When people talk about journals as 'black boxes' usually they mean things like how reviewers are chosen, and the decisions made by editors. We don't talk much about journals as capitalist ventures, beyond decrying the cost of subscriptions (see my posts on Open Access) or predatory journals. But there are many models of journal ownership and management, and it's important to understand how they work. Here are five with which I am acquainted. I'd be interested to know of others, or for corrections to my models.
Model A: the Learned Society journal: this is a journal that is published by a learned society or disciplinary group. Ownership is vested in the professional association, who contract with a publisher to publish the journal, and with a group of academics (often based in one department) to edit it for a number of years. Some journals stay in certain foundational departments, more often they move around, and departments are invited to bid for them periodically. Learned socities often use journal funds to pay for their administrative costs, run conferences, and other activities of benefit to their members. Possible examples*: BJPS, ASR
Model B: The Association Journal: whereby a collective body or organization owns the journal, but editors are selected, rather than solicited. A contract with a publisher ensures that the journal is published. Funds again usually go to both administrative costs, editorial costs, and to support other endeavours of the owning group. Possible examples*: African Affairs, Africa
Model C: the Collective Model. Ownership is vested in the editorial collective, and decisions about editors and often about papers are made collectively. Both journals that I know which run on this model date from the 1970s. Again, a contract is negotiated with a publisher, and money is often used to run conferences, sponsor academic travel etc, as well as cover the costs of editing the journal. Examples: JSAS, ROAPE
Model D: the Private Ownership model. The journal is owned by an individual and profits accrue to them. That individual might contract out the editorial work, or take it on themselves. Possible examples*: TWQ
Model E: the Publisher-Owner model. The publisher 'owns' the journal and contracts with an editor or editors to run the journal. Possible examples*: JMAS
In the case of models D and E, I presume that proceeds of the journal pay salaries and other expenses, but are less likely to support conferences and other activities, but I may be wrong. Certainly editors could negotiate such agreements into their contracts, if publishers were amenable, but I'm not aware of examples that do this. It's also difficult to know whether journals are Model D or E, since we're not usually privy to their contractual arrangements.
How any of these groups relate to their editorial boards is generally unrelated to their ownership, with the exception of Model C, where the editorial board is resposinble for both organizational and editorial decision making. In Models D and E editorial boards usually serve at the pleasure of the editor.
Finally, peer review is unrelated to both ownership and management.
*This is based on the best of my knowledge. Please let me know if I'm wrong, so I can correct it.