There is nothing wrong with it, the only trusted component is (obviously) in terms of the initial distribution of the staked coins, which is the central feature of proof of stake.