当主成分存在条件相关时,正交广义自回归条件异方差模型(orthogonal generalized autoregressive conditional heteroskedasticity,O-GARCH)可能导致预测结果与实际情况不相符。基于独立分量分析的广义自回归条件异方差模型(GARCH model...当主成分存在条件相关时,正交广义自回归条件异方差模型(orthogonal generalized autoregressive conditional heteroskedasticity,O-GARCH)可能导致预测结果与实际情况不相符。基于独立分量分析的广义自回归条件异方差模型(GARCH model based on independent component analysis,ICA-GARCH)能有效解决此问题,但ICA-GARCH模型中的梯度下降算法易陷入局部最优,收敛精度也有待提高。为克服此缺点,本文提出了一种基于粒子群优化算法的ICA-GARCH模型(ICA-GARCH model based on particle swarm optimization,PSO-ICA-GARCH),并将其用于证券市场收益波动率建模,以最终提高收益率预测效果。通过对阿里巴巴概念股收益波动率的实证分析,结果显示PSO-ICA-GARCH模型相较于O-GARCH和ICA-GARCH模型,具有更高的分离精度和更准确的模型预测效果。展开更多
The emphasis of this study is on the practice of the Pooled Mean Group （PMG） estimators to investigate the magnitude of macroeconomic performances： Real Gross Domestic Product （RGDP）, Foreign Exchange Rate （EX）...The emphasis of this study is on the practice of the Pooled Mean Group （PMG） estimators to investigate the magnitude of macroeconomic performances： Real Gross Domestic Product （RGDP）, Foreign Exchange Rate （EX）, and Deposit Interest Rate （DINT） affecting on the rate of financial sector returns in Southeast Asian Stock Markets including Stock Exchange Of Thailand （SET） index （Thailand）, the Kuala Lumpur Composite Index （KLSE） index （Malaysia）, Financial Times Share Index （FTSI） （Singapore）, Philippine Stock Exchange （PSE）, and the Jakarta Composite Index （JKSE） （Indonesia）. The Panel Autoregressive Distributed Lag （Panel ARDL） is applied to model the relations. The study applies the Levin, Lin, and Chu （LLC） test （2002） and Im, Pesaran, and Shin （IPS） test （2003） to investigates a set of time series data to examine whether the determinants and the rate of financial sector returns contain a unit root, the next step is investigated the cointegration and causality relationship of the determinants of financial sector influencing on long-run rate of returns of financial sector in Southeast Asian Stock Markets.展开更多
文摘当主成分存在条件相关时,正交广义自回归条件异方差模型(orthogonal generalized autoregressive conditional heteroskedasticity,O-GARCH)可能导致预测结果与实际情况不相符。基于独立分量分析的广义自回归条件异方差模型(GARCH model based on independent component analysis,ICA-GARCH)能有效解决此问题,但ICA-GARCH模型中的梯度下降算法易陷入局部最优,收敛精度也有待提高。为克服此缺点,本文提出了一种基于粒子群优化算法的ICA-GARCH模型(ICA-GARCH model based on particle swarm optimization,PSO-ICA-GARCH),并将其用于证券市场收益波动率建模,以最终提高收益率预测效果。通过对阿里巴巴概念股收益波动率的实证分析,结果显示PSO-ICA-GARCH模型相较于O-GARCH和ICA-GARCH模型,具有更高的分离精度和更准确的模型预测效果。
文摘The emphasis of this study is on the practice of the Pooled Mean Group （PMG） estimators to investigate the magnitude of macroeconomic performances： Real Gross Domestic Product （RGDP）, Foreign Exchange Rate （EX）, and Deposit Interest Rate （DINT） affecting on the rate of financial sector returns in Southeast Asian Stock Markets including Stock Exchange Of Thailand （SET） index （Thailand）, the Kuala Lumpur Composite Index （KLSE） index （Malaysia）, Financial Times Share Index （FTSI） （Singapore）, Philippine Stock Exchange （PSE）, and the Jakarta Composite Index （JKSE） （Indonesia）. The Panel Autoregressive Distributed Lag （Panel ARDL） is applied to model the relations. The study applies the Levin, Lin, and Chu （LLC） test （2002） and Im, Pesaran, and Shin （IPS） test （2003） to investigates a set of time series data to examine whether the determinants and the rate of financial sector returns contain a unit root, the next step is investigated the cointegration and causality relationship of the determinants of financial sector influencing on long-run rate of returns of financial sector in Southeast Asian Stock Markets.
