When two galaxies merge, their central supermassive black holes are expected to form a bound binary that shrinks by interacting with stars and gas, but at separations of order a parsec, standard loss-cone scattering and dynamical friction become inefficient, so the binary can stall before gravitational-wave emission takes over and completes the merger (Begelman et al. 1980; Vasiliev 2014). In ΛCDM this “final parsec problem” raises a tension between theoretical expectations of long-lived stalled binaries and observations and PTA inferences that suggest frequent supermassive black hole coalescences, forcing models to invoke special galaxy shapes, gas physics, or dark-matter microphysics to bridge the gap (Khan et al. 2013; Merritt 2013).