They don't have to eat their failures because making them do so is too injurious to everybody else. That is "too big to fail" in a nutshell.
Too big to fail is where the intertwining of things financial differs from say K-Mart, which can go to bankruptcy court and downsize, or an airline which can go to bankruptcy court in order to screw the unions and airline workers.
There is no domino effect to K-Mart or the airlines because the only counterparty risk is consumers of K-Mart being stuck having paid in advance for some delivery and having to suffer thourgh delay and possible small loss. Or with airlines the risk is pilot or mechanical error and death, one flight at a time, and the world's interrelationships are not at risk.
So, banking is different, and the big guys know that and manipulate things. I always viewed the Bush bank bailout [with Barney Frank a willing partner] as these guys saying, "We've kidnapped the world, financially, and if you do not pay the ransom we kill it. Here's a severed finger [Leahman Brothers] to prove we are serious."
Others might view it differently.
Whether the Miller bill will be the answer or a starting point is unclear. It recognizes something needs to be done, and is vague because nobody really knows the best way to bring about a too-big-to-fail reform from the present situation, but everyone with a brain knows it is needed and that Goldman Sachs cannot recklessly continue to rob everybody else by manipulating the market by its size alone and having the most advanced fast trading system platforms.
In every trading market worldwide the big players, (some from other nations [a problem for US legislation]), are big local players - having big player footprints in each and every trading market, regardless of locale.
They can arbitrage, and put in fiie-then-quick-kill order pairings that influence short term prices but where no short term counterparty has a quick enough trigger-finger chance to close a trade before it has been pulled as an offer.
Do the big guys do that, phantom offers? There are things your gut knows, but cannot prove. This seems a question in that category. Can manipulating a market be done, profitably? Turn that one around, if markets were to fit the economic theories of ideal equilibrium markets these firms could not have had the opportunity to get as big as they are. Something gave them an edge. And presumably they used it. Once only? Twice? Get real. They milk any advantage dry.
How do regulators, (on a global basis fitting the reality that the world has only one "economy"), police things, and set rules that optimize efficiency but disarm wrong favortism of some over others? That is the ongoing problem of good regulation, with a companion problem being how do you set up a regulatory system and over time keep the foxes from ruling the henhouse, as a practical matter of money and politics and human nature.
To read of the role of Ellison as a cosponsor, and for text of a letter the bill sponsoring group sent colleagues, THIS MINN INDEPENDENT LINK.
A pdf download of the bill text, THIS LINK.
More info on bill progress, here.
With Ellison on the House Financial Services Committee, he does something, unlike Michele Bachmann on the same committee, who does nothing, but does it with Sean Hennity, in their gossipy dumb Rupert Murdoch now-and-again manner.