When taken across an entire demographic, that seems to be the case, and there are countless number of nature/nuture reasons for why that could be. But if you were to get, say, a racially diverse bunch of quantum physicists in a room together and test them out, they'd all end up scoring about the same.
But there'd likely be fewer blacks than asians taking the test, right?
So while the black population as a whole tends to score lower, it says nothing about the potential of an individual of that group.
Of course that's a given. In such populations, the bell curve holds sway, with exceptional individuals at both ends of the curve.