分数阶积分误差下趋势结构断点的推断

Inference on a Structural Break in Trend with Fractionally Integrated Errors

Journal of Time Series Analysis · 2016
被引 16
ABS 3

中文导读

将Perron和Zhu(2005)关于趋势斜率变化的结构断点估计方法推广到分数阶积分误差情形,发现记忆参数d*在(-0.5,0.5)时断点估计收敛速度不变,其他情况随d*增大而减慢,并讨论了虚假断点问题。

Abstract

Perron and Zhu (2005) established the consistency, convergence rate and limiting distributions of parameter estimates in time trends with a change in slope with or without a concurrent level change for the cases with I(1) or I(0) errors. We extend their analysis to the general case of fractionally integrated errors with memory parameter d ∗ . Our results uncover interesting features; e.g., with a level shift allowed, the convergence rate for the break date estimate is the same for all d ∗ ∈(−0.5,0.5). In other cases, it is decreasing as d ∗ increases. We also provide results about the so‐called spurious break issue.

时间序列分析结构断点分数阶积分计量经济学