Much attention focused on the large decrease in U.S. real GDP for the second quarter of 2020 announced by the U.S. Bureau of Economic Analysis (BEA) late last week. At least some have also noted that the implicit deflator for the U.S. decreased in the second quarter. Although aggregate supply factors likely helped to cause the decrease in U.S. real GDP, clearly the combined decrease in real GDP and price levels must have been the result of a decrease in aggregate demand. Recessions in general and particularly negative-AD-shock recessions may be periods during which and/or near when money velocity falls. Comparatively little attention has been given to the ongoing decline in money velocity in the U.S., although greater attention to this decline could be helpful.
Data from the FRED web page of the St. Louis Federal Reserve Bank show that the velocity of the M1 money stock fell sharply in the second quarter of 2020 to about 3.9 times per year (at an annualized rate). While it is true that M1 velocity resumed its pattern of decreasing most quarters probably starting with the fourth quarter of 2018, the rate of decline accelerated sharply in the U.S. last quarter. In the second quarter of 2020, the estimated value for U.S. M1 velocity was at its lowest level since the first quarter of 1961. The velocity of the U.S. M2 money stock also decreased relatively rapidly, as noted below.
From the algebraic definition of velocity (V=P*Y/M), what caused U.S. M1 velocity to plummet last quarter? In addition to both the implicit deflator and real GDP (or P and Y in the above formula) falling, the M1 money stock (M in the formula above) skyrocketed. Again using data from the FRED web page of the St. Louis Fed, I estimate that U.S. real GDP decreased by roughly 9.5 per cent. Annualizing that drop produces a decrease in real GDP of nearly 33 per cent per year. The FRED web page of the St. Louis Fed reports that the implicit price deflator fell at an annualized rate by about two per cent last quarter. Thus, the decrease was about 0.5 per cent when not projected over a full year. With real GDP falling by about 9.5 per cent and the implicit deflator decreasing by approximately 0.5 per cent, neither annualized, it may be no surprise that the St. Louis Fed web page FRED shows data that I use to calculate that U.S. nominal GDP decreased at a rate of about ten per cent in the second quarter, not annualized.
My calculations using data from the St. Louis Fed web page (but perhaps originating with the Board of Governors of the Federal Reserve) find that the quarterly average M1 money supply increased by roughly 23 per cent in the second quarter, and that is not annualized. Combining this increase in the money supply with a decrease in nominal GDP spending of about ten per cent resulted in M1 velocity falling nearly 26.8 per cent. This was the largest quarterly percentage decrease in M1 velocity in the quarterly data sample period, which started in 1959. Readers may want to note that U.S. M2 velocity decreased more than 20 per cent last quarter.
I mentioned in my 2015 book, It’s Velocity, Stupid! (short title) the possibility of a money supply level above which the velocity of that monetary aggregate would automatically decrease. I referred to this theoretical possibility as either a non-decreasing-velocity money supply level (NDVMSL) or a maximum constant velocity money supply level (MCVMSL). If the MCVMSL or the NDVMSL exists, then clearly the U.S. M1 money supply was well above it in the second quarter of 2020. It’s Velocity, Stupid! explains why the NDVMSL or MCVMSL must be time varying. To explain why briefly, suppose the money supply exceeded the NDVMSL last quarter, and so the velocity of money fell last quarter. If in the current quarter nominal GDP and the money supply remain exactly the same as they were last quarter, then there is no change in velocity in the current quarter. But that means that the NDVMSL or MCVMSL must have increased to be at least equal to the current money supply. Otherwise, money velocity would decrease again. It’s Velocity, Stupid! also offers preliminary thoughts on how to estimate the NDVMSL or the MCVMSL.
Plummeting money velocity suggests at least two things to me. For one, in addition to expansionary monetary policy, the U.S. economy and likely other economies need more expansionary fiscal policies to try to complete the recoveries from their recessions. It also suggests that inflation will probably not be a problem in the short term until money velocity starts to increase. Let’s hope that the additional $600 of unemployment compensation in the U.S. is extended and implemented right now. That would not only help millions of families, but it would probably also help the overall economy by preventing a further decrease in aggregate demand.
What about future levels of money velocity in the U.S. for the third quarter of 2020 and beyond? If the M1 money supply keeps growing rapidly, then nominal GDP spending will need to grow by an amount at least as large in order to keep the velocity of M1 from falling. Thus, although there is no guarantee, if monetary policy remains extremely expansionary, further decreases in money velocity are quite possible, if not likely, at least in the very short term. Well-designed fiscal policy can help to ensure that dollars trade in GDP transactions – something that appears lacking based on the sharp drop in U.S. money velocity last quarter.
Readers may want to realize that future data revisions could impact the analysis of this blog entry. Thanks for your understanding.
Readers can find data and/or more information at the following web pages.
https://www.bea.gov/news/2020/gross-domestic-product-2nd-quarter-2020-advance-estimate-and-annual-update
https://fred.stlouisfed.org/series/GDPC1
https://fred.stlouisfed.org/series/A191RI1Q225SBEA
https://fred.stlouisfed.org/series/GDP
https://fred.stlouisfed.org/data/M1SL.txt
https://fred.stlouisfed.org/series/M1V
https://fred.stlouisfed.org/series/M2V
www.economagic.com
