B0707
Title: Multidimensional financial connectedness
Authors: Alessandro Celani - Universita Politecnica delle Marche (Italy)
Paolo Pagnottoni - University of Pavia (Italy) [presenting]
Abstract: The increasing availability of high and multidimensional data generated over time in finance has put severe limitations on standard approaches in multivariate time series econometric models. While it is common to model vectors of observations through standard vector time series analysis, the potential matrix form of data often reflects different types of structures of time series observations which can be further exploited to model interdependencies across financial securities. We propose a novel autoregressive model in a bilinear form which is able to: a) handle high and multidimensional financial data; b) yield enhanced interpretability given by the autoregressive model in matrix form; c) unveil dependencies across different sources of risks, i.e. price, volatility and liquidity risks. We illustrate the properties of our model through a real example of cryptocurrency market data.
