Sunday, January 23, 2011
This first chart shows the level of M2, arguably the best measure of the amount of "money" in the economy (M2 consists of checking accounts, currency in circulation, retail money market funds, small time deposits, and savings deposits). The y-axis uses a semi-log scale, so that a constant slope equates to a constant rate of growth, which in this case appears to be about 6% per year, compounded. There are times when M2 grows faster or slower, but on average it has increased about 6% per year, which not coincidentally is only about 1% per year faster than the average annual growth rate of nominal GDP over the past 25 years. A good portion of that faster-than-nominal-GDP growth can be explained by the huge amount of U.S. dollars that are now held by foreigners. When money grows in concert with growth in the nominal value of transactions in the economy, it's hard to worry that inflation is going to rise because there is "too much money chasing too few goods and services."
What about those periods during which M2 has grown much faster than 6% a year (i.e., 2001-5 and 2008-9)? The second chart explains what has happened. There is a strong correlation between big increases/decreases in refinancing activity and faster/slower growth of M2. That's because the refinancing process temporarily creates a lot of extra "money" as money flows into and sits in escrow accounts. Most recently, M2 growth has slowed sharply (the result of rising interest rates since last summer), just as refinancing activity has dropped rather rapidly. The recent slow growth in M2 (2.5% annualized growth over the past three months) is nothing to worry about, because it is most likely the natural result of a decline in mortgage refi activity.
We see a similar story—not much of note going on with the money supply of late—in this chart of bank reserves. As of early January, there is no evidence that the Fed's QE2 program has resulted in any unusual increase in bank reserves, the raw material for any expansion of the money supply. In fact, bank reserves today are almost exactly the same as they were at this time last year ($1.1 trillion). Apparently, the Fed's additional purchases of Treasuries in recent months have mostly been offset by other actions which have drained reserves.
It may be hard to believe, but to date, there is no evidence to be found in the money supply statistics that the Fed's addition of over $1 trillion of reserves to the banking system since Sep. '08 (an 11-fold increase!) has resulted in any unusual increase in the amount of money sloshing around the economy.
What we do observe, however, is that the value of the dollar has declined relative to other currencies, and the dollar buys a lot less gold and commodities (e.g., since Sep. '08 gold is up over 60%, and spot commodities are up 30%). This suggests that there has been an effective increase in the amount of dollars in the system relative to the demand for dollars. It is an excess of dollars in relative terms that is likely responsible for the lower value of the dollar and the higher prices of commodities.
With the Fed holding the amount of bank reserves relatively constant, but at a level that is orders of magnitude higher than we what we observe during "normal" times, the key thing to watch is the demand for dollars, since that is the thing that can change on the margin. If the world wants fewer dollars, then the existing supply of dollars will be "un-hoarded," and banks will want to reduce their holdings of reserves, by using those reserves to support an increase in lending. The result of weaker money demand and more bank lending likely would be faster nominal GDP growth.
I think this process is underway. It's still difficult to see with clarity, but a careful reading of the monetary tea leaves tells us that the economy is "reflating" as we speak—the size of the economy is growing, and the volume of transactions in on the rise.
Posted by Scott Grannis at 12:33 PM