[belorn]

Looking at a few studies, the data is much more complex. For example, higher levels of circulating testosterone were associated with lower risk aversion among women (r = −0.1793; P = 0.01), but not among men (P = 0.11). However at comparably low concentrations of testosterone the gender difference in risk aversion disappeared, suggesting that testosterone has nonlinear effects on risk aversion regardless of gender.

Or to make a summery of the summery of the study: there is an association of risk aversion when going from low levels of testosterone to very low levels, but "male with high testosterone" is not supported in that study as increased risk taking behavior vs males with normal levels of testosterone.

(https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2741240/)