"But they are not great at doing that in the absence of trust and in the presence of tremendous fear."

James Suroweicki on the Lehman question:

"I mean that creating environments in which panic flourishes—like the environment that followed the collapse of Lehman—and trust vanishes leads to decisions and choices that are different, and significantly worse, than those that would have been made in calmer times, even if those times turned out to be bad ones. Markets are great at incorporating information quickly. But they are not great at doing that in the absence of trust and in the presence of tremendous fear.

Dixon and Cass’s implicit assumption is that since the bad news and the deleveraging had to happen eventually, it didn’t really matter (or it was perhaps even better) that it happened in a month rather than a year. In other words, they assume the order in which events occur doesn’t, in the end, matter that much. I think this assumption is simply wrong. In fact, the evidence suggests that sequencing matters a lot in markets, which means that because Lehman collapsed the way it did we’ve ended up in a very different place than we would have if the government had stepped in to save it."

I think that this is a very good argument. The reaction to Lehman, and following avalanche, was not a good thing, and should have been prevented, especially since the government ended up intervening anyway. It's hard to see a panic as good, since it generally leads to very poor decisions.