The problem with trying to determine break-even is that, as with Lockeed's TriStar project, many of the 387 revenue airframes weren't 'sold' as such but were placed with airlines under complex financing projects with BAe assuming the risk if they fell-through.
IIRC PSA bought their fleet outright with cash but most of the small operators relied on BAe to corral-together various banks and finance packages. Even Dan-Air had some sort of partial-sale-partial-lease arrangement.
Edit: turns-out PSA didn't do straight-cash either, here's a snippet from Flight:
Payment for the 146s initially is through a leverage lease on the first ten,
arranged by BAe and allowing BAe and PSA to take advantage of US tax laws.
BAe is providing a proportion of debt support on the next ten aeroplanes, as well
as on the Rediffusion phase 3 simulator which PSA will install.
ECGD financing would have been available for the second ten, but overall it is not
appropriate for the full package.
The creation of BAE AMO in the 1990s was an atempt to eke-out the 146's service life and keep the program cashflow moving. Not sure if they ever achieved that before selling-off the entire portfolio, but they did manage to sell-for-cash a fair few formerly-leased airframes.
Given the target of 350 noted up-thread, the construction of 394 airframes and the costs of the RJ and RJX re-launches I'd guess ( as an uninvolved outsider ) that it overall lost money.