基于LPPL模型的股市暴跌风险预警

LPPL-Model-Based Risk Prediction of China’s Stock Market Slump

文章以2015年6月上证指数和创业板指数暴跌为研究对象。运用改进的LPPL模型,利用暴跌之前的数据,对暴跌时点进行预测,并结合Lomb谱分析和O-U过程检验对预测结果进行验证。结果表明,改进的LPPL模型对上证指数和创业板指数的暴跌时点均具有较好的预测效果,同时模型预测能力不受拟合区间起点和终点变更影响,且拟合结果通过了相关检验。

This paper takes the Shanghai Composite Index and GEM index collapse in June 2015 as the research object. By constructing an improved LPPL model and using the data before the collapse of the crash point, the paper predicts when the crash point will appear. And then the paper combines Lomb spectrum analysis and O-U-process inspection to verify the predicted resuits. The result shows that the improved LPPL model has favorable effect in predicting the crash point, and that the predictive power of the model is not distracted by the change of fitting interval, and also that the fitting result agrees with the corresponding test.