Author: peter.c.rupert

By Thomas Cooley and Peter Rupert

The BEA announced third quarter real GDP increased 33.1% seasonally adjusted at an annual rate…or 7.4% if not annualized. So, one might ask, “Why does the BEA annualize?” The answer is to make it easier to compare data since some data come out monthly, quarterly or annually. Most countries do not annualize the GDP numbers at all. For example, news reports show 3rd quarter GDP in Germany rose 8.2% and France up 18.2%, these are NOT annualized.

As a quick aside, note that 33.1% is not 4 times 7.4%. What annualization does it assumes that the growth rate for each quarter in the year would be 7.4%, so compounding is involved. The formula to annualize quarterly GDP say is: (((GDP_t/GDP_t-1)^4)-1)*100.

In any event, this was by far a record increase for the US with the next highest being 16.67% in Q1 of 1950. Given the crazy gyrations seen in 2020: Q1 down 5.0%, Q2 down 31.4% and Q3 up 33.1%, where are we relative to where we were at the beginning of 2020? Real GDP in 2019 Q4 was $19.25 trillion and today stands at $18.58 trillion, so we are down about 3.5%.

Over a longer time span, the Great Recession and COVID-19 have pushed our economy well below our long run (linear) trend on a per capita basis. And although the record bounce-back looks like a sharp V shaped recovery there has been significant “scarring” of the economy and there are sectors that simply won’t recover anytime soon – possibly never. Most obvious are the sectors everyone cites – travel, leisure, hospitality, retail, aeronautics, higher education, commercial real estate and fossil fuels. It goes without saying that all of these imply substantial resource reallocation.

Many, many records have been set for the various GDP components. Personal consumption expenditures up 40.7% with durable goods up 82.2%. Gross private domestic investment up 83%, equipment up 70.1%, nonresidential structures down 14.6% and residential up 59.3%. Households are investing in their home-based consumption, buying bigger houses, better appliances, newer cars and so on. But investment in non-residential structures dropped like a rock and shows no sign of recovery.

Obviously the pandemic looks nothing like a “standard” recession. It is easy to see from the graphs that typically it takes years for the economy to reach bottom and then many more years to get back to where it was before the recession hit. During the pandemic things are happening at warp speed. Massive one month or one quarter declines with a rebound the next month or quarter.

The labor market continues to be stressed and with the failure of Washington to deliver more stimulus we may soon see increased labor market pain and a drop in consumption as well. Next week will give us the latest snapshot of labor market conditions.

By Thomas Cooley and Peter Rupert

The BLS reported nonfarm payroll employment rose 661,000 in September and reported 145,000 in upward revisions , 27,000 for July and 118,000 for August. The private sector gained 877,000 jobs while the public sector shed 216,000. Although the headline number fell below a million for the first time since May, the gains were broadly spread with every major sector except government increasing employment. Moreover, the not seasonally adjusted increase was 1,113,700. This is an impressive rebound but there is more to be done.

Average hourly earnings ticked up slightly and average hours of work increased to 34.7. Both May and September average hours at 34.7 are the highest levels dating back to the beginning of the series in March, 2006.

The household survey showed a decline in the labor force of 695,000 and nearly a million fewer unemployed persons (970,000). The unemployment rate fell to 7.86% from 8.42%. After the great recession it took about 7 years for the unemployment rate to fall by half, from about 10% to 5%. The fall to half of the unemployment rate after the pandemic has been fast, about 5 months, it was the government, stupid. Keep in mind also that a drop in labor force participation caused by people opting to sit on the sidelines lowers the unemployment rate as well. The labor force declined by 695,000 and the labor force participation rate fell to 61.4 from 61.7.

The relationship between unemployment and job vacancies is known as the Beveridge Curve. What has been remarkable is that for all of the labor market turmoil, the job openings rate has not changed much. During the Great Recession the job vacancy rate fell roughly in half while during the pandemic it fell slightly and is now near its highest ever recorded. When the job vacancy rate is over 4% the unemployment rate is typically at or under 4%.

Seasonally adjusted initial claims fell somewhat, to 840,000 and continued claims fell to . Once again, however, the not seasonally adjusted claims actually rose slightly.

