This article studies the co-movement between large daily revisions of short- and long-term inflation expectations using copulas. The main findings are: first, the co-movement between unusually large changes in short- and long-term inflation expectations increased markedly since mid-2012, which implies that long-term inflation expectations might not be, in a precise sense, well-anchored. Second, this co-movement measure is quite noisy. Finally, the result is shown not to be an artifact of the methodology or of the specific data used in the analysis.