ABSTRACT
Under the influence of the COVID19 pandemic, Bitcoin, gold, copper and silver prices have exhibited sudden changes. For this reason, in this paper, it was aimed to investigate contagion behavior, and the volatility of bitcoin, gold, copper and silver prices by using Markov Switching GARCH Multilayer Perceptron (MS-GARCH-MLP) Copula method in the period of February 02, 2012–May 29, 2020. Firstly, the nonlinear, uncertainity and chaotic structure of Bitcoin, gold, silver, and copper were determined by Largest Lyapunov Exponent and Shannon Entropy techniques. Following, the MS-GARCH-MLP Copula method was emerged and applied to explore the existence of persistence and contagion. Our findings presented that there are presence of persistence and the evidences of contagion between the variables. At the final stage, forecast performance at our model was analyzed. The forecast results showed that the best performance is observed at bitcoin and silver for the long run.