The overall picture in the labor market is of an economy that is resilient and capable of recovering fairly quickly. Key to that continued recovery is that individuals have the purchasing power to buy goods and services and the willingness to do so. In terms of consumption it is typical during a recession for durable goods expenditures to fall and take time to recover. As the following graph shows, however, the US is largely a service economy. All three components took a major hit with services the largest decline.

To get a better look at what has happened to the components, the graph below indexes the expenditures. Services remains much below its peak before the pandemic. Durable consumption, however, has rebounded sharply, taking only 4 months to recover to its pre-pandemic peak in January, 2020, and is now 9% higher than in January. The Great Recession showed a much different pattern. It took durable consumption about 5 years to get back to its previous high.

While spending has begun to return to prior levels, household debt has fallen for most categories. Given the lower levels of debt, households are in a position to ramp up spending when some of the uncertainty gets resolved. While the personal savings rate has fallen from historic highs, it is still about 14%. Moreover, nonfinancial corporate business debt as a percentage of the market value of corporate equities is also quite low.

The CARES Act went a long way to putting a floor under the fortunes and consumption of the weakest consumers. These “hand-to-mouth” consumers are the service workers who do not have assets to cushion them through an economic shock. Detailed transaction data show that spending by consumers in lower income zip codes has nearly recovered its level in January 2020 but spending in high income zip codes is still way off. The key to the recovery will be the return of well-to-do consumers and that will depend entirely on how safe they feel, i.e. controlling the virus.

By Thomas Cooley and Peter Rupert

The BLS establishment survey showed an increase of 1.371M workers and 1.027M in the private sector for the month of August. The bulk of the gain was in service producing jobs, up 984,000, with retail employment the largest gainer in that category.

Average weekly hours increased slightly, from 34.5 to 34.6. Average hourly earnings rose 11 cents to $29.47.

From the household survey (seasonally adjusted) there was an increase in the labor force and an increase in the labor force participation rate from 61.4 to 61.7. The employment to population ratio increased from 55.1 to 56.5. The number of unemployed persons fell by 2.788M. Thus, the unemployment rate fell from 10.2% to 8.4%. However, there are roughly 13.5M persons unemployed and only 5.9M vacancies, so it will likely take a while before the labor market is back to something like “normal” whatever that may be.

Initial and Continued Claims

The BLS seasonal adjustment methodology for initial claims and continued claims (insured unemployment) has been altered to reflect the widening gap between the adjusted and unadjusted data. The switch from multiplicative to additive factors will bring the seasonally adjusted data more in line with the not seasonally adjusted data. This issue was pointed out in this blog a couple of months ago, maybe they listened to us! The stark comparison can be seen in the latest large drop in the seasonally adjusted data reflecting the new factors. However the unadjusted claims were slightly higher than the previous week.

Overall, the labor market continues to mend. However, the mending is certainly not even across income and sector types. The lower earning employees were those hardest hit and are those that will likely take the longest to recover.

The Shape of the Recovery

While everyone would love to see a v-shaped recovery, the continued growth in employment from its trough is not a harbinger of one. Unemployment is still only a shade above where it bottomed in the great recession. And the mismatch between vacancies and unemployed workers is larger than it was during that recovery. This recovery will most likely be sluggish and a lot depends on how fast people get back to work. However, many of the jobs are gone forever and the speed of the recovery will depend on how fast the labor gets reallocated.

By Thomas Cooley and Peter Rupert

The BEA’s second estimate showed an upward revision of 1.2 percentage points…so that real GDP fell by “only” 31.7% on an annual basis instead of the 32.9 stated in the advance estimate. Private inventory investment and personal consumption expenditures (PCE) decreased less than noted in the advance estimate. Note that what “on an annual basis” means is that if GDP fell by as much as it did this quarter (9.1%) for four quarters it would lead to a 31.7% decline for the year. Make no mistake, however, 9.1% is an extremely large quarterly decline. Real PCE

Initial and Continuing Claims

Initial claims (not seasonally adjusted) fell to 822,000 after increasing the previous week. Continued claims fell to to about 14 million.

Personal Income and Outlays

The BEA released personal income and outlays for July with a 0.35% gain to personal income and a 1.92% gain in PCE. The personal savings rate fell to 17.8%.

While the economy continues its slow march back to something like normality, we are certainly not out of the woods, maybe the trees have thinned some though. Moreover, the stock market seems to have ignored the symptoms and the housing market has sprung back as well.

