Wednesday, July 25, 2012

Safe Assets, Money, and the Output Gap

In the past, I made the case that the shortage of safe assets is really just an excess money demand problem.  That is, the sharp decline in the stock of safe assets that began in 2008 matters because it means there are fewer assets that can facilitate exchange relative to the demand for them.  This relative shortage of transaction assets or money implies a deficiency of aggregate nominal expenditures and can explain the ongoing slump.  This notion of excess money demand is not novel, but what is new and makes this view a compelling narrative of the crisis is our expanded understanding of what constitutes money.

Prior to the crisis, most observers thought of some measure of retail money assets like M2 as an appropriate measure of money.  Thanks to efforts of Gary Gorton (2008), Wilmot et al. (2009), Sing and Stella (2012), and others we now know that a more accurate measure of money should also include institutional money assets that facilitate exchange for institutional investors.  One attempt to measure this broader notion of money comes from the Center for Financial Stability (CFS).  Using their data, one can show that the supply of money has fallen sharply since 2008 and has yet to recover.  By itself, this decline in the stock of money assets implies an unsatisfied demand for money.  Throw into this mix the heightened demand for safe assets arising from economic fears and you have a pronounced excess money demand problem.  One would never know this, though, by looking at traditional measures of money.

The nice thing about this excess money demand view is that it helps shed light on the ongoing debate as to whether there is a large negative output gap.  Since money assets are on every market, excess money demand implies a general glut that in turn should create a negative output gap.  Thus, relative money shortages should be correlated with the output gap.  So is there any evidence for this view? 

Michael Belongia and Peter Ireland provide a clever technique in a recent paper that can answer this question.  They solve for the optimal amount of money assets by plugging in potential Nominal GDP (as estimated by the CBO) and actual trend money velocity (as estimated by the Hodrick-Prescott filter) into the equation of exchange (i.e. M*t= NGDP*t/V*t ).  I reproduce their procedure here using the CFS's M4 divisia money supply1 and come up with the following figure:

The figure shows that the quantity of money assets is currently about 7% below what is needed to generate full potential NGDP growth.  Now if one takes the percent difference between the actual and needed M4 divisia in the figure above--the M4 divisa gap--and plots it against the the output gap you get the following figure:

I find this figure striking.  With a R2 of about 60%, it shows that the output gap typically tracks the M4 gap.  For the recent crisis in particular, it shows the acute shortage of money assets (or excess money demand) is matched by the large output gap.  This figure, then, indicates the excess money demand explanation for the recent crisis is a compelling one.

Now some observers like James Bullards and Stephen Williamsons believe that the shortage of safe assets or money is the consequence of real shocks to financial intermediation that have permanently lowered the productive capacity of the U.S. economy.   The relationship evident in the figure above, however, suggests that there is in fact a large output gap given the significant shortage of money assets.  And even if the shortage were caused by a real shock, there are still policy options that could close the M4 gap.   

First, the government could create more safe assets in the form of treasuries.  This approach, however, is politically controversial as it requires more budget deficits.  It is also not clear to me that this approach would be able to create enough safe assets to completely close the M4 gap.  Second, the Fed could create the incentive for the private sector to start producing more safe assets by adopting a NGDP level target.  Such a target would raise the expected level of future NGDP and, in turn, raise the current demand for financial intermediation services.  That would lead to more privately-produced safe assets and ultimately a recovery. There are ways out of this economic morass.

1I specifically use the CFS' "M4 minus" money supply measure. It had a better fit than the regular M4.


  1. Glad you are back.

    Doesn't QE basically add to safe assets by adding interest bearing bank reserves which are generally less risky than MBS or long duration treasuries?

  2. David,

    In your first graph what accounts for the trend change in the "M4 Divisia Needed for..." line that starts at the beginning of 2009? It seems like this would be the equivalent of the CBO marking down potential output but did anyone do such a thing at that point in the recession? Is it an effect of using the HP filter?


  3. David writes,

    "Our expanded understanding what constitutes money."

    Actually, this is NOT a new understanding of what constitutes money. It's an old one, straight out of Hayek's _Prices and Production_, as noted by a team of Credit Suisse economists in 2009:

  4. David writes,

    "this decline in the stock of money assets implies an unsatisfied demand for money"

    You seem to be missing the other side of any supply and demand story -- the shifting COST of the supply side.

    The cost of the supply side of asset or shadow money during the boom was falsely priced, falsely perceived, and artificially lowered by all sorts of pathological government interventions, government regulations, Fed policies as well as transparency problems in the financial markets.

    In other words asset or shadow money as falsely priced -- at its real HIGHER cost, real demand will be LESS.

    The decline in the stock of asset money -- correctly incorporating its real supply costs & money value -- does NOT necessarily imply an unsatisfied demand for money.

    At a higher cost there will be less demanded.

  5. dwb, glad to be back at it.

    Yes, you are right. But I think the key is getting expectations right. If the public expects the Fed to pull out and eliminate the increased monetary base at a later date, then they won't be seen as a permanent addition to the stock of safe assets. And without a permanent increase the public will not expect a permanently higher level of nominal spending. Thus, no recovery.

  6. John, yes there is a notable change in trend after 2009. Most of it does reflect the CBO's downward revision on potential output (which would by implication also lower its potential NGDP). I suspect that some of the decline in potential output can be traced to cyclical factors and thus can be reversed. So if anything, the 7% number I cited is a conservative one.

  7. Greg,yes, some of the safe asset creation during the boom was not warranted by the fundamentals. But a large portion of safe asset collapse during the bust does not seem warranted to me either. For example, Gary Gorton shows in his work that subprime mortgages were only a small portion of all MBS and CDOs during the panic. Another example is France during the current Eurozone crisis: it was downgraded not because it was inccorectly priced before but becuase of the economic slump brought on by too tight money policy of the ECB.