Presentation #102.91 in the session Poster Session.
The now established high abundance of exoplanets around low-mass stars makes them ideal to place constraints on planet formation. Here, we compare the latest planet occurrence rates from the radial velocity survey CARMENES and HARPS to a synthetic planet population obtained from the Bern model of planet formation and evolution. We predict a high, intrinsic occurrence of up to one temperate, rocky planet per mid M dwarf, but also a decrease of the number of planets toward later types. Such an outcome is in agreement with the relatively low number of rocky planet detections found by CARMENES. However, key discrepancies still exist between theory and observation. For giant planets, a general feature of core accretion models with runaway gas accretion is the existence of a lower stellar mass limit around 0.5 solar masses below which no giants can form. This contradicts the detection of several giant planets around late-type M dwarfs. In addition, we find a dependence of the observed detection rate of short-period planets on stellar mass, which is different from the model. We show that these discrepancies can be reconciled by adapting the migration timescale in the models. This is further motivated by the high abundance of sub-structures in disks which could act as planet traps.