Functional Data Analysis in Finance

About the Project

In Functional data analysis (FDA), the variable of interest can be naturally viewed as a smooth curve or function, rather than scalars in univariate analysis or vectors in multivariate analysis. The field has witnessed rapid development over the last two decades. While the central ideas and methods of FDA has achieved a certain mature level, its applications in various subjects is still in an accelerating speed. Among them, finance is one of fields that are largely benefited from the widely use of tools from FDA. In finance, some of the prominent examples that can be naturally viewed as curves include: intraday price curves (Kokoszka et al, 2015), term structure of interest rates (Barsley, 2017), forward curves of commodity futures (Horváth et al, 2019), and price signatures (Oomen, 2019).

The nature of this PhD project is employing newly developed tools in FDA to produce new insights for financial study, which cannot be revealed from the conventional methods. Thus, your first and second chapters will mainly be applied work. It is likely that you may find some limitations in the current toolkit of FDA for some specific finance problems. Then, the third chapter can develop a new method in FDA, devoting to expand the applicability of FDA in finance.

Department of Economics, University of Reading:

The Department of Economics has a long and established track record of research, working with a wide variety of industrial and academic partners to achieve significant social and economic benefits. Research activity within the Department is broad and extensive; among our most active fields are business economics, development economics, behavioural economics, labour economics and sports economics.

We have an active community of 20-30 PhD students and provide you with the opportunity to carry out your studies and learn in a highly collaborative environment, putting you on the path to a successful career.

We offer flexible modes of study designed to fit with your needs. Our PhDs are available for study on a full-time basis over three years and part-time over four to six years, starting in the autumn term of the academic year. Both full-time and part-time variants are available for study in Reading, or at a distance for students who live outside the UK.

Eligibility:

Applicants should have a good master degree in Economics or on a strongly-related discipline. Applicants will also need to meet the University’s English Language requirements. We offer pre-sessional courses that can help with meeting these requirements. Submit an application for a PhD in Economics at http://www.reading.ac.uk/pgapply

Pre-requisites:

A solid background from mathematics or statistics with knowledge in economics and finance; proficiency in using Matlab, R, and Python; a good understanding of major textbooks in FDA, including Ramsay and Silverman (2005), Horváth and Kokoszka (2012), Kokoszka and Reimherr (2017).

Funding Notes

Funding opportunities are available on a competitive basis through the Economic and Social Research Council (ESRC) South East Network for Social Sciences (SeNSS). You can find more information here: View Website

References

• Bardsley, P., Horváth, L., Kokoszka, P., & Young, G. (2017). Change point tests in functional factor models with application to yield curves. The Econometrics Journal, 20(1), 86-117.
• Horváth, L., & Kokoszka, P. (2012). Inference for functional data with applications (Vol. 200). Springer Science & Business Media.
• Horváth, L., Liu, Z., Rice, G., & Wang, S. (2019). A functional time series analysis of forward curves derived from commodity futures. International Journal of Forecasting.
• Kokoszka, P., Miao, H., & Zhang, X. (2015). Functional dynamic factor model for Intraday price curves. Journal of Financial Econometrics, 13(2), 456-477.
• Kokoszka, P., & Reimherr, M. (2017). Introduction to functional data analysis. CRC Press.
• Oomen, R. (2019). Price signatures. Quantitative Finance, 19(5), 733-761.
• Ramsay, J. O., & Silverman, B. W. (2005). Functional data analysis. Springer.