A Sequential Bayesian Change-Point Analysis of BRICS Currency Returns

We examine the occurrence of structural breaks in the mean return and volatility of the BRICS currency returns. We propose a novel methodology to identify regimes with same statistical variance based on sequential Bayesian change point model. We find frequent occurrence of structural breaks in the mean returns and volatility of BRICS currencies. The segments with same statistical properties are merged into regimes. We sorted the regimes from low to high regarding their volatility. We show that the last regime exhibits the highest returns and incorporates the shortest periods across all currencies with exception to the Chinese and Brazilian currency. This is an indication that periods of high volatility are few and offers high returns compared to periods of low volatility.