By Thomas Cooley and Peter Rupert

The BLS announced that seasonally adjusted nonfarm payroll employment increased 1,763,000 in July. The increases in May, June and July are the three largest gains in the history of the series that began in January of 1939. Over the last three months payroll employment has increased nearly 9.3 million. Of course March and April set records for declines, 1.4 million and 20.8 million respectively, totaling a loss of 22.2 million jobs. The May numbers were revised up by 26k while June saw a downward revision of 9k.

These reported numbers come from seasonally adjusted data. As mentioned by the BLS, seasonal adjustment accounts for “predictable” seasonal patterns and there is the problem! -COVID-19 is anything but a predictable pattern:

Looking at the not seasonally adjusted numbers reveals an increase in employment of 591,000. The two series were fairly close in March, April, May and June but, at times, may differ, as in January of this year and the latest numbers. Given the very unpredictable event, the not seasonally adjusted numbers may be more revealing.

The seasonal adjustment problem affects not just the Employment Situation Reports, but as we have pointed out the weekly new jobless claims numbers. In those unadjusted data recent months have continued declines in the new claims while the adjusted data show increases. The financial media almost never make the distinction.

Back to the reported seasonally adjusted numbers, service producing jobs accounted for the bulk of the gain, increasing 1.4 million.

Average weekly hours decreased slightly from 34.6 to 34.5. Average hourly earnings increased from $29.32 to $29.39.

The household survey showed an increase in employed persons of 1,350,000 seasonally adjusted (SA) and 1,681,000 not seasonally adjusted (NSA). The labor force participation rate fell from 61.5 to 61.4 (SA) and increased from 61.8 to 62.0 (NSA). The unemployment rate fell almost a full percentage point, from 11.1% to 10.2% (SA) and from 11.2% to 10.5% (NSA). The teenage unemployment rate fell by almost 4 percentage points, from 23.2% to 19.2% (SA). The employment to population ratio increased a half a percentage point, from 54.6 to 55.1. Before the COVID-19 shutdown the employment to population ratio had not seen a change like 0.5 for about 60 years. The magnitudes of all these changes make it very difficult to interpret the underlying condition of the labor market.

By Thomas Cooley and Peter Rupert

The BEA announced that Q2 real GDP declined 32.9%, pretty much as expected, but still hard to fathom. In fact, due to the government shutdown, these numbers are very difficult to interpret. To what extent has the new surge in COVID-19 cases, prompting more shutdowns in some areas, raised uncertainty of the future path of GDP for Q3 and Q4? It is certainly not out of the realm of possibility to see a 30% increase for Q3! These massive gyrations have made it even more problematic to gauge the underlying health of the economy.

To be clear the 32.9% number is quarter on quarter growth annualized. The quarter on quarter growth is -9.5%. Whichever you prefer, this is the largest drop ever recorded!

The Great Recession seems to have had an impact on the trend rate of growth of real GDP and the current pandemic has moved us even farther from the historical long run trend.

Our view is that the government shutdowns, CARES/PPP and the change in the due date for taxes have made the data impossible to put in historical context. For example, the graph below shows the percent change in real GDP along with the percent change in real disposable income. In general the two move more or less together. The recent shutdowns combined with the recent transfer payments have thrown a monkey wrench into the behavior of these series. While real GDP fell 32.9% real disposable income increased 44.9%. Of course the disposable income numbers also reflect the fact that tax deadlines were moved from April 15 to July 15.

Monthly disposable income and personal consumption expenditures was released by the BEA today (July 31) again showing large zig-zag patterns. Real disposable income in June decreased 1.8% following a 5.2% decline in May, while real personal consumption expenditures increased 5.2% in June following an 8.4% increase in May.

The quarterly declines were seen in virtually every component. Personal consumption expenditures down 34.6%, real investment fell nearly 50% with fixed investment down 30%, exports down 64.1% and imports down 53.4%. Non-defense government spending was up 39.7%.

Initial claims were also announced today and, like in our recent post on initial claims, the seasonally adjusted claims increased while the not seasonally adjusted decreased as they did last month.

The stock market appears to have shrugged off the early signs of illness and has been climbing, with the Nasdaq setting records almost daily.

Equity markets seem to be pricing in a V shaped recovery. We have argued for a short V-shaped response followed by a long slow slog back to potential. A lot of this depends on the consumer and how they respond as the bills come due, how employment recovers, and how the government resolves its current impasse over unemployment benefits and further stimulus.

The Federal Reserve is doing all it can think of to do and it remains to be seen how effective is the accommodation they have made so far. This drama has many episodes and we are only a few months in.

by Thomas Cooley and Peter Rupert

Data on initial claims for unemployment insurance and continued claims (aka insured unemployment) were released today. Many of the headlines took the announcement and provided articles titled Rise in Weekly Unemployment Claims Points to Faltering Jobs Recovery (WSJ) or Weekly Unemployment Claims Rose to 1.4 Million in U.S.: Live Updates (NYT). While technically true these articles are misleading. The reason is that the headline numbers that come out are almost always seasonally adjusted. Here is a graph of the seasonally adjusted data for initial claims starting at the peak on March 28:

As mentioned by the media, indeed claims increased from 1,307,000 to 1,416,000, an increase of 109,000 people. However, seasonal adjustment alters the raw data by applying seasonal adjustment factors. When times are crazy, when something very odd happens, seasonal adjustment can give misleading information. Here is a graph of not seasonally adjusted initial claims:

The actual number of people applying for initial claims actually fell from 1,512,763 to 1,370,947, a decline of 141,816! Moreover there was an increase in initial claims from July 4 to July 11.

So, what is going on? Think about what happens when, say, Christmas is approaching. We know every year that Christmas is on December 25. Knowing that there will be increased shopping in the preceding months, producers and stores ramp up employment to deal with it. The following graph plots payroll employment both seasonally adjusted and not. Note how the not seasonally adjusted (red) data is so regular. October and November build up employment and every January and February it falls. The seasonally adjusted data (blue), on the other hand is quite smooth. Why? Essentially the statistical adjustment accounts for the fact that every November employment goes up in anticipation of Christmas. What the seasonal adjustment does is to compare, say, this November with last November.

Imagine now the government mandated that Christmas would be on November 25 instead of December 25. In anticipation of that change the employment build up would most likely be shifted by a month. However, the seasonal adjustment factor would apply the old seasonal.

This pandemic has altered the traditional employment unemployment cycles. Stores were closed, schools were closed, auto retooling changed months, and so on. To understand where we are then we should really be looking at the not seasonally adjusted data. This begs the question as to why we look at seasonally adjusted data at all. Back to the Christmas example. Is the holiday season a boom or a bust? The only way to think about that is to compare this season to last season and not this December to this November. That is essentially what seasonal adjustment does.

To be sure, given all of the caveats above, we still ain’t doing so hot.

By Thomas Cooley and Peter Rupert

The BLS announced that nonfarm payroll employment increased 4,800,000 in June. The largest increase was in private service producing, increasing 4,263,000. It should be noted however that COVID-19 has had a large impact on the measurement of labor market variables. In particular, the raw counts should be used with caution as there are potentially large misclassification issues. For example, the BLS writes:

In the establishment survey, workers who are paid by their employer for all or any part of the pay period including the 12th of the month are counted as employed, even if they were not actually at their jobs. Workers who are temporarily or permanently absent from their jobs and are not being paid are not counted as employed, even if they are continuing to receive benefits.

Initial claims have come down steadily although the continuing claims have not.

The bulk of the added jobs were in leisure and hospitality (2.1 million) as bars, restaurants and related business began to reopen. Many of these jobs may disappear again as some authorities begin to respond to the dramatic surge in Covid cases by shutting bars, restaurants and other services down again. Retail trade added 758,000 jobs. Education and health services accounted for 568,00 jobs and other services added 357,000 jobs. Many of these service sector jobs reflect re-allocation across firms – jobs lost in retail replaced by increased hiring at Amazon and Walmart. Some of them reflect recall from furlough to employment.

Manufacturing added 356,000 jobs after losing 757,00 since February. Most of these were in durable manufacture and the automotive sector.

Average hourly earning decreased by 1.2% reflecting the fact that the service sector jobs added or reactivated are the lower paid employees.

According to the household survey the labor force increased by about 1.7 million so that the labor force participation rate increased from 60.8 to 61.5. The number of persons unemployed fell by 3.2 million. However, again the BLS warns of data issues in terms of misclassification:

If the workers who were recorded as employed but absent from work due to “other reasons” (over and above the number absent for other reasons in a typical June) had been classified as unemployed on temporary layoff, the overall unemployment rate would have been about 1 percentage point higher than reported (on a not seasonally adjusted basis). However, this represents the upper bound of our estimate of misclassification and probably overstates the size of the misclassification error.

