MIXED SIGNALS ABOUT THE U.S. ECONOMY (PART TWO)

This is the second part of a blog entry about the difficulty of forecasting what will happen to the U.S. economy in the relatively near future based on current and very recent conditions.  The data that we are receiving may not give a clear indication.  To give an example, the unemployment rate in the U.S. remains low, which could indicate economic growth to come.  However, the unemployment rate is often a lagging indicator that does not rise much, if at all, until after economic growth slows or goes negative.

Further, if my memory is correct, the yield curve based on the U.S. Treasury ten-year interest rate when compared with (1) the federal funds rate, (2) the U.S. Treasury one-year rate and (3) the U.S. Treasury two-year rate all went negative, at least briefly, during the year 2019.  This means that interest rates on debt with more time to maturity were lower than at least one shorter-term interest rate.  Further, the two-year U.S. Treasury interest rate has been higher than the five-year U.S. Treasury rate much of the time since early December 2018, according to data shown on CNBC.  As I pointed out in part one of this two-part blog and as many others have noted before then, the inverted yield curve could indicate a recession coming relatively soon.

My understanding is that an inverted yield curve and other developments similar to a few of the points in this blog are factors in a new index constructed by some at State Street Associates and the Sloan School of Business at MIT estimating that the probability of a recession starting in the next half year reached seventy per cent in November 2019.  For more, refer to sources such as an article written by Ed Adamczyk on the UPI web page (available at https://www.upi.com/Top_News/US/2020/02/05/MIT-study-70-percent-chance-of-recession-within-six-months/6441580933374/).

To follow up on my point about falling money velocity in part one of the blog, if not measured on a seasonally adjusted basis, fourth quarter money velocity may have increased last quarter due to holiday spending.  However, I calculate using M1 money supply data and nominal GDP data both from the Saint Louis Federal Reserve web page, both not seasonally adjusted, that M1 money velocity decreased in the fourth quarter of 2019.  Given that my calculations could be of the ‘back of the envelope’ variety, I cannot be certain that my estimates are correct.  (For example, I calculated simple quarterly averages using the not seasonally adjusted monthly money supply data, failing to factor into the analysis that not every month has the same number of days.)  Still, this possible decrease in the fourth quarter of last year in M1 velocity not seasonally adjusted could be noteworthy. 

With holiday shopping, I would think that usually, money velocity would rise in the fourth quarter of a year compared with the third quarter.  Nominal GDP spending not seasonally adjusted did rise in the fourth quarter of 2019 based on my calculations (using data at https://fred.stlouisfed.org/series/NA000334Q accessed February 26, 2020), but the M1 money supply not seasonally adjusted (available at https://fred.stlouisfed.org/data/M1NS.txt) increased at a faster rate if my calculations are correct.  Also, remember that seasonally adjusted money velocity fell in the fourth quarter of last year.  Readers should realize that money velocity has decreased most quarters after the fourth quarter of 2007.  Falling money velocity raises the issue of how effective expansionary monetary policy has been in completing the recovery from the Great Recession.

U.S. real GDP growth was probably either approximately at its recent trend or above its recent trend during at least part of the year 2019.  That appears good or at least acceptable, except for the fact that trend growth is probably much slower than it was years ago.  Additionally, the release from the U.S. Bureau of Economic Analysis (BEA) available at https://www.bea.gov/system/files/2020-01/gdp4q19_adv_0.pdf shows that after a seasonal adjustment, inventories fell by more than $60 billion in the fourth quarter of 2019.  I am outside of the area of my expertise in seasonal adjustments.  However, does it seem that inventories not seasonally adjusted would normally decrease in the fourth quarter of a year due to holiday shopping?  If so, then does the BEA estimate for inventories from the fourth quarter of last year mean that inventories fell more than normal during that quarter?  If so, then does that mean that sales were greater than expected during the fourth quarter of 2019, or does it indicate pessimism by firms about their ability to sell their products?

Related to that, according to an article on page A8 of The Atlanta Journal-Constitution Saturday, February 15, 2020 attributed to Bloomberg with the headline “U.S. factory output drops on Boeing production halt,” U.S. consumption may have been tepid in January 2020.  The article also notes decreases in both U.S. factory production and total U.S. industrial production last month.  Thus, based on factory and consumption estimates, the article surmised that U.S. economic growth was clearly not at its most robust level last month. 

Moreover, part one of this blog entry noted the BLS announcement that employment estimates were revised downward for recent months, in some cases by more than one half million jobs.  Will that lead to downward revisions to the real GDP series and its reported growth rate and upward revisions to the unemployment rate for recent quarters and months?  The Bureau of Labor Statistics (BLS) has estimated based on household surveys that the number of those employed in a recent month was roughly 158 million.  (https://www.bls.gov/news.release/empsit.a.htm)  Does this imply that real GDP from prior quarters could be revised downward by roughly 0.2-to-0.4 percentage points? 

