ABSTRACT An empirical model is presented that links, for the first time, the demographics of active galactic nuclei (AGN) to their ensemble X-ray variability properties. Observations on the incidence of AGN in galaxies are combined with (i) models of the power spectrum density (PSD) of the flux variations of AGN and (ii) parametrizations of the black hole mass versus stellar mass scaling relation to predict the mean excess variance of active black hole populations in cosmological volumes. We show that the comparison of the model with observational measurements of the ensemble excess variance as a function of X-ray luminosity provides a handle on both the PSD models and the black hole mass versus stellar mass relation. We find strong evidence against a PSD model that is described by a broken power law and a constant overall normalization. Instead, our analysis indicates that the amplitude of the PSD depends on the physical properties of the accretion events, such as the Eddington ratio and/or the black hole mass. We also find that current observational measurements of the ensemble excess variance are consistent with the black hole mass versus stellar mass relation of local spheroids based on dynamically determined black hole masses. We also discuss future prospects of the proposed approach to jointly constrain the PSD of AGN and the black hole mass versus stellar mass relation as a function of redshift.