- The May jobs report is both better and worse than it looks on the surface, but mostly worse.
- Though the data may be unreliable right now, the report showed a small recovery, with 1 out of 7 lost jobs returning. This is not yet V-shaped; curb your enthusiasm.
- One important indicator got better, but several others got worse. State and local government layoffs may become a wet blanket for recovery.
- Nominally, the largest losses had been in restaurants, which saw 26% of lost jobs returning. That other 74% is the biggest single chunk of unemployed remaining.
- Other services that rely on density and volume to make up for high fixed costs will also have trouble coming back.
CNBC, But Dumber Than Usual
This is an old trick of using changes instead of levels to make a recovery look better. Here’s the chart they should have shown:
Another way of looking at it is comparing the relative job losses with the last recession, until now the greatest economic calamity in my lifetime.
BLS. “NSA” = not seasonally adjusted
Or maybe you prefer the unemployment rate as a measurement?
Curb your enthusiasm.
- You have $100, but lose 80% in the market. You now have $20.
- But then your portfolio goes up 80%. Do you have $100 again? No, you only have $36. You have still lost $64.
This is simple arithmetic, something everyone seems incapable of right this moment.
One out of seven jobs returned in May, which is another way of saying six out of seven did not.
On top of that, our data is unusually terrible right now. Either BLS is really struggling with the circumstances, as they claim, or there is something worse going on, because the errors are all on the side of undercounting unemployment.
The Pandemic Unemployment Assistance (PUA) is being counted separately, at the bottom of the unemployment insurance (UI) tables. The current number is 11 million people receiving PUA and other non-UI benefits for the week ending May 16, the same week as the jobs report surveys were done. These overlap to some degree with regular UI claims, but to some degree, that 10 million is additive. We just don’t know.
Complicating matters, the household survey, from where the headline unemployment rate comes, is really a mess right now. These surveys are normally done via in-person interviews, with plenty of staff at BLS HQ and local offices. None of this was true for the April or May surveys, and they had record low response rates in the 70s for both, normally in the 90s.
On top of that, BLS interviewers have made the same coding error two months in a row, classifying some unemployed as “absent from work for other reasons.” That it happened in April is understandable; that it happened again in May is suspicious. That all the errors are understating the true numbers is doubly suspicious.
So here are the “real” unemployment rates, but even they are probably wrong. By how much, and in what direction, we don’t know.
There was actually a bigger decline in the real unemployment rate than in the headline number without the errors added back in, since the error was smaller in May than April. Still, the number now is probably in the mid-high teens. Again, we just don’t really know because of how poor the data quality is right now.
Ironically, if they fix this in the June report, for which the surveys begin this week, the actual rate may come down, while the headline rate goes up. Maybe CNBC will report the error then. This is where we are.
But with what we have, let’s dig into the report and see where the bounce came from.
Unlike most analysts, I have been using the not seasonally adjusted numbers since this all started. Why?
Under normal circumstances, the adjustments are important, because without them, every January and July would look like recessions.
But never has seasonality been less relevant. The adjustments are multiplicative, not additive, so they can be very large when we are dealing with large numbers like we are now. The adjustments in May are small, but earlier in this, adjusted initial UI claims were overcounting by 2 million in the first 2 weeks.
Also, I want the answer to the question, “How many Americans were employed in mid-May?” The not seasonally adjusted numbers answer that question, or at least try to.
Problem for a later date: the seasonal adjustments are based off previous years’ NSA numbers. They are going to have to come up with some way of adjusting the adjustments for all the highly unusual data coming through in 2020. If they don’t, Q2 2021 will look like the greatest economic miracle of all time, only to be crushed hard in Q3 and Q4.
The Household Survey
Again, take this part with a grain of salt, because this is where the big errors and low response rate were. I have adjusted the numbers where I could, but BLS only identified the errors as being about 5 pp in April and about 3 pp in May. Even they don’t know.
We’ve already seen the top-line number, about 16% unemployed adjusted for the error in May, but there are some key indicators I like to keep an eye on in the details. First, the participation rate:
This is the percentage of the working age population that has a job or tells interviewers they want one. I’m showing you that whole chart since 1965 so you can see the grand sweep of the trends. The rise from 1965-2000 is primarily due to women entering the workforce, but also to the population bubble of the Boomers aging in. But in 2000, the oldest boomers were 54, and the participation rate for 55+ is much lower than for the general population.
After leveling off at the end of the previous cycle at around 66%, the participation rate kept declining all the way through 2015. Two things were happening. First, Boomers were beginning to age out of the workforce. But also, we saw the unprecedented rise in “marginally attached” workers and a subset of that, “discouraged” workers
Marginally attached workers have had, or looked for a job in the past year, but not in the past 4 weeks. Discouraged workers are a subset who have told interviewers they want a job but have given up looking, at least for now.
As you can see, those numbers kept growing through 2011, and really put a ceiling on the recovery. Let’s zoom in on 2020:
The labor force participation rate ticked up from 60% flat to 60.7% in May, but still down from 63.3% in February.
Another key indicator from the household report is going to be temporary and non-temporary job losses. Ideally, the job losses now being categorized as temporary will stay that way and the jobs come back. What we don’t want to see is those job losses switching categories to non-temporary. That has begun, but only a little so far:
Right now, that’s only 300,000 workers switching categories, but we want to keep a close eye on this number going forward.
- The household survey is a real mess. The actual unemployment rate is likely in the mid to high teens.
- Amongst key indicators, only the labor force participation rate is going in the right direction. It is still very early in the game, however, and that can easily turn around.
The Employer Survey
The employer survey should be more reliable. The response rate was off a little, but not like the household survey, and there were no giant errors, at least that were reported. Here we see a small bounce of just under 3 million jobs, after losses of 21 million from February to April.
A big caveat here is that the last few years have seen giant annual revisions to these numbers, all downward, so these are likely overstating absolute levels.
For starters, there is a divergence in the report, as gains of 3.6 million private-sector jobs were offset by another 662 thousand government job losses, now totaling 1.6 million.
The reason this concerns me is this:
That’s state and local government employment added together. These governments cannot run huge deficits like the Fed, and were forced to continue reducing their workforces through 2013. This was a huge wet blanket on the last recovery.
You see what is happening now on the far right of the chart.
- The employer survey is much more reliable right now than the household survey. Even so, there is evidence that it may be way off as well. We just don’t know right now.
- It shows job losses of 21 million from February to April. Then May returned 3 million of these jobs, about 1 out of 7.
- Private sector and government employers are going in opposite directions, with private employers adding 3.6 million jobs, and state and local governments pulling out another 662 thousand.
- Continued job losses in the public sector were a huge wet blanket on the last recovery, so we need to keep a close eye on that.
So Who Got Their Jobs Back, and Who Didn’t?
We get two looks at this, one from each of the surveys. However, we can’t adjust the household survey’s occupational splits for their error, so that’s pretty unreliable.
But the employer survey seems more reliable this month, and always more detailed.
By percentage, by far the biggest gainer was construction, almost getting 80% of job losses back. As you can see, by nominal numbers, the biggest return was to restaurants and bars, but this was also the largest category of job losses in the first place. 74% of restaurant and bar workers have still not come back to work. There can be two interpretations of that data point:
- This is low hanging fruit, and additional job gains will be much harder to come by.
- There is still a lot of room to run.
You can probably guess how CNBC reported this data:
Employment in the entire category is down 39% from February levels. Restaurants and bars saw 26% of job losses return, but the rest of the category lags far behind.
Unsurprisingly, CNBC missed the real big turnaround in the report, which is construction. Also, when we dig down on healthcare, we see that some categories had very nice turnarounds:
Other than dentists, these are much smaller job losses than leisure and hospitality services. But those first 3 columns combined saw a 29% rebound off the bottom, almost twice the restaurant bounce. I expect the rest of the categories to the right to follow. This is the one area I think we can be pretty confident that we will see a full recovery soon, even in the absence of a safe, effective vaccine.
But let’s turn to a few other areas that have giant job losses that I don’t believe are coming back soon, or may even get worse. The first place is in passenger transportation. Like in the leisure and hospitality services we already looked at, these are businesses that have very high fixed costs, and that rely on density and volume to make up for it. Full capacity will not be possible until there is a safe and effective vaccine.
Another area that worries me is business services, the recovery in which will be dependent on the recovery of the economy as a whole.
The big loser so far is employment services, which is almost entirely temp agencies. There was a tiny bounce there in May, but companies will have enough trouble hiring back their full time employees, and temp hiring may be very slow for some time.
The left column contains a wide range of business services, with the biggest losses coming from accounting offices, which doubled in May.
My fear is that this entire category is people who could have easily worked from home, but still 2 million of them were laid off. That is not encouraging for the future. As you see, employment in the left column has only declined 6% since February, one of the lowest numbers you will see. My fear is that it gets much larger, and many of the people now working from home wind up laid off like that 2 million.
Finally, state and local governments, which we have already discussed, but the splits there:
- The biggest single chunk of job losses was in restaurants, and also the biggest nominal recovery. Still, 74% or restaurant workers have not returned to work.
- The biggest percentage recovery came in construction and several healthcare categories. We can expect a full recovery in the latter, even in the absence of a safe, effective vaccine.
- Like restaurants, other hospitality and leisure services have high fixed costs, and make up for it with density and volume. Returning to full capacity cannot be done in a safe way until there is a vaccine.
- Passenger transport services also have the same business model and will be as challenged to return to full capacity any time soon.
- Business services will rise or fall with the rest of the economy. What concerns me here is that 2 million people were laid off, even though they could have worked from home.
- State and local government employment can be a key hindrance to recovery as it was in the last recession.
About That Vaccine...
I keep using the qualifier “absent a safe, effective vaccine” for a reason. This is the only thing that is truly a game-changer. At best, we could expect it by the beginning of 2021.
I am the last person to ask about the medical aspects of this, but I do want to comment on one thing. Moderna (MRNA) has received a lot of attention for its vaccine candidate, and it is now one of the 5 finalists in the US. When I went to look at their SEC filings, something popped out immediately:
I have never seen so many Form 4s, insider transactions, in such a short period. Surely, management must be buying in anticipation of a successful outcome.
Quite the opposite. Rather than hoarding every share they can get, Moderna leadership was selling like crazy in May, and now in June.
- CEO Stephane Bancel sold $3.2 million in shares from his children’s trust.
- President Stephen Hoge sold a modest 20,000 shares for $1.2 million.
- But the big winners were CFO Lorence Kim and the Chief Medical Officer Tal Zaks. Three Tuesdays in a row in May, and again this past week, they exercised options, and instead of keeping them, sold them all. Dr. Zaks netted $27 million and Kim an astounding $50 million. All in a 3-week span from May 12 to June 2. Nice work if you can get it.
- The company’s Chief Medical Officer, Dr. Zaks, currently holds zero shares, and has since May 5.
Does this sound like a company with a golden ticket? Please be careful here, and take all press releases from pharma companies that reference C19 with a grain of salt.
What’s The Actionable Analysis?
This is one of the notes I get from Seeking Alpha editors on my macro articles. Under normal circumstances, they are right - sometimes I forget what the implications of the macro landscape are for the stock market, which is what this site is about after all.
If in January, I had told you that the unemployment rate, then at 3.5%, was going to rise to 16% by May, your response may have been two-fold:
- Are you crazy?
- That would be the biggest calamity of all time.
Both would have been true in January, and they are still true now.
Now, suppose I told you that an auto rental company, the largest, had declared bankruptcy. Their bonds are secured by their fleet, which is worth a lot less today than it was when those bonds were written. There is a fire sale of used autos coming, when there are already giant inventories and declining prices. It’s a disaster of the first order; a perfect financial storm.
But not to the folks on Robinhood:
I am of course talking about Hertz (HTZ). What you see in that chart is the Robinhoodies driving the price of a disaster up 190% over 2 days.
So I have no actionable analysis for you until this fever breaks, and people realize that 16% unemployment is a bad, not good thing.
It’s not just ignorant, but also dangerous:
To be honest, the audience for my macro articles has declined dramatically since this began. People seem more interested in analysis that looks like it was based off a 10-second skim of the first paragraph of the FactSet report.
The purpose of these articles going forward will not be to provide actionable analysis, because I have none in an environment that totally eschews math and logic. For those of you left, I will report what I see in the data, and you can make your own judgements on what to do.
See you at the bottom.
This article was written by
Confirmation Bias Is Your Enemy.
Tech and macro. Deep analysis of long term sectoral trends, and the opportunities arising from them. I promise not to bore you. Author of Long View Capital, a Marketplace service for long-term investors. Risk Factors: I am also wrong sometimes.
Analyst’s Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Seeking Alpha's Disclosure: Past performance is no guarantee of future results. No recommendation or advice is being given as to whether any investment is suitable for a particular investor. Any views or opinions expressed above may not reflect those of Seeking Alpha as a whole. Seeking Alpha is not a licensed securities dealer, broker or US investment adviser or investment bank. Our analysts are third party authors that include both professional investors and individual investors who may not be licensed or certified by any institute or regulatory body.