arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv physics/0607180physics.soc-phq-fin.GN
收藏

产出增长率分布是什么样的?关于OECD国家的一些时间序列证据

How Do Output Growth Rate Distributions Look Like? Some Time-Series Evidence on OECD Countries

Giorgio Fagiolo, Mauro Napoletano, Andrea Roventini

摘要:

本文研究了国内生产总值和工业生产增长率分布的统计特性。近年来,许多实证研究成果指出,企业、行业和国家的横截面增长率都服从拉普拉斯分布。在这项工作中,我们测试是否也在国内,时间序列GDP和IP增长率可以近似帐篷形分布。我们拟合输出增长率与指数幂(Subbotin)的密度,其中包括作为特殊情况下的高斯分布和拉普拉斯分布。我们发现,对于包括美国在内的大量经合组织国家来说,GDP和IP增长率均呈拉普拉斯分布。此外,我们表明,厚尾分布鲁棒出现后,即使控制离群值,自相关性和异方差。

英文摘要:

This paper investigates the statistical properties of within-country GDP and industrial production (IP) growth rate distributions. Many empirical contributions have recently pointed out that cross-section growth rates of firms, industries and countries all follow Laplace distributions. In this work, we test whether also within-country, time-series GDP and IP growth rates can be approximated by tent-shaped distributions. We fit output growth rates with the exponential-power (Subbotin) family of densities, which includes as particular cases both the Gaussian and the Laplace distributions. We find that, for a large number of OECD countries including the U.S., both GDP and IP growth rates are Laplace distributed. Moreover, we show that fat-tailed distributions robustly emerge even after controlling for outliers, autocorrelation and heteroscedasticity.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