Author: John Phillips
Our data is expected to report employment slightly below but very close to consensus level.
- January 3, 2022, 10:31 am New York time.
In December, the restaurant looked empty due to the decline in the number of diners in the restaurant. This may be a bad sign for the leisure and hotel industry, which has seen a significant decline in the proportion of new jobs in the leisure and hotel industry.
Non-farm employment in December: higher than November, slightly below the consensus level
Our big data-driven "predicted value" of non-farm employment in December is 380,000, higher than the growth of 210,000 in November, and lower than the consensus level of 412,000 at the time of writing this report (10:00 am on Monday, January 1, 2022, New York time).
Overall non-farm employment (institutional survey) had a downward surprise last month, however, due to the sharp increase in employment in the household survey, we still believe the report is strong. This month, we expect some of the strong data from the household survey to be reflected in institutional surveys, but are offset by relatively negative signals in some data and ongoing COVID-19 headwinds. Overall, we believe that non-farm employment will be higher than last month, but it is still relatively low compared to recent months.
The drivers of the predicted value of non-farm employment in December are as follows:
- Online "recruitment" activity: The number of new jobs and the number of new jobs have both declined on the basis of quarterly adjustments. The number of new jobs fell by 6.7%, while the number of new jobs fell by 1.8% (see Figure 1a). In addition to the decline in job levels, we also saw a sharp increase in average jobs in December, jumping to 57 days (see Figure 1b). These data put negative pressure on the model.
Figure 1a: new jobs and new jobs, January 1, 2020-December 18, 2021 (21-day moving average)
Source: DeepMacro and LinkUp/Smart Market Data.
Note: Orange = the survey period for the November report; blue = the survey period for the December report.
Our main source of employment data is about 30,000 U.S. companies on the human resources websites. When a company publishes job ads on its website, we see it as a "new job." If an ad such as this is removed from the website, we will count it as a "job position". New jobs represent the company's demand for labor and can therefore become the main indicator of employment growth. We also found that the total number of new jobs, plus other variables, such as the DeepMacro "economic growth factor" that measures the strength of the overall economic cycle, can explain the number of non-agricultural jobs in the monthly period.
Figure 1b: The average time for employment in , March 2020-December 2021 (days)
Source: DeepMacro and LinkUp/Smart Market Data.
- Catering industry: This month, we continue to track the details of the catering industry. Figure 1c shows employment levels in the hotel/leisure industry, comparing the percentage of changes in the number of restaurant diners compared to 2019 (sourced from OpenTable). For the first time since last winter, we have seen the industry’s employment potentially dropping. This is bad news for employment reports, as the industry has been the driving force for employment reports this year, at least recently (see next section).
Figure 1c: Employment in restaurant diners vs. Hotel/leisure industry, June 1, 2020-December 18, 2021 (% change compared to 2019, unit in 1,000 people)
Source: DeepMacro Inc. and OpenTable.
Note: The blue shaded area indicates the survey period for the December report. The blue dot is the predicted number of hotel employment in December.
- Overall economic growth remains high and has recently shifted from a slightly negative momentum to a slightly positive momentum.While this shift brings some positive pressure to the model, we want to emphasize that the impact is slight.
- Unemployment benefits application count December was relatively flat but low, and continued its long-term downward trend, bringing positive pressure to the model.
The impact of the leisure industry has weakened
In the past year, one of the biggest factors that have affected employment reports is the leisure and hospitality industry. The industry accounts for more than 100% of total employment growth in April, but its impact has since declined steadily, with only 11% in November (see Figure 2a). While the early employment growth of this industry is mostly fairly easy, the employment growth rate in this industry has slowed sharply and many jobs are still open.
Figure 2a. The ratio of employment growth in leisure/hotel industry to total employment growth, April 2021-November 2021 (percentage)
Source: US Bureau of Labor Statistics.
This can be clearly seen in the data. As shown in Figure 2b, we compare the recovery of employment in the hotel and leisure sectors with three main industries: the commercial services industry, the trade and transportation industry, and the construction industry. The first point is that the hotel/leisure industry collapsed much larger than other industries, so the employment growth in this industry also had a greater impact on total employment growth. The second point is that most of the gaps have been filled, so the industry has limited potential to drive further significant growth in overall non-agricultural employment. The third point is that this industry can still drive some employment growth in the process of resuming the decline caused by the COVID-19 pandemic - the employment level in the hotel/leisure industry is still 3% lower than the pre-epidemic level. But it remains to be seen when and how this last gap can be filled; it seems that as long as the COVID-19 pandemic continues, it will be difficult to make up for it.
Figure 2b. Comparison of employment growth in the leisure/hotel industry with the business services industry, trade and transportation industry, and construction industry, April 2021-November 2021 (percentage)
Source: DeepMacro and LinkUp/Smart Market Data.
How to allocate asset positions before data is released: "Mixed and moderate" type report
When the prediction value of DeepMacro is below the consensus level, it is like the non-farm employment forecast in December (DeepMacro): 380,000 vs. Bloomberg consensus level: 412,000), our simple trading rules recommend that go long USD Interest rate / short USD 4/ go long S&P 500 index . Note that these are short-term recommendations, while DeepMacro's portfolio is essentially a medium-term portfolio. Therefore, these suggestions are not direct medium-term suggestions in themselves, but rather that is designed to to help short-term risk management in the event of a significant economic event. In addition, the gap between the forecast value of DeepMacro this month and the market consensus level is only 32,000, so if our predictions are accurate, the market is unlikely to have too many surprises.
However, from the direction, let’s review the interest rate/currency/asset allocation model of DeepMacro and the possible market reactions. DeepMacro short-term interest rate model STR-1 currently predicts that interest rates will rise and exceed market forward rates. This is consistent with the "short selling" strategy (contact us to view). Our forecast for the number of non-farm employment shows that there are certain risks before and after the number of non-farm employment is released. On the forex side, the FX-1 model shorted the US dollar against the euro and yen (contact us to see), so our forecast is consistent with these positions. But from a net value perspective, against other currencies (such as GBP, AUD and Swiss franc ), FX-1 goes long US dollars, so investors may want to cut these positions when non-farm employment is released.
Finally, in terms of stocks, our monthly asset allocation model (target annualized volatility 5%) is currently slightly reducing its holdings (compared to the neutral reference value). The driver of stock reductions is the rise in risk aversion sentiment (at least before the recent light rebound).Good but not-so-hot job growth is a "best of both worlds" situation for the stock market, so investors may want to increase their exposure as non-farm employment is about to be released.
DeepMacro (DeepMacro) was created by Wall Street senior economist and famous IT data scientist . It uses artificial intelligence to analyze a large amount of economic data and quantify it into indicators such as economic growth factor, inflation factor, and global investment risk factor to predict the global macroeconomic trend. The fully automatic algorithm system uses "big data" to analyze economic conditions and obtain data that is important to the market but does not cover well before the official data is released. Based on these indicators, DeepMacro has built medium-term investment portfolios in various asset classes, including short-term interest rates, foreign exchange and global asset allocation, and has achieved good results. DeepMacro works closely with financial institutions around the world, including major banks, sovereign wealth funds and global hedge fund , to provide paid content, proprietary indicators and consulting services. For more in-depth data analysis and professional explanations, please contact [email protected].
