An insurance model that uses data on financial stability in a society where a specific race of people have been, through bigotry alone, forced into a less stable position by default, will be bound to perpetuate the original bigoted patterns of behavior. It's not inaccurate to describe the model as therefore racist because it's modeling itself off a risk assessment where certain groups have historically been forced into the fringe and are therefore inherently risky.

I'm sure if black wall street wasn't razed to the ground by a racist white mob (among several other attempts to gain stability, wealth, etc. by black people being destroyed by racist white anger) then maybe the model wouldn't need to "reflect the underlying stats that correlate with race...". But those underlying stats, and those correlations, didn't just happen in a vacuum, hewn out of the aether like a magical and consequentless thing.