Here are the crucial paragraphs:
Above all, the key promise of securitization — that it would make the financial system more robust by spreading risk more widely — turned out to be a lie. Banks used securitization to increase their risk, not reduce it, and in the process they made the economy more, not less, vulnerable to financial disruption.
I don’t think this is just a financial panic; I believe that it represents the failure of a whole model of banking, of an overgrown financial sector that did more harm than good. I don’t think the Obama administration can bring securitization back to life, and I don’t believe it should try.
Krugman's essential point, and why he doesn't want securitization resurrected (is it actually dead?) is that banks improperly leveraged securitized instruments, thereby creating a debt-deflation when they misfired.
If securitization isn't leveraged, does it pose the deathly systemic risk currently enveloping our system? With one exception, no. Take away leverage (taking out loans to bet on the securities) and the Wall Street meltdown doesn't critically endanger the Main Street economy.
Here's the exception: securitization atomizes risk. But if atomizing risk leads you to take more risk, then atomizing risk creates systemic risk.
A good analogy is anti-lock brakes. They're a great invention. The problem is, too many people take them into account in how they drive, taking more risk of tailgating, thereby defeating their purpose. Similarly, once securitization reaches critical mass, then too many people take too many cumulative small risks, creating the chance of systemic meltdown if the risks misfire in non-random fashion (like in a housing bust).
There are other problems as well. As readers of the late Tanta know, splitting securities into tranches creates a tremendous gordion knot of opacity and inability to modify the terms in cases of stress (how many mortgages can get modified when hundreds of owners of the various tranches need to sign off?). And a large part of the problem with Lehman Bros. and AIG was how interconnected all their various CDS contracts were.
Like Barry Ritholtz, I believe the issue of causality is clouded to say the least by the spectacular overindulgence in laissez faire deregulation in the last decade. Imho, securitization is fine so long as (1) it is properly regulated to ensure that the agents creating it bear real risk if it goes wrong when they sell it; (2) it isn't leveraged; and (3) there is not so much securitization in the system so as to create a credit bubble.
As I read Krugman, he believes (3) is inevitable, but he does not marshall any evidence in support of that crucial point. My verdict is: not proven.