摘要
上证基金指数反映了基金市场的整体变动情况,研究其波动结构特征对基金市场参与者具有重要作用。研究结果表明:(1)上证基金指数序列可由经济基本面决定的趋势项、重大事件带来的低频分量和短期不均衡导致的高频分量构成,而且趋势项主导上证基金指数的长期走势,低频分量在中期对该指数有较大影响,而高频分量的影响可忽略不计;(2)与直接SVM预测法相比,EEMD-SVM组合预测法有更高的预测精度,说明EEMD分解得到的各结构分量有效地刻画了上证基金指数的内在运行特征。
Shanghai Securities Fund Index reflects the overall changes in fund market, so investigating its volatility structure is of the utmost importance to investors in this market. The results show that Shanghai Securities Fund Index can be composed of the trend series dominated by economic fundamen-tals, high frequency series created by short-term disequilibrium and low frequency series caused by big events; moreover, the trend decides the long-term trend of Shanghai Securities Fund Index, low fre-quency at median term exerts large influence upon it, but high frequency has little impact on it. Addi-tionally, the forecasting results demonstrate effectiveness and attractiveness of the proposed EEMD-based SVM ensemble learning paradigm compared with single SVM, which illustrates that the EEMD decomposition method effectively depicts the inherent characteristics of Shanghai Securities Fund Index.
出处
《金融理论与实践》
北大核心
2014年第1期80-85,共6页
Financial Theory and Practice
基金
国家社会科学基金(11CJY104)
福建省高等学校新世纪优秀人才支持计划(2012FJ-NCET-SK02)
福建省高等学校杰出青年科研人才培育计划(11FJPY04)
