Markets aren't efficient, except when I want them to be

It really saddens me when Brad DeLong uses an obviously moronic argument. That is not a thing that often happens. But here's the proof:

And [Niall Ferguson writes]:

Last, crucially, explosions of public debt incur bills that fall due much sooner than we expect. For the world’s biggest economy, the US, the day of reckoning still seems reassuringly remote. The worse things get in the eurozone, the more the US dollar rallies as nervous investors park their cash in the “safe haven” of American government debt. This effect may persist for some months, just as the dollar and Treasuries rallied in the depths of the banking panic in late 2008. Yet even a casual look at the fiscal position of the federal government (not to mention the states) makes a nonsense of the phrase “safe haven”. US government debt is a safe haven the way Pearl Harbor was a safe haven in 1941.... Only two things have thus far stood between the US and higher bond yields: purchases of Treasuries... by the Federal Reserve and reserve accumulation by the Chinese monetary authorities. But now the Fed is phasing out such purchases... the Chinese have sharply reduced their purchases.... Small wonder Morgan Stanley assumes that 10-year yields will rise from around 3.5 per cent to 5.5 per cent this year...

To this the right answer is: perhaps, but probably not. Look at the yield curve [on U.S. Treasuries]...

If ten-years go from 3.5% to 5.5%, then 30 years will go from 4.6% to 6.2%--and holders of thirty-year bonds will lose forty percent of their wealth. The private holders of thirty-year bonds, and there are many of them, are all making a very large and powerful bet that Morgan Stanley and Niall Ferguson are wrong.

Brad DeLong's argument is this: "Bond markets, as of February, 2010, do not expect America to default on its sovereign debt in the next 30 years. Therefore, America is not likely to default on its sovereign debt in the next thirty years."

Now change the wording a little bit: "Bond markets, as of February, 2006, do not expect American housing prices to fall in the next 30 years. Therefore, American housing prices are not likely to fall in the next 30 years."

In other words, DeLong is using the strong version of the Efficient Markets Hypothesis to justify his blithe lack of concern about sovereign default risk. Yes, the same EMH that he routinely derides (second link). This is a sad thing to see.