The Structure of Financial Returns
14 Pages Posted: 5 Jun 2020
Date Written: December 24, 2019
Financial returns at unit time are modeled as non-Gaussian limit laws. They may reflect random walks or additive processes reflecting some predictability. Mixtures of these two constructions are formulated and estimated on one minute data. It is observed that the random walk fraction is generally below 10%. The results argue against a strict random walk in favor of the presence of a predictable component representing returns as perpetual motion machines responding to past price movements.
Keywords: Bilateral Gamma, CGMY, Sato Process, Random Walk
JEL Classification: G10, G17, G19
Suggested Citation: Suggested Citation