Searching for chaotic components in financial time-series
Abstract
This study attempts to distinguish between random and chaotic motion in several U.S. Treasury issues as reflected in 596 secondary market week-end interest rates covering the period January, 1980 to May, 1991. The Treasury series studied are the 3 month and 1 year Bills, the 1 year and 10 year Notes, and the 30 year Bond. The series are rescaled to a value range of.0000 to 1.000 and compared to 596 random data values and 596 values of the Logistic function which is known to be chaotic in nature. The Treasury series as well as the Logistic and random data series are reformulated into three dimension series using the Packard-Takens method and are evaluated using Lyapunov Exponent and Correlation Dimension measures. The investigation focuses on phase-space portrait differences rather than an extensive quantitative analysis. The central hypothesis of non-randomness in the Treasury series behavior is strongly supported, but the evidence for chaotic behavior, while indicative, is ambiguous. An extensive introductory section on chaotic behavior is also included.
Subject Area
Finance|Computer science|Management
Recommended Citation
Enright, Arthur J, "Searching for chaotic components in financial time-series" (1992). ETD Collection for Pace University. AAI9226411.
https://digitalcommons.pace.edu/dissertations/AAI9226411
Remote User: Click Here to Login (must have Pace University remote login ID and password. Once logged in, click on the View More link above)