From my essay on mortgage interest-rate risk:
Interest-rate risk migrates toward those financial institutions that enjoy the highest level of perceived government protection with the least effective form of regulation. The perceived government protection enables them to serve as reliable counterparties to households and firms seeking risk protection. The relatively less stringent regulation leads the risk-bearing institutions to reap gains when interest rates are relatively stable, without having to hold sufficient capital to survive a major shock.
…Suppose that regulators actually succeed in locating all of the bearers of interest-rate risk and holding them to strict capital standards. I believe that the consequence of this will be that the interest rate on 30-year fixed-rate mortgages will rise significantly relative to mortgage instruments where rates are fixed for a shorter period.
Before the Housing Bust I read that the average mortgage only lasts about 7-8 years (because of moves primarily). Wouldn’t that mitigate the premium between 15 and 30-year mortgages?
You don’t get at the fundamental problem with a 30-year fixed rate mortgage: it is unstable because it relies on knowing how inflation will play out over the next 30 years, which is fundamentally unknowable. I don’t see why you think ARMs put too much risk on the borrower. It has to be carried by someone and if they’re going to get the benefit of declining rates, they may as well carry the risk of increasing rates too. The problem is the 30-year fixed mortgage would not exist without heavy govt manipulation of the economy and we’ve seen what that gets us: the recent housing bust and the collapse of Freddie and Fannie. If we can’t finally realize the huge problems with such long-term fixed rate mortgages, it’s hopeless.