基于GARCH模型的股票市场波动预测——以沪深300指数为例

Predicting Fluctuation in Stock Market Based on GARCH Model

1990年上海证券交易所和深圳证券交易所正式成立。在这将近三十年时间里,我国对于股票证券市场的监管和体制方面都有很大的改变。可是我国股票市场相比国外比较健全的股票市场还是存在一定的差距。自从2008年金融危机以后,人们逐渐将注意力放到了股票市场的预测上。股票的预测主要是对股票的波动的分析。如何分析描述股票的波动及股票未来收益率成为当今最热门的话题之一。AR CH族模型是目前最常用的分析股票波动的模型,其中GAR CH模型更是可以很好地拟合股票波动率。其次蒙特卡罗模拟是分析随机问题的一种必要手段,将GARCH模型和蒙特卡洛模拟结合可以更加准确的分析股票波动率。本文结合沪深300指数,利用软件R对不同阶数的GARCH模型进行模拟分析,根据各种条件确定相对最优模型,发现GARCH(1,1)对股票收盘指数模拟效果最好,再利用GARCH(1,1)结合蒙特卡洛模拟对未来沪深300日收盘价指数进行短期预测。通过预测,可以大概了解股票的波动规律,能有效地对是否进行股票交易提供合理的参考。

China has made a lot of changes and improvements in regulation and supervision of its stock exchange market in the last three decades since Shanghai and Shenzhen Stock Exchanges were founded, but there is still great gap between China＇s stock market and the sound foreign market. After the economic crisis in 2008,the prediction on stock market comes to attract people＇s attention. The prediction mainly concerns the analysis of the stock fluctuation. As a result, it becomes one of the hot topics how to analyze and describe the stock fluctuation and the future return rate.ARCH series are the most commonly used models in stock fluctuation analysis, among which ARCH is particularly good in fitting the fluctuating rate of the stock. Besides, Monte Carlo simulation is the necessary means to analyze the random phenomena. So with the combination of ARCH and Monte Carlo simulation, stock market fluctuation can be more accurately predicted. With SCI 300 index, the paper stimulates and analyzes the ARCH model in its different stages using R software and establishes the comparatively best model based on various conditions. It is found that GARCH（1,1） can best simulate the stock closing index. And then making use of GARCH（1,1） combined with Mote Carlo model, daily closing price index of SCI 300 is predicted. Thus the regular pattern of stock market fluctuation is roughly revealed, offering useful reference to the decision of stock exchange.