The publishing of academic journals generates billions of profit per year. One could easily decimate these profits and still have a viable industry, whatever definition of decimate one cares to use.
Profit margins doesn't mean much in this case. Most science related work (e.g peer review) is done for free. Typesetting costs around $30 per article [1]. It looks like big part of spending is administration and lawyers for taking down copies of the articles.
Profit margins around 40% if I recall...