You wouldn't expect to see difference because the members of the different "races" as you say, are all carefully selected for high performance to get into Oxford. Those at Oxford, Cambridge, or the Ivies in the U.S. are a tiny, highly selected (biased) sample, not representative of the general population, at all. As you have described your qualifications in previous posts, you should know that.
Carefully selected on the basis of performance... exactly. You're making my point in an obverse sort of way. Everyone at Oxford is selected on the basis of high performance - those whose innate abilities, developed habits, parental expectations, reward structure (and a long list of other "et ceteras") have resulted in their achieve admission. No inferences about the race-innate ability connection can be inferred, because of all the other factors.
Applying the same logic, groups of persons perform, on average, worse or better on IQ tests in public schools, than others. A whole bunch of other factors are thrown in the mix, with no attempt to factor them out separately: preparation, home stability, behavior habits, early childhood malnutrition/trauma, cultural and parental expectations, emotional state (another long list of et ceteras).
Part of science is to draw out cause and effect. Every item in the list above needed to be studied and factored out carefully before any "analysis" could be done. And the quantity "IQ" is still an ill-defined measure of innate ability.
A very good book on this subject is Stephen Jay Gould's "The Mismeasure of Man", in which he explores in depth the fallacy of making "linear" extrapolations from a multidimensional system such as human intelligence.