Dinâmica de correlações no mercado financeiro Bovespa&BMF

摘要: In systems where many agents interact, allowing for measures that may be erratic, times we can extract behavior patterns denote a group of agents, this is the case financial market and its emerging structure correlations. This work aims to reproduce synthesize what perceived as correlation in markets. Analysis will consist 2 parts, one dynamic, accessing temporal dependencies, other topological economical by importance connections between assets. analysis dynamics are investigated instantaneous correlation, it’s across scales time, not instantaneous, decay from maximum correlation. The topology analyzed graph matrix analyzing connectivity vertices, starting most connected (called root) analyzes various clusters shares obtained comparing with known economic classification. performed at several order instantly compare them. Introduce general notions complex Chapter 1. give brief description through important variables their behavior, ie ranges price time. 3 describes methods used market, estimator presented Pearson’s linear Kruskal method obtain tree containing all assets minimize sum edges (weighted distance defined correlation). 4 I present results Bovespa. keywords: systems, econophysics, ultrametrics, epps effect, minimal spanning