Capturing Financial Volatility Through Simple Network Measures

Measuring the inner characteristics of financial markets risks have been proven to be key at understanding what promotes financial instability and volatility swings. Advances in complex network analysis have shown the capability to characterize the specificities of financial networks, ranging from credit networks, volatility networks, and supply-chain networks, among other examples. Here, we present a price-correlation network model in which Standard & Poors' members are nodes connected by edges corresponding to price-correlations over time. We use the average degree and the frequency of specific motifs, based on structural balance, to evaluate if it is possible, with these simple measures, to identify financial volatility. Our results suggest the existence of a significant correlation between the Index implied volatility (measured with the VIX Index) and the average degree of the network. Moreover, we identify a close relation between volatility and the number of balanced positive triads. These results are shown to be robust to a wide range of time windows and correlations thresholds, suggesting that market instability can be inferred from simple topological features.