That really isn't an accurate statement. Nobody knows. The research is very sparse. For instance, in this very recent review, a pro-trans-athlete scientist makes the case that trans women should not be excluded because research is too poor to know and they shouldn't be excluded unless the research gets good enough to show that they have advantage:

"Is there evidence that transgender women athletes have a physiological advantage? Not according to Eric Vilain, a geneticist at the University of California, Irvine, who specializes in gender-based biology. Very little research has been published on transgender athletes, and what has been published didn’t provide enough results to create evidence-based policies, says Vilain, who does not identify as transgender. “It’s not black and white.”"

(Emphasis mine. Note that he's in favor of allowing trans women to compete as women.) Source: https://www.science.org/content/article/world-athletics-banned-transgender-women-competing-does-science-support-rule.

Also, meldonium, for instance, was banned because its effects are plausibly of the sort that might enhance performance and athletes were taking it, presumably in the hope that it would enhance performance: https://www.usatoday.com/story/sports/olympics/2016/04/05/meldonium-experts-wada-performance-enhancing-drug/82663156/

The issue is what to do in the case where the data isn't clear. Some biophysical measures stay in a more masculine range, which suggests there could be an advantage. Some sporting bodies say, "First, prove there's no advantage." Others say, "It's fair unless it's proven there is an advantage."

With potential performance-enhancing drugs, generally the bar is shifted pretty far towards assuming a drug is guilty unless proven innocent. People make arguments that the social cost is such that the burden should go the other way here, but you at least need to firmly make the argument that it's (1) plausible but (2) not shown and (3) in this case, trans women should be allowed to compete.