A new class of statistical methods for analysing long memory time series models with heteroskedasticity

一类新的统计方法，用于分析具有异方差性的长记忆时间序列模型

• 批准号: DP1094010
• 负责人: Prof David Harris
• 金额: $ 13.46万
• 依托单位: Monash University
• 依托单位国家: 澳大利亚
• 项目类别: Discovery Projects
• 财政年份: 2010
• 资助国家: 澳大利亚
• 起止时间: 2010-01-01 至 2013-12-31
• 项目状态: 已结题
• 来源: https://dataportal.arc.gov.au/RGS/Web/Grants/DP1094010
• 关键词: new; class; statistical; methods; analysing

This project will result in a class of statistical methods that will aid policy makers and financial analysts when examining and predicting key international and Australian macroeconomic and financial variables that exhibit long memory. Leading applications of long memory modelling in the literature include GDP, CPI, asset pricing models, stock returns, exchange rates and interest rates. It will be possible to robustly and efficiently analyse such series in the presence of changes in variability, such as the overall reduction in variability that has occurred since the 1970's, called the "Great Moderation". The utility of the new methods will be demonstrated by a robust and efficient analysis of the Purchasing Power Parity hypothesis.

该项目将产生一类统计方法，这些方法将帮助政策制定者和金融分析师在检查和预测表现出长期记忆的关键国际和澳大利亚宏观经济和金融变量时提供帮助。在文献中，长记忆模型的主要应用包括GDP、CPI、资产定价模型、股票收益、汇率和利率。在存在变异性变化的情况下，例如自20世纪70年代以来发生的变异性总体减少，即所谓的“大缓和”，将有可能对这些序列进行稳健而有效的分析。新方法的效用将通过对购买力平价假设的稳健而有效的分析来证明。