This comment discusses two previous papers in this journal by J. Watson. It is shown that a confusion between necessary and sufficient conditions for the stability of the inter‐country correlation structure, and inconclusive empirical results led Watson to some misleading conclusions. It is also demonstrated that even if inter‐country correlation coefficients are stationary, ex‐ante correlations cannot be estimated directly from ex‐post data unless the underlying stochastic process is known. Alternative methods of examining the stability of the correlation structure indirectly are suggested. Furthermore it is argued that the choice of the base currency further limits Watson's f