Mark Roe at Harvard has concluded that in addition to everything else, to big to fail (2B2F) is a petri dish for incompetent insulated management:

Corporate governance incentives at too-big-to-fail financial firms deserve systematic examination. For industrial conglomerates that have grown too large, internal and external corporate structural pressures push to re-size the firm. External activists press it to restructure to raise its stock market value. Inside the firm, boards and managers see that the too-big firm can be more efficient and more profitable if restructured via spin-offs and sales. But for large, too-big-to-fail financial firms (1) if the value captured by being too-big-to-fail lowers the firms’ financing costs enough and (2) if a resized firm or the spun-off entities would lose that funding benefit, then a major constraint on industrial firm over-expansion breaks down for too-big-to-fail finance.

His insight is two fold.

First is the point made by plenty of economists that 2B2F institutions are able to borrow money at lower rates, because, notwithstanding the law, if they implode, their creditors expect to be the beneficiary of a government bailout, because the consequences of not doing so are perceived to be catastrophic.

The second point is far more interesting, and original. He believes that one of the constraints on executive behavior is the potential takeover by any of the many vultures out there (Icahn, Pickens, etc.), and that they are too big to be taking:

These lower financing costs from the too-big-to-fail subsidy are a shadow poison pill — the corporate governance defense that managers and boards have used to ward of unwanted takeovers in the industrial sector. Worse, the shadow financial pill impedes restructurings more strongly than a conventional poison pill. It impedes not just outsiders, as does the conventional pill, but insiders as well — a controlling shareholder where there is one, the board of directors and the CEO where there is no controlling shareholder — even if restructuring the firm would be operationally wise.

James Kwak further expands on this by noting that a corporate takeover is effectively impossible at this scale:

Not so with too-big-to-fail banks. For one thing, TBTF banks are impossible to acquire in one piece: no other bank could absorb JPMorgan, even if there weren’t the rule against a banking conglomerate having more than 10 percent of all U.S. deposits. The other option is to engineer a breakup, which is what all manner of shareholder advocates have been arguing for. But, Roe argues, if being too big to fail is your competitive advantage, that would kill the golden goose. Therefore, the market for control doesn’t work properly, and these behemoths continue bumbling along their way—not just threatening the financial, but doing a lousy job at their job of providing credit to the economy.

So, even if you believe that basic market forces serve to regulate corporate governance, (I don’t) the market breaks down at this scale, and government intervention is essential.