This week the BLS will release price data for August, and although we believe it is too little too late, inflation data is getting a lot of attention from market watchers. Investors are hoping that price data will begin to decelerate, and if so, it will ease the pressure on the Federal Reserve to be aggressive in terms of raising interest rates. However, they may be disappointed. A deceleration from the CPI’s 9% YOY pace in June would be nice; yet many are pinning their hopes on the fact that WTI oil futures are down more than 20% from their May close. But this ignores the fact that oil prices are up 33% YOY, a percentage that will continue to add to inflation pressures. Moreover, the geopolitics around energy is complex, volatile, and unpredictable, particularly since the leaders in many oil-producing countries like Brazil, Iraq, Iran, and Russia are facing a multitude of domestic issues, making any long-term forecast for oil prices nearly impossible.
Still, it is clear that inflation data for July and August will be important, and they will set the tone for the next FOMC meeting set for September 20-21. So too will employment data and that is one of our two main topics this week.
The July employment report showed a surprisingly strong monthly increase of 528,000 jobs and an equally surprising dip in the unemployment rate from 3.6% to 3.5%. Given the gloomy backdrop from other surveys such as the ISM manufacturing and nonmanufacturing surveys, this big jump in employment was clearly unexpected.
However, with July’s increase, the total number of people employed in the US rose to 152.54 million, beating the previous record of 152.50 million workers in February 2020 by 32,000. This was important in our view, since the fact that the total level of employment had not exceeded its February 2020 peak was one indication that the post-pandemic recovery was weak. It also helped explain the declines in GDP.
Also in July, the participation rate inched lower to 62.1% while the employment-population ratio rose 0.1 to 60% in July. However, both remained below their respective February 2020 peaks of 63.4% and 61.2%. These ratios show the relationship between the labor force and/or employment to the overall population. What July’s data indicated was that both remain below the 2020 peak levels. See page 3.
One reason for this weakness is that the labor force has been shrinking. In July, those no longer or “not in the labor force” rose to 100.15 million, the highest level seen since October 2021. There can be a variety of reasons for people to leave the labor force, but the percentage of those no longer in the labor force yet indicating they want a job increased to 6.3%, up from the February 2022 level of 5.3%. Separately, discouraged workers jumped from 386,000 to 471,000 in July. Keep in mind that the decline in the labor force is what contributed to the unemployment rate falling in July. See page 4.
Since employment data can have a major impact on monetary policy in coming months, we dug into the data in greater detail and we noticed several interesting things about July’s job report. The establishment survey showed seasonally adjusted employment rising 528,000 to a record 152.54 million, but not-seasonally-adjusted data showed a decrease of 385,000 jobs to 152.25 million. In short, unadjusted employment remained well below its peak of 153.1 million in November 2019. See page 5.
In addition to the establishment survey, the BLS conducts a broader household survey each month. This survey showed July employment at 158.29 million, a bit less than the 158.87 million recorded in February 2020. However, the not-seasonally-adjusted household series showed 159.1 million workers in July, matching the previous October 2019 record. All in all, a deep dive into job data left us questioning whether employment actually reached a record level in July.
And the BLS will add another complication to employment data. On August 24, 2022, the Bureau of Labor Statistics will release a preliminary estimate of the upcoming annual benchmark revision to the establishment survey. These benchmarks are derived from state unemployment insurance tax records that nearly all employers are required to file. A final benchmark revision will be issued with the publication of the January 2023 job report released in February 2023. Since benchmarks adjust data retroactively, it is nearly impossible at this moment to know if employment has really exceeded its February 2020 peak. We are skeptical particularly since seasonal adjustments are done on an active basis and pandemic layoffs have undoubtedly impacted normal seasonal patterns in employment. This may be too much information for some; but since the Fed is required to maintain full employment with moderate inflation, it is important to understand where US employment stands today. We think it could be weaker than the headlines imply.
Employment data is also a tale of the have’s and have not’s. The unemployment rate for workers with a bachelor’s degree or higher was 2% in July, well below the average, whereas the unemployment rate for workers without a high school degree rose to 5.9% in July, well above the headline 3.5%. The US workforce with a college degree has grown from 26% of all workers to 44% in July. It eclipsed all other groups in 2000. However, since the pandemic it is the only group that recovered to peak levels of employment. This means the other 56% of the workforce is yet to recover to pre-pandemic employment levels. See page 7.
Reducing Earnings Forecasts
Our other deep-dive topic is earnings. As we noted last week, financial headlines are full of reports of better-than-expected earnings results for the second quarter, but this too is misleading. Last week we discussed the difference between consensus estimates and whisper numbers. The whisper numbers, primarily among hedge fund managers, were far worse than the actual consensus earnings expectations and from this perspective, earnings were a positive surprise. Nevertheless, the S&P Dow Jones consensus earnings estimates for 2022 and 2023 fell $1.40 and $1.59, respectively, this week. Refinitiv IBES consensus earnings forecasts fell $1.52 and $1.97, respectively. Over the last seven weeks the S&P Dow Jones estimate for 2022 has dropped $7.36 and for 2023 has declined $6.26. IBES forecasts in the same seven weeks fell $4.07 for 2022 and $7.63 for 2023. As a result, the nominal earnings range for 2022 declined and is now $216.88 (S&P) to $225.50 (IBES). EPS growth rates for this year fell to 4.2% and 8.4%, respectively. To adjust for the weakness seen in second quarter earnings, we are lowering our DRG 2022 estimate from $220 to $218 and our 2023 estimate from $240 to $237, and fear that there may be more downside risk to these estimates. See page 10.
Technicals Of all the indices the Russell 2000 has had one of the best performances from its June low. This is encouraging. We used the RUT as a leading indicator at the top and it may prove to be a good predictor of the low as well. Our 25-day volume oscillator is approaching an overbought reading. Bear markets rarely have overbought readings, and if they do, they are brief. Therefore, the current rally may be at a turning point. Without follow through in coming sessions and a solid overbought reading in this indicator, we would label the current advance as a bear market rally. We continue to focus on recession-resistant stocks and sectors. See page 15.
PLEASE NOTE: Unless otherwise stated, the firm and any affiliated person or entity 1) either does not own any, or owns less than 1%, of the outstanding shares of any public company mentioned, 2) does not receive, and has not within the past 12 months received, investment banking compensation or other compensation from any public company mentioned, and 3) does not expect within the next three months to receive investment banking compensation or other compensation from any public company mentioned. The firm does not currently make markets in any public securities.