Title: Observational Support for Massive Black Hole Formation Driven by Runaway Stellar Collisions in Galactic Nuclei

Abstract: We explore here an scenario for massive black hole formation driven by stellar collisions in galactic nuclei, proposing a new formation regime of global instability in nuclear stellar clusters triggered by runaway stellar collisions. Using order of magnitude estimations, we show that observed nuclear stellar clusters avoid the regime where stellar collisions are dynamically relevant over the whole system, while resolved detections of massive black holes are well into such collision-dominated regime. We interpret this result in terms of massive black holes and nuclear stellar clusters being different evolutionary paths of a common formation mechanism, unified under the standard terminology of being both central massive objects. We propose a formation scenario where central massive objects more massive than $\rm \sim 10^8 \, Msun$, which also have relaxation times longer that their collision times, will be too dense (in virial equilibrium) to be globally stable against stellar collisions and most of its mass will collapse towards the formation of a massive black hole. Contrarily, this will only be the case at the core of less dense central massive objects leading to the formation of black holes with much lower black hole efficiencies $\rm \epsilon_{BH} = \frac{M_{BH}}{M_{CMO}}$, with these efficiencies $\rm \epsilon_{BH}$ drastically growing for central massive objects more massive than $\rm \sim 10^7 \, Msun$, approaching unity around $\rm M_{CMO} \sim 10^8 \, Msun$. We show that the proposed scenario successfully explains the relative trends observed in the masses, efficiencies, and scaling relations between massive black holes and nuclear stellar clusters.