Karl Smith hits some of the themes that I've expressed recently:
First, its actually not that easy from a technical point of view. A significant portion of the US debt is in very short term securities, like 90 days. If I remember correctly the bulk of foreign holdings is in securities with a maturity of less than a year. This is because they are using Treasuries as a liquidity vehicle and under those circumstances you want the shortest maturity possible.
So what happens with this massive inflation? It simply drives up the interest rate of Treasuries. You get to screw investors for 90 days and then much of the jig is up.
Second, large domestic holders of Treasuries are folks the US government has implicitly agreed to backstop: pension plans, major US banks, insurance companies, state and local governments. What is the point of screwing these folks out of their money if you are just going to turn around and give the money right back to them?
Third, why is this better than a structured default? Its not as if people are going to be oblivious to the fact that they were screwed. Having done it raises the risk premium on US debt. Not the default risk premium but the inflation risk premium. What do you really accomplish with inflation versus structured default?
Fourth, what do you accomplish period? Structured default scares everyone because of the prospect that it will bring a crushing recession. An effective attempt to screw holders of US Treasury bonds should accomplish the same thing. Whose interest is that in? The Fed? Politicians? Bureaucrats?
I just don't see the internal constituency for saying, yeah lets inflate away the debt.
Never mind the constituency; I don't see the point. In fact, we'd get longer than Karl postulates, because money supply data is noisy; it would take a while for the markets to realized what had happened, and that it was US policy, not a blip. But we wouldn't have much more time--a few years, not a few decades.