The most obvious case of this is how, perversely, the intensive demand for AAA securities has been universally taken as a sign of excess, because the institutions that are organized around meeting that demand happen to have financially leveraged business models. But, most of the owners of that debt aren't leveraged. In fact, this is acknowledged in parts of the conventional narrative about the period. The global savings glut wasn't about leveraged savers. Even domestic money markets were not leveraged. Remember how it was a big deal that some funds were down by a percentage point or two and "broke the buck"? There was trillions worth of savings that was not leveraged at all, and this is simply thrown into the memory hole whenever it needs to be ignored in order to keep believing that everything was about excess and recklessness.
As I have pointed out, homeownership peaked in early 2004. I have tried to demonstrate in various ways how there was a massive outflow of home equity into low risk securities. The appearance of synthetic CDOs in 2006 and 2007 should have been a stark warning that accommodation was desperately needed. But, this also was misinterpreted as excess and speculation.
Maybe this is the graph that finally presents the point as I see it. Here, both the change in home equity and the change in homeownership rate (right scale) are inverted. From 2005 to 2007, there was a massive outflow of unencumbered and lightly encumbered homeowners from the housing market. And they piled all of their cash into things like money market funds.
The reason is that risk aversion had become so strong that savers weren't even willing to take on home equity. New homebuyers tended to be leveraged. This is partly because lifecycle effects create a natural stable flow of first time homebuyers in all markets. But, also, because of risk aversion home buying with leverage is basically a form of call options. And, again, call options can be framed as a tool of speculation or of insurance. We chose to view it as speculation. I have come to see it as insurance.
In this graph, we can see the parallel movements out of homeownership (and home equity) and into money markets in 2006 and 2007. The housing ATM idea has it backwards. Homeowners weren't using their homes to make withdrawals. They were using their homes to make deposits.
After 2007, we wiped out homeowners, so since then homeownership has continued to fall, but former homeowners have nothing to show for it.