The present study investigates financial arbitrage strategies, overcoming limitations of techniques such as pairs trading. Through multivariate cointegration in the Johansen sense, the aim is to construct a long-short portfolio by identifying an equilibrium relationship among multiple assets. Parameters were estimated using maximum likelihood, while arbitrage opportunities were discovered by modeling the spread, targeting mean-reversion trades. This approach is consistent with economic theory and allows for determining portfolio asset weights. The empirical analysis collected data from Economatica, covering 203 stocks from the Bovespa index from 1997 to 2023. Results were compared to a benchmark and subjected to robustness tests, confirming the strategy's effectiveness in generating returns without significantly increasing risk.
Complete tool for submission, evaluation and publication of proceedings for scientific events.
