首页   按字顺浏览 期刊浏览 卷期浏览 Cumulative distribution function values for symmetric standardized stable distributions
Cumulative distribution function values for symmetric standardized stable distributions

 

作者: Don B. Panton,  

 

期刊: Communications in Statistics - Simulation and Computation  (Taylor Available online 1992)
卷期: Volume 21, issue 2  

页码: 485-492

 

ISSN:0361-0918

 

年代: 1992

 

DOI:10.1080/03610919208813030

 

出版商: Marcel Dekker, Inc.

 

关键词: Stable distributions;simulation;numerical integration

 

数据来源: Taylor

 

摘要:

There exists a significant body of empirical evidence indicating that the underlying distribution from which samples of stock returns have been drawn may reasonably be characterized as stable Paretian with characteristic exponent less than two. This paper employs the methodology of Woxsdale (1975) to calculate five-decimal-place cumulative distribution function (c.d.f.) values for standardized stable distributions with characteristic exponents alpha [α] between one and two. The standardization is that originally proposed by Fama and Roll (1968). In addition to extending the Fama and Roll four decimal c.d.f. table to five decimal places, the current paper demonstrates that the Worsdale procedure using numerical integration techniques readily overcomes the computer overflow problems found by Fama and Roll in employing the gamma expansions of Bergstrom (1952).

 

点击下载:  PDF (192KB)



返 回