Likelihood ratio test for the model allowing for different volatilities of the two time series (alternative hypothesis) versus a model enforcing the same volatility for the two time series (null hypothesis). Given scarcity of the data, the modeling framework can be simple AR(1).