However, another article (with no listed author) also on page A8 of the February 15, 2020 edition of The Atlanta Journal-Constitution announced that the National Retail Federation predicted a large increase in Valentine’s Day shopping compared to the previous year.  But, CNBC reported relatively large drops in the U.S. stock indexes on February 24^th and February 25^th.

In conclusion, it may be very difficult to predict with high accuracy what will happen in the U.S. economy in the next year or two.  We may need to watch developments as they occur.

(Please realize that subsequent data revisions possibly as early as tomorrow (Thursday, February 27, 2020) could change the analysis of this blog entry.)

MIXED SIGNALS ABOUT THE U.S. ECONOMY (PART ONE)

Trying to get a sense of the short-term future for the U.S. economy could be difficult due to recent potentially mixed signals.  This blog entry will probably be the first part of a two-part blog.  This first part will focus mainly on employment, the yield curve, personal disposable income, and money velocity.

Yesterday (Friday, February 7, 2020), the U.S. Bureau of Labor Statistics (BLS) reported that total U.S. nonfarm payroll employment increased by 225,000 jobs in January 2020 based upon its survey of workplace establishments.  However, using data from the BLS web page, I calculate that employment measured by the BLS household survey fell in January 2020 by roughly 89,000 jobs.  This is the second time in the last three months when the BLS establishment survey reported an increase in total nonfarm payroll employment but the household survey reported a decrease.  Despite two decreases in employment from the household survey in the last three months, readers should realize that the reported decrease for November 2019 was relatively small, so that estimated employment from the household survey for January 2020 is still greater than it was in November 2019.  (Note:  I think that both the establishment survey data and the household survey data for employment are seasonally adjusted.)  Readers can read more about these statistics and releases at the following web pages. 

    

https://www.bls.gov/news.release/empsit.nr0.htm

https://beta.bls.gov/dataViewer/view/timeseries/LNS12000000

https://www.bls.gov/news.release/empsit.a.htm

https://data.bls.gov/timeseries/CES0000000001?output_view=net_1mth

The analysis may be less clear due to downward revisions in previously reported estimates.  The Atlanta-Journal Constitution reported on page A3 of today’s (Feb. 8) newspaper that previous employment estimates from 2018 and part of 2019 overstated job growth by more than 500,000 jobs.

Could the reported decrease in employment from the household survey help to explain (1) the re-inversion of the yield curve based on the two-year and five-year rates for U.S. Treasury debt, and (2) the one-day decrease in stock indexes, both shown on CNBC yesterday (Feb. 7)?  Inverted yield curves could possibly signal a recession on the horizon. 

 

If I remember correctly, Steve Liesman mentioned that a possible explanation for the one-day lower stock indexes was that before receiving the latest employment release from the BLS, some investors may have expected another federal funds rate reduction announcement by the Federal Open Market Committee (FOMC) of the Federal Reserve Bank.  Now that appears less likely due to the establishment survey exceeding expectations for the number of people employed in January 2020.  (The thinking may have been that if another interest rate cut would come in the next two months, then investors who would be essentially indifferent between stocks and U.S. government debt would prefer stocks given that the return associated with newly issued U.S. government debt would likely decrease.)

Perhaps comparable to the recent employment estimates from the household survey, data from the U.S. Bureau of Economic Analysis (BEA) web page show that U.S. real disposable personal income (measured after-taxes in U.S. dollars with constant purchasing power) has fallen in two out of the last three months of data available.  However, the decreases in real disposable personal income are in October 2019 and December 2019 rather than November 2019 and January 2020, as real disposable personal income data are not available as of this writing for January 2020.  Even though U.S. nominal personal disposable income (measured in after-tax dollars that have varying buying power as prices in the economy change) did not decrease in October, November, or December of 2019; decreases in real disposable personal income in two of those three months reduce the amount of goods and services that households can afford to buy in those months of decrease, all other things equal.  If consumer spending is rising more than income, what if anything does that indicate about debt levels in the U.S.?  Readers can see more about the personal disposable income data at the web link below.

https://www.bea.gov/news/2020/personal-income-and-outlays-december-2019

As expected in my earlier blog post (https://harrisonhartman.blogspot.com/2019/12/is-velocity-of-us-m1-continuing-to.html), data on the FRED web page of the Saint Louis Federal Reserve Bank indicate that money velocity for both the M1 and M2 monetary aggregates fell in the fourth quarter of 2019, with velocity measured on a seasonally adjusted basis.  Nominal GDP spending failed to keep up with money supply growth, as has been the case most quarters in the Not-So-Great Recovery.  Falling money velocity, which happened most quarters in the U.S. after the fourth quarter of 2007, raises the issue of how effective expansionary monetary policy has been in completing the recovery from the Great Recession.  I will probably have more about U.S. money velocity and other macroeconomic variables soon.

(Readers should note that subsequent data revisions could change the analysis of this blog entry.)