Off-campus UNL users: To download campus access dissertations, please use the following link to log into our proxy server with your NU ID and password. When you are done browsing please remember to return to this page and log out.
Non-UNL users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Statistical Arbitrage in Momentum and Pairs Trading by Machine Learning Models and Copulas
Abstract
A method to buy and sell in markets based on predefined rules to make trading decisions is a market-neutral trading strategy if it exhibits zero correlation with the unwanted source of risks. In addition, researchers are interested in obtaining higher returns than markets with special methods such as momentum and statistical arbitrage. In this dissertation, machine learning models and deep learning models were applied to predict stock’s movement with momentum trading in the first two topics, and a multivariate pairs trading strategy was developed in the third topic. Unlike most previous literature that used machine learning models in a single market, machine learning models with momentum trading in four different markets – the United States, mainland China, Hong Kong, and the United Kingdom with varying data sizes were explored, which made the result more robust. Next, a stacking ensemble of multiple deep learning models was used to predict the movements of Standard and Poor’s 500 (S&P 500) components by natural language processing (NLP) with Securities and Exchange Commission (SEC) 8-K files. Finally, a canonical-vine (C-vine) + drawable-vine (D-vine) copula model was developed to make bivariate decomposition of a four-dimensional dataset that captured temporal and cross-sectional relationships among the dataset to create the multivariate pairs trading signals in the Hong Kong market.
Subject Area
Statistics|Commerce-Business|Artificial intelligence
Recommended Citation
Xu, Jiaqi, "Statistical Arbitrage in Momentum and Pairs Trading by Machine Learning Models and Copulas" (2021). ETD collection for University of Nebraska-Lincoln. AAI28652984.
https://digitalcommons.unl.edu/dissertations/AAI28652984