Initial Claims Interactive
Figure 2
Figure 3
17-Feb-2024. How the claims cycles coincide with recessions and S&P500 bear markets? In previous cycles, Initial Claims rising from the bottom of the cycle practically preceded the beginning of the recession and then the bottom on the S&P500. In the current 2022-2024 cycle, Initial Claims increased from the bottom (September 2022) to February 10, 2024 by only 30k (and it took 16.5 months). For example, in the 2006-2008 cycle from the bottom of Initial Claims (28-Jan-2006): - until the recession start, Initial Claims increased by 68k (in 22.5 months), and - to the bottom on the S&P500 Initial Claims increased by 378k (in 37.2 months). In the 2000-2002 cycle from the bottom of Initial Claims (15-Apr-2000): - until the recession start, Initial Claims increased by 134k (in 11 months), - to the bottom on the S&P500 Initial Claims increased by 146k (in 29.8 months), but in this cycle we first had a peak on Initial Claims and only then a low on the S&P500. In the 1989-1991 cycle from the bottom of Initial Claims (21-Jan-1989): - until the recession start, Initial Claims increased by 85k (in 17.7 months), - to the bottom on the S&P500 Initial Claims increased by 153k (in 20.4 months). All the previous 7 cycles are presented in Figure 1. The chart is a bit difficult to read, so Figure 2 shows only the ranges of the (i) beginning of the recession, (ii) the bottom on the S&P500 and (iii) the peak of the Initial Claims upward trend - for each of the last 7 cycles. However, the most interesting is Figure 3, which shows the same as Figure 2, but without the two outliers (in the case of the Initial Claims low, we omit the one from 1981; and in the case of the S&P500 low, we also omit the one from March 2009). The current cycle is closest to the one from 2006-2008, when the Fed's pause was also very long (due to the strong economy). Back then, over 22 months passed from the bottom of Initial Claims to the entry into the recession (the longest of all 7 cycles). In the current cycle, 22 months will end only in July 2024.
2024 vs 2006
2024 vs 2000
2024 vs 1989
2024 vs 1981
2024 vs 1978
2024 vs 1973
2024 vs 1968
2006 timeline
2000 timeline
1989 timeline
1978-82 timeline
1973-75 timeline
1968-70 timeline
US vs Canada unemployment
8-Dec-2023. US Unemployment Rate (U3). U3 fell to 3.7% in November from 3.9% in October (expected 3.9%). Let's look at the details - Figure 1. The number of unemployed decreased by 215k (point 1 on the Figure 1). The number of employees increased by 747k (point 2). The sum of these two data gives +532k people and the Civilian Labor Force increased by this amount (point 3). The unemployment rate is Total Unemployed / Civilian Labor Force and amounted to 3.74% (point 4). This is a decrease of 14 bps (much less than (after rounding) 20 bps as officially reported by BLS), of which 13 bps is the contribution of the decline in the unemployed and only 1 bps is the contribution of the increase in the Civilian Labor Force. So the decline in the number of unemployed people was critical here (as a rule, it is). Interestingly, we are now only 35 bps above the April 2023 cycle low. In Canada, the unemployment rate has increased by 80 bps since April (Figure 2). Since 1960, the correlation between the U.S. and Canadian unemployment rates has been 0.65 (Figure 3).
US vs Canada employment
1-Dec-2023. US vs Canada Employment. Why wait another week for the U.S. November jobs report? Today we got the Canadian employment report for November. It’s well known that the economies of both countries are closely interconnected. Employment in Canada increased by 24.9 thousand people in November, above expectations of 15.8 thousand (vs. 17.5k in October and 63.8k in September; the 3-month average increase is 35.4 thousand). Figure 1. Figure 2 shows the annual change in employment for both countries since 2000. This increase in employment was enough for the annual employment change to remain at +2.52% (vs. 2.53% in October), while the 3-month annualized employment change dropped from 2.43% to 2.12% - see Figure 3. The unemployment rate increased from 5.7% to 5.8% in November (and is already up +80 bps from the recent low). For comparison, the U.S. unemployment rate is currently 50 bps above the low seen in April this year - Figure 4. Figure 5 shows the unemployment rates of both countries since 1960. We can probably expect a similar, i.e. neutral report in the USA (published on December 8).
job-worker gap
3-Nov-2023. Job-Worker Gap. The labor market is slowly but surely slowing down, also if we look at the job-worker gap - i.e. the difference between demand and supply in the US labor market (this is one of Powell's favorite charts). The Gap decreased in October, but only by 91 thousand people (in September it increased by 573 thousand people). However, the Gap is still above 3 million people, i.e. 3 million people more on demand side than supply! Definitions: Demand (number of available jobs) = Employment Level (number of employees according to Household Survey) + Number of Job Openings Supply (number of available employees) = number of people in the labor force - see chart 1. The job-worker gap decreased by 91,000 in October, of which: +56 thousand change of Job Openings -348 thousand change Employment Level +201 thousand Labor Force change – see chart 2.
US employment situation
3-Nov-2023. Employment Situation. X 2023. The 3-month average change in employed persons is negative and amounts to -13k persons (Household Survey). The change in October 2023 was -348k. Payroll employment (number of jobs, Establishmengt Survey) increased in October by only 150k (3-month average change is 204k). See chart 1. Chart 2 – the difference between Payroll Employment and Employed Persons is growing.
US wage growth
3-Nov-2023. Employment Situation. October 2023. Wage growth is also clearly slowing down... Powell's 3 favorite series (chart 1): 1) Average Hourly Earnings of All Employees: Total Private: only +4.10% YoY 2) Employment Cost Index: Wages and Salaries: Private Industry Workers: only 4.49% YoY 3) Atlanta FED Wage Tracker - Unweighted Overall: only 5.20% YoY On top, the 3-month average change in “Average Hourly Earