Mitochondrial outer membrane permeabilisation (MOMP) is a hallmark of apoptosis and is controlled by antagonising members of the BCL 2 protein family. However, whether the effector proteins for MOMP, BAX and BAK, require a separate activation step (direct activation model) or if the sole inhibition of otherwise active effectors by anti-apoptotic BCL 2 proteins suffices (indirect activation model) is controversial. To address this question, we recently identified that, in most cells, effector proteins were more abundant than their inhibitors. We now employ systems modelling to elucidate fundamental consequences of this imbalance for the indirect activation model by integrating its key features into the recently validated implementation of the direct activation model. First, by modelling constitutively active BAK, several cell lines were predicted to be unstable against MOMP even in the absence of stress. This instability could be rectified by assuming BAK pores to be subject to degradation, which however demanded BAK to underlie a high and biologically unreasonable protein turnover. Stability against MOMP in the absence of stress was similarly reconstituted by assuming BAK oligomerisation to be weak, but predicted some cells to be MOMP resistant even under high stress. Assuming also BAX to be constitutively active required highly effective BAX re-translocation to the cytosol along with weak BAX oligomerisation, which again rendered some cells insensitive to induce MOMP in the presence of stress. Our results suggest that an activation step of the two effectors is required to allow stability against MOMP in the absence of stress and execution of MOMP under stress.