Financial Applications of Forecasting Models for Functional Time Series

Functional data analysis is a growing field in statistics that studies data that are observed in the form of functions. These types of high-dimensional data appear in many real-world applications such as the study of demographic curves, estimation of the concentration of atmospheric aerosol particles (such as PM10) and the study of the daily electricity demand profiles. Consequently, the need for models that are able to obtain accurate estimations of these continuous data have increased. When analyzing financial markets, one often encounter these high-frequency data in the form of financial transaction data, intraday price curves or yield curves, among others. The analysis of how the recently developed functional data techniques can be applied to financial data would be of major interest, possibly leading to an improvement of forecasting models in financial markets.