By Karl Smith
Paul Krugman is warming up to the idea of a safe-asset shortage or at a minimum that the QE might have next to no direct positive effects on the real economy:
What the Fed has been doing recently, however, is something else: buying long-term Treasuries. This looks like this:
Fed purchases of Treasuries
This should indirectly reduce the interest rate on MBS, but maybe not by much. If MBS and Treasuries are poor substitutes, the MBS curve may be very steep; and if Treasuries and short-term assets are close substitutes, the Treasury curve may not move much.
If we can get this far on standard analysis, I don’t think (with very thin trading) complementary between Treasuries and MBS and hence a safe asset shortage is a big leap. We simply have to note that (A) Some financial intermediaries are going to hold long bonds in an effort to mimic the standard borrow-short, lend-long mechanics of traditional banking. (B) Treasuries and private sector bond prices moved in opposite directions in response to the peak events of the 2008 crisis.
Those intermediaries would like to be long solely MBS, because this would offer the highest yield. However, their creditors will fear that will be insolvent in the face of a second crisis. The financial intermediaries could hold a cushion of cash but this would eat quickly into their margins. What’s far better are long-term treasuries. Not only do they pay higher interest rates than cash but their balance sheet value will increase in a crisis, making them a more effective hedge. For this group, cash and long-term Treasuries are not close substitutes, while long-term treasuries and MBS are functional complements.
As long as the long-dated Treasury market is thick, this is not much of an issue. However, as the Fed’s balance sheet grows the marginal demanders of long-dated Treasuries are going to be increasingly the very institutions that see short-dated Treasuries as poor substitutes. Thus the larger the Fed’s balance sheet grows the more likely it is that this complementary effect will or could dominate. And, of course market participants know this and so as the Fed’s balance sheet grows the risk that this will happen grows and so the risk of being heavy on MBS grows.
In this way we can get a safe asset shortage effect even if the instantaneous equilibrium conditions in the market produce a gross substitutability between MBS and long Treasuries.