COSMOS2020: The cosmic evolution of the stellar-to-halo mass relation for central and satellite galaxies up to z~5

We use the COSMOS2020 catalogue to measure the stellar-to-halo mass relation (SHMR) divided by central and satellite galaxies from $z=0.2$ to $z = 5.5$. Starting from accurate photometric redshifts we measure the near-infrared selected two-point angular correlation and stellar mass functions in ten redshift bins and fit them with an HOD-based model. At each redshift, we measure the ratio of stellar mass to halo mass, $M_*/M_h$, which shows the characteristic strong dependence of halo mass with a peak at $M_h^{\rm peak} \sim 10^{12}\, M_{\odot}$. Our results are in accordance with the scenario in which the peak of star-formation efficiency moves towards more massive halos at higher redshifts. We also measure the fraction of satellites as a function of stellar mass and redshift. For all stellar mass thresholds the satellite fraction decreases at higher redshifts. At a given redshift there is a higher fraction of low-mass satellites. The satellite contribution to the total stellar mass budget in halos becomes more important than centrals at halo masses of about $M_h > 10^{13} \, M_{\odot}$ and always stays below by peak, indicating that quenching mechanisms are present in massive halos that keep the star-formation efficiency low. Finally, we compare our results with three hydrodynamical simulations Horizon-AGN, Illustris-TNG-100 and EAGLE. We find that the most significant discrepancy is at the high mass end, where the simulations generally show that satellites have a higher contribution to the total stellar mass budget than the observations. This, together with the finding that the fraction of satellites is higher in the simulations, indicates that the feedback mechanisms acting in group-and cluster-scale halos appear to be less efficient in quenching the mass assembly of satellites, and/or that quenching occurs much later in the simulations.