method to deal with this problem. Investor’s all position can be seen as a portfolio. What if this portfolio has a strong correlation with stock index futures? Then investor can lock their total positions with futures. So, what are the Weights of the stock in the portfolio? That is the point in this paper. To have a maximum correlation between the portfolio and stock index futures, we need to minimum tracking error to get this index portfolio.In theory, every portfolio of fixed stocks has his own optimal proportions to track the aimed index. But, if the stocks in the portfolio have no correlation with stock index future entirely, the portfolio has the same performance. This paper can’t deal with this kind of portfolio. We have some basic constraint on the stocks in the portfolio. First, unless one stock that have a strong correlation with stock index is needed in the portfolio:a strong correlation means0.8or above. Second, there only CSI300Index Futures exit in the China’s market. Its Underlying is CSI300Index. CSI300Index reflects the price trend of A-share market. So, the stocks of portfolio should issue in A-share market. The exiting research about index investing all based on the constituent stocks or stocks that have strong correlation with CSI300Index. This paper tries to make those uncorrelated stocks correlating in portfolio. That is the innovation of this paper.Based on the author’s vision, first step of the paper is choosing the optimization of tracking index. A lot of research found that data mining methods have a better tracking effect. Wu (2011) compared the different types of data mining methods index tracking effect. His study showed that there is no different effect between different data mining methods. Fan (2006) found that Generic Algorithms is a better choice to minimum the tracking error. Genetic algorithms progress multi-point search of the solution space. Under certain conditions, the genetic algorithm can always converge to the optimal solution of the problem with probability1.Considering Generic Algorithms have a better way in searching optimum solution, this paper choose Generic Algorithm to optimize the proportions of the portfolio. The target function is minimizing the tracking error. After optimization, we get the index portfolio. The index portfolio tracks the index well:its yield correlation coefficient with CSI300Index is0.922in-sample and0.8269out-of-sample. We think this index portfolio can be hedge with futures. Next Step, to make the research more rigorous, we compare CVaR of hedging and no hedging index portfolio to get the hedging feasibility of index portfolio. The CVaR is based on Copula-GARCH model. After that we find CVaR of hedging index portfolio is lower than no hedging index portfolio. We get our conclusion that hedging hedge the market risk, then index portfolio can do hedging. Finally, we estimate the optimal hedging ratio for the index portfolio then study the hedging effect. Comprehensive considering the CVaR and the hedging effect, we get summary of this paper. This index portfolio, optimal weights through Generic Algorithms, finally achieves hedging the market risk, which can’t be done with only one or some stock in the portfolio. |
PDF Full Text Request