The Volatility Forecasting Power of Financial Network Analysis

Nicolás S. Magner, Jaime F. Lavin, Mauricio A. Valle, Nicolás Hardy

Research output: Contribution to journalArticlepeer-review

7 Scopus citations

Abstract

This investigation connects two crucial economic and financial fields, financial networks, and forecasting. From the financial network's perspective, it is possible to enhance forecasting tools, since econometrics does not incorporate into standard economic models, second-order effects, nonlinearities, and systemic structural factors. Using daily returns from July 2001 to September 2019, we used minimum spanning tree and planar maximally filtered graph techniques to forecast the stock market realized volatility of 26 countries. We test the predictive power of our core models versus forecasting benchmarks models in and out of the sample. Our results show that the length of the minimum spanning tree is relevant to forecast volatility in European and Asian stock markets, improving forecasting models' performance. As a new contribution, the evidence from this work establishes a road map to deepening the understanding of how financial networks can improve the quality of prediction of financial variables, being the latter, a crucial factor during financial shocks, where uncertainty and volatility skyrocket.

Original languageEnglish
Article number7051402
JournalComplexity
Volume2020
DOIs
StatePublished - 2020
Externally publishedYes

Fingerprint

Dive into the research topics of 'The Volatility Forecasting Power of Financial Network Analysis'. Together they form a unique fingerprint.

Cite this