The safe asset shortage problem is back. Actually, it never went away but drifted into the background as the symptoms of this problem--sluggish growth and low interest rates--became the norm. As the figure above shows, yields on government bonds considered safe assets have been steadily declining since the crisis broke out.
This problem is now manifesting itself in a new form: central banks tinkering with negative interest rates. Many view this development as the latest manifestation of central banks running amok. A more nuance read is that central banks are continuing to imperfectly respond to the safe asset shortage problem. The above figure indicates the downward march of global safe yields since 2008 is a global phenomenon. This decline has occurred outside of QE and prior to negative rates. It is a far bigger development than central banks playing with negative rates.
But many observers miss this point. They confuse the symptom--central bankers tinkering with negative interest rates--for the cause--the safe asset shortage. So I want to revisit the safe asset shortage problem in this post by reviewing what exactly it is, why it has persisted for so long, and what can be done to remedy it.
So what exactly is the safe asset shortage? It is the shortage of those assets that are highly liquid, expected to be stable in value, and used to facilitate exchange for institutional investors in the shadow banking system. They effectively function as money for institutional investors and include treasuries, agencies, commercial paper, and repos. During the crisis many of the privately issued safe assets disappeared erasing a large chunk of the shadow banking's money supply. This happened just as the demand for safe assets was increasing because of the panic. As recently shown by Caballero, Farhi, and Gourinchas (2016), if this excess demand for safe assets is big enough it will push down the natural interest rate below zero. Since the central bank cannot push its policy rate very far past 0%, an interest rate gap will emerge and cause output to fall below its potential. That seems to fit post-2008 fairly well. (HT Josh Hendrickson)
This problem can be illustrated by using a standard supply and demand graph for safe assets, as seen in the two figures below. These graphs are fairly standard except for the second vertical axis which allows me to show that interest rates move inversely with bond prices. The first figure shows the supply and demand of safe asset pre-2008. Note that interest rates are near 5% which is roughly where treasury yields were during that time.
Now comes the crisis in 2008 and two changes occur, as seen in the next figure. First, the demand for safe assets rises given the fear and uncertainty during the panic. This shifts the demand curve to the right. Second, the the supply of safe assets is reduced as many of the private-label safe assets disappear. This pushes the supply curve to the left.
Given these developments, two rather striking results emerge. First, because of the zero lower bound (ZLB), interest rates remain too high. The ZLB, therefore, acts as price floor that prevents the market from clearing along the interest rate dimension. Second, the ZLB also means safe asset prices are too low. It creates a price ceiling that prevents the market from clearing along the bond price dimension.This implies treasury prices have been too low since the crisis!
Sometimes I am asked how can there be a safe asset shortage? Shouldn't safe asset prices simply adjust upward to meet the increase demand for them? The above analysis shows why this cannot happen with the ZLB. The safe asset market has been constrained by the ZLB and prevented from working its market-clearing magic.
Why has this problem persisted for so long? Why are safe assets yields still so low? I am not entirely sure. I suspect the demand for safe assets has remained elevated given the ongoing spate of bad news shocks since the crisis: Eurozone crisis, China concerns, fiscal cliff talks, 2016 recessions fears, etc. Also, some of the new banking and finance regulations is probably suppressing the growth of private-label safe assets. That is has taken this long, though, is a bit puzzling.
So what are the solutions to safe asset shortage problem? One solution is for the government to issue more safe assets. That the 10-year U.S. treasury is now 1.73% seven years after the crisis suggest there is still strong appetite for treasury securities. This is the solution proposed by folks like Stephen Williamson and Paul Krugman. Another solution is for policymakers to 'shock and awe' the public into believing a robust recovery is coming and thereby decrease their demand for safe assets. QE was supposed to do this but was not that effective. Scott Sumner and Michael Woodford's call for NGDP level targeting would also fall under this option. The final option is that policy makers could try to work past the ZLB via negative interest rates. Under this option policymakers would help the safe asset market clear by pushing interest rates and prices to their market-clearing levels.
As noted by Narayana Kocherlakota, the difficulty of doing the first two options seems to be making negative interest rates the default option. I am not convinced it will be very effective given how it is being implemented (very different than Miles Kimball's suggested approach). But it is a response--a groping in the dark by central bankers--to the deeper safe asset shortage problem. And that is something we all need to better understand.