Testosterone and estrogen studies are quite interesting in that they are marinated in cultural bias. A lot of studies that historically linked aggression with testosterone or estrogen with mood swings have later been found to be on shaky ground.
Funny enough, estrogen has also been linked to increased aggression just as testosterone. You take a baboon and look at when females act most violently and it spikes just as estrogen spikes. Same for violence in women prisons. At the same time, depression also spikes which further suggest that estrogen causes wild mood swings.
But when we look at more modern studies a new pattern is emerging. For high social status individuals we see the exact same pattern as old studies, but as we go down the social ladder the picture change. For testosterone the current findings seem to link it to increases learned behavior when a individual has their social status challenge. That response in baboons is always more violence, but directed at the nearest individuals below them on the social ladder. For estrogen you see a drop in violence as you go down the social status ladder, while isolationing behavior goes up.
An other relative recent finding is that hormone secretion increase in men after winning social status. Unsurprising that correlated in animal studies when males has the highest probability of reproduction.
Hormones are useful when explaining behavior linked to reproduction, and in that you do get some gender differences. I would however be a bit cautious. A good talk I would recommend is https://www.youtube.com/watch?v=2bnSY4L3V8s&t=30m17s as it explore this a bit further.