The robo-signing stuff is straight up illegal. But that falls under my original point 1: The people doing that were tiny, tiny fish, who had nothing to do with the CDOs.
> The rating agencies could have been defrauded and if people sold those while not really believing they were of the rated quality IMO it's still fraud.
Legally that's a harder question than you might imagine. However...
...there's absolutely zero evidence of this on a systemic basis. Banks overwhelmingly made it clear that they believed the assets were good. The bailouts went primarily to institutions that had originated utter crap and then held on to it because they thought it was a great investment. Studies have been done to look at the personal investments of bank executives; it turns out they were closely correlated with their job; the more you knew about CDOs and MBSs, and the more closely linked to the system you were, the more you thought CDOs were safe as houses, and the more willing you were to have your retirement invested in it.
Conversely, the small handful of Cassandras were utterly ignored by the world in large part because they were complete outsiders who had no inside knowledge. The arguments boiled down to: "What do they know? The people who know the most about CDOs are all in; these people must just not understand the models."
(There are exceptions in specific cases. One that springs to mind is Fabrice Tourre[1]. But he was very much the exception to the rule, and it's probably not a coincidence he worked at Goldman...)
> It could have all been innocent but it certainly smells and warrants investigating.
And it was. Unfortunately (or fortunately, depending on your view) no smoking guns emerged.
[1]: http://www.reuters.com/article/us-goldmansachs-sec-tourre-id...