In ΛCDM with statistically isotropic, Gaussian initial conditions, the CMB temperature two-point correlation function is expected to show substantial power even at very large angles, corresponding to low multipoles of the angular power spectrum (Spergel et al. 2003; Planck Collaboration 2016). However, observations from COBE, WMAP, and Planck reveal that the temperature correlation above about 60° is anomalously low, with the commonly used S_1/2 statistic far below the ΛCDM expectation, a feature that appears robust to different masks and estimators and that is difficult to attribute unambiguously to chance, foregrounds, or simple parameter shifts within the standard model (Copi et al. 2010; Schwarz et al. 2016).