Moreover, the data are as recorded by the interviewers as per usual and the BLS has not altered the data to account for such misclassification…and rightly so.

The increase in employment and the fall in the unemployment rate will give optimists further hope for a V-shaped recovery. But the measured unemployment rate is a still staggering 11.1% as measured and is more likely higher. The unemployment rate including all marginally attached person, plus total employed part time for economic reasons declined from 22.8% in April to 21.2% in May and now sits at 18%.

By Thomas Cooley and Peter Rupert

As expected, the US economy experienced a devastating collapse in the labor market. Payroll employment fell 20.2 million according to the BLS. It is nearly impossible to compare the magnitude to anything in the past. As seen below, it now makes it nearly impossible to discern differences in graphs. Looking at a long history, as in the employment differences graph beginning before WWII shows only small wiggles as the scale has to be changed to accommodate recent events. All major sectors felt the shock. What is most striking from this sudden collapse is the speed at which the labor market has plunged.

A surprising feature of this report is that average hourly earnings increased by 7.9% Y/Y, but this was simply an artifact of the fact the so many low wage workers lost their jobs. It was pure carnage for those in the most vulnerable occupations and also led to the unemployment rate of teens 16-19 rising to its highest level ever recorded. The same forces affected the average weekly hours.

The household survey from the BLS also shows unprecedented changes. The number of unemployed persons now is over 23 million. However, looking at the initial claims data there has been about 33 million individuals claiming unemployment insurance since March 21. The report also includes a disclaimer saying that the response rate for the household survey plunged last month and the payroll survey required an adjustment. This suggests we should should expect big revisions to all these numbers in the coming reports as the BLS tries to fill in the blanks. They won’t be happy revisions!

The BLS is candid in it’s report about the shortcomings of the data for this month:

“….. there was also a large increase in the number of workers who were classified as employed but absent from work. As was the case in March, special instructions sent to household survey interviewers called for all employed persons absent from work due to coronavirus-related business closures to be classified as unemployed on temporary layoff. However, it is apparent that not all such workers were so classified. If the workers who were recorded as employed but absent from work due to “other reasons” (over and above the number absent for other reasons in a typical April) had been classified as unemployed on temporary layoff, the overall unemployment rate would have been almost 5 percentage points higher than reported (on a not seasonally adjusted basis). However, according to usual practice, the data from the household survey are accepted as recorded. To maintain data integrity, no ad hoc actions are taken to reclassify survey responses.”

The coming months will bring the full extent of the carnage into sharper focus. Until then we can only hope that the CARES Act and other programs will slip a safety net under these unfortunate workers.

By Thomas Cooley and Peter Rupert

The BEA announced that 2020 Q1 real GDP fell 4.8% at a seasonally adjusted annual rate. This is the first decline in 6 years after the longest expansion in US history.

As a comparison to the Great Recession declines, the following graph shows GDP changes since 2007.

Personal consumption expenditures took a major hit, falling 7.6% as people are on a forced “staycation”. Durable goods consumption fell 16.1% while nondurable consumption rose 6.9%. The consumption of services fell 10.1%.

Gross private investment fell 5.6% with nonresidential structures declining 9.7% and equipment investment 15.2%. Note that nonresidential investment has declined every quarter over the past year. Residential investment, on the other hand, increased 21.0% this quarter and saw increases over the past three quarters as well. That pattern will reverse itself as the housing market has begun to follow the rest of the economy.

The Q1 data only partially reflect the large shutdowns of business, much of which happened around the ides of March. The data likely overstates consumer spending and understates the decline in other services. We expect the next release to contain substanital revisions to what is in this report.

Unfortunately, the news will only get worse for the next quarter. Signs of opening are starting, but are months off for many states. Estimates of the size of the second quarter decline vary widely but all signs point to Great Depression magnitudes. What the rebound will look like like is the subject of much speculation but a lot depends on the effectiveness of the massive stimulus and the aggressive stance of the Federal Reserve, who reiterated they will use everything in their arsenal to help smooth the hardships experienced by firms and households.

Although governments can shut down an economy as we have just witnessed throughout the world, it won’t be the government who will open it…that lies in the hands of the consumers. How fast consumers will feel safe in returning to stores, restaurants and travel is unknown.