Date of Award
2019
Degree Type
Thesis
Degree Name
Master of Science in Statistics
Department
Computer Science and Statistics
First Advisor
Gavino Puggioni
Abstract
High-frequency data are observations collected at fine time scale. Such data largely incorporates pricing and transactions, of which institutional rules prevent from drastically rising or falling within a short period of time. This results in data changes based on the measure of one tick, a measure of the minimum upward or downward movement in the price of a security. The discreteness brings that the observations are in Z. A Skellam distribution has a unique property that returns values in Z.
We are interested in studying the Skellam process where the time-dependent intensities are Gaussian process. Such doubly stochastic Poisson process, also known as Cox process, is a point process which is a generalization of a Poisson process. We then investigate if this Skellam model provide better fit to high frequency financial data and how Gaussian process can capture the market microstructure.
Recommended Citation
Lee, Tingfang, "SKELLAM PROCESS MODELING FOR FINANCIAL HIGH-FREQUENCY DATA" (2019). Open Access Master's Theses. Paper 1520.
https://digitalcommons.uri.edu/theses/1520
Terms of Use
All rights reserved under copyright.