本研究基于对数周期幂律模型LPPL(Log Periodic Power Law Model),针对金融时间序列将一维价格波动翻译成反映市场泡沫微观结构的多维变量。通过对多维变量的动态监测,把握市场中泡沫的演变并预测泡沫破裂的临界点,从而有效降低或防范...本研究基于对数周期幂律模型LPPL(Log Periodic Power Law Model),针对金融时间序列将一维价格波动翻译成反映市场泡沫微观结构的多维变量。通过对多维变量的动态监测,把握市场中泡沫的演变并预测泡沫破裂的临界点,从而有效降低或防范金融资产泡沫破裂所导致的风险。为检验LPPL模型在中国金融市场中的适用性,本文分别使用上证综指、四个期货连续合约以及两支个股检验模型效果。实证结果表明当金融资产价格序列呈现超指数加速震荡上升或下降时,该模型能获得稳定的估计效果,有效预测泡沫破裂临界时点。展开更多
Let F=C（x1,x2,…,xe,xe＋1,…,xm）, where x1, x2,… , xe are differential variables, and xe＋1,…,xm are shift variables. We show that a hyperexponential function, which is algebraic over F,is of form g（x1, x2, …,xm...Let F=C（x1,x2,…,xe,xe＋1,…,xm）, where x1, x2,… , xe are differential variables, and xe＋1,…,xm are shift variables. We show that a hyperexponential function, which is algebraic over F,is of form g（x1, x2, …,xm）q（x1,x2,…,xe）^1/lwe＋1^xe＋1…wm^xm, where g∈ F, q ∈ C（x1,x2,…,xe）,t∈Z^＋ and we＋1,…,wm are roots of unity. Furthermore,we present an algorithm for determining whether a hyperexponential function is algebraic over F.展开更多
Using the hyper-exponential recurrence criterion,we establish the occupation measures’large deviation principle for a class of non-linear monotone stochastic partial differential equations(SPDEs)driven by Wiener nois...Using the hyper-exponential recurrence criterion,we establish the occupation measures’large deviation principle for a class of non-linear monotone stochastic partial differential equations(SPDEs)driven by Wiener noise,including the stochastic p-Laplace equation,the stochastic porous medium equation and the stochastic fast-diffusion equation.We also propose a framework for verifying hyper-exponential recurrence,and apply it to study the large deviation problems for strong dissipative SPDEs.These SPDEs can be stochastic systems driven by heavy-tailedα-stable process.展开更多
文摘本研究基于对数周期幂律模型LPPL(Log Periodic Power Law Model),针对金融时间序列将一维价格波动翻译成反映市场泡沫微观结构的多维变量。通过对多维变量的动态监测,把握市场中泡沫的演变并预测泡沫破裂的临界点,从而有效降低或防范金融资产泡沫破裂所导致的风险。为检验LPPL模型在中国金融市场中的适用性,本文分别使用上证综指、四个期货连续合约以及两支个股检验模型效果。实证结果表明当金融资产价格序列呈现超指数加速震荡上升或下降时,该模型能获得稳定的估计效果,有效预测泡沫破裂临界时点。
基金The research is supported in part by the 973 project of China(2004CB31830).
文摘Let F=C（x1,x2,…,xe,xe＋1,…,xm）, where x1, x2,… , xe are differential variables, and xe＋1,…,xm are shift variables. We show that a hyperexponential function, which is algebraic over F,is of form g（x1, x2, …,xm）q（x1,x2,…,xe）^1/lwe＋1^xe＋1…wm^xm, where g∈ F, q ∈ C（x1,x2,…,xe）,t∈Z^＋ and we＋1,…,wm are roots of unity. Furthermore,we present an algorithm for determining whether a hyperexponential function is algebraic over F.
基金supported by National Natural Science Foundation of China(Grant Nos.11431014 and 11671076)supported by University of Macao Multi-Year Research Grant(Grant No.MYRG2016-00025-FST)Science and Technology Development Fund,Macao SAR(Grant Nos.025/2016/A1,030/2016/A1 and 038/2017/A1)the Faculty of Science and Technology,University of Macao,for financial support and hospitality。
文摘Using the hyper-exponential recurrence criterion,we establish the occupation measures’large deviation principle for a class of non-linear monotone stochastic partial differential equations(SPDEs)driven by Wiener noise,including the stochastic p-Laplace equation,the stochastic porous medium equation and the stochastic fast-diffusion equation.We also propose a framework for verifying hyper-exponential recurrence,and apply it to study the large deviation problems for strong dissipative SPDEs.These SPDEs can be stochastic systems driven by heavy-tailedα-stable process.
