Financial intelligence and strategy extraction via visual analysis approaches

Abstract

Visualization techniques have been widely utilized to facilitate the analysis in different financial fields, such as trading markets, risk assessment, anomaly transaction detection, asset management. Visual analysis techniques could contribute to systematic intelligence generation and strategy extraction for domain practitioners and overcome the perceptual barrier for the public. However, previous works mainly focus on the novel visual representation or internal system which relies on high domain knowledge. In this thesis, we focused on the visual analysis of financial transaction data to generate intelligence, strategies, and patterns that are more easily accepted by the public and end-users, rather than being limited to the professionals with strong domain knowledge. We grounded our study on specific applications: cryptocurrency exchange and quantitative investment. The first research problem focuses on the evolutionary transaction patterns of cryptocurrency exchanges. Delving into the analysis of the transaction patterns of exchanges can shed light on the evolution and trends in the cryptocurrency market, and participants can gain hints for identifying credible exchanges as well. Specifically, we present a visual analytics system named BitExTract, which is the first attempt to explore the evolutionary transaction patterns of Bitcoin exchanges from two perspectives, namely, exchange versus exchange and exchange versus client. Our second focusing area is quantitative investment. The essence of quantitative investment is the multi-factor model, one that explains the relationship between the risk and return of equities. The challenge is to develop visualization tools that can effectively analyze financial factors in stock selection and portfolio construction. Also, the portfolio measurement has also been expanded to factors-level except the return and position which is insufficient for actionable insights and understanding of market trends. We introduce the progress to date by summarizing the methods we have developed that address the aforementioned research problems. Thus we present sPortfolio, which is the first visual analytic system that attempts to explore the factor investment area. In particular, sPortfolio provides a holistic overview of the factor data and aims to facilitate quantitative market analysis. We also design iQUANT, an interactive quantitative investment system that assists equity traders to quickly spot promising financial factors from initial recommendations suggested by algorithmic models, and conduct a joint refinement of factors and stocks for investment portfolio composition. In the last, we briefly discuss future research works as well as open questions in financial visualization area.

Date of Award	2019
Original language	English
Awarding Institution	The Hong Kong University of Science and Technology