This document discusses clustering random walk time series. It introduces the concept of clustering and discusses challenges in clustering time series, such as how to define a distance between dependent random variables. It proposes using the copula transform and empirical copula transform to map time series to uniform distributions before calculating distances. The hierarchical block model for nested data partitions is presented. Experimental results show the proposed distances perform well in recovering the nested partitions on both synthetic hierarchical block model data and real credit default swap time series data. Consistency of clustering algorithms is defined in relation to recovering the hierarchical block model partitions.