Liz Warren is a dangerous woman, because she’s willing to say what others only imply or are in denial about: we may have to break up those institutions "too big to fail."

On MSNBC’s Rachel Maddow Show Tuesday night, Warren was asked about the implication of having financial institutions like A.I.G. and CitiGroup "too big to fail" when we can’t figure out how to save the financial system working through them. Should we prevent them becoming that large? Her response was unambiguous:

It not only makes sense; it’s absolutely essential. . . . We cannot be in a position where the American taxpayer is essentially the implicit insurer for anyone who gets big and complicated and can’t be regulated. That just can’t be how we run our economy.

If we do, then the taxpayer is hostage to every large institution in America.

That answer has been hiding in plain sight ever since the bailouts of Wall Street began last Fall, but few if any senior officials have been willing to say it straight out. But now the logic is penetrating into the public’s growing anger about the bailouts, and it’s hard to see how the Administration or Congress can avoid it.

That answer brings much needed clarity to the debates about whether and how to deal with toxic assets, and systemic risks, and whether/how to nationalize/take over/pre-privatize the zombie banks. If our solutions aren’t working to create a more manageable and less risky set of financial institutions, then what are we doing? Why are we giving unmanageable and dangerous institutions another bailout?

Warren is not alone; she’s just more explicit. Bernanke’s speech Tuesday acknowledged that federal regulators do not have sufficient means for regulating financial derivatives, don’t have the means to take over mega-financial institution or regulate the far-flung activities of an A.I.G.

These are the activities that AIG warned were so pervasive that AIG’s failure would jeopardize not merely banks and other counterparties to credit default swaps, but the entire insurance system, the funding of municipal entities [and their transit systems] and so on — all the functions we were told last fall were not at risk. Now AIG’s leaked presentation says it’s all at risk.

Simon Johnson at the econ blog, The Baseline Scenario, draws the right conclusions:

Perhaps the financial situation – e.g., in and around derivatives – really is too complex for anyone to understand, unless they have the inside knowledge of regulators. This would mean, of course, that going forward no one can question Treasury about anything important. But that, in turn, makes congressional oversight impossible – even if we move to closed door hearings.

And it raises the question: if our financial system has become so economically complex that President Obama is right, then is it also too complex to be politically sustainable?

Big financial players now know they have a colossal potential put or bailout option. They can also construct interconnected structures that no one can understand, except possibly the Treasury. So every 10-20 years (or more often?) we will experience a crisis of current proportions?

There is a growing consensus that large banks should be broken up; no more “too big to fail”. But the President’s implied point about economic/political complexity suggests that derivatives – for all their obvious potential benefits – are too dangerous to be allowed at anything like their current scale. Who will be willing down the road to let Treasury, without outside comment or oversight, repeatedly provide massive amounts of resources to financial system insiders?

Derivatives have the potential to create a rent-seeking structure that is unparalleled in human history. No society can afford to allow that kind of financial system to operate. Either we figure out how to make it much more transparent – and amenable to outside review – or the re-regulation process currently in the hands of Senator Dodd and Congressman Frank needs to consider more radical alternatives.

It’s time to think anew about what Treasury is trying to accomplish . . . or we need a lot better explanation than we’ve gotten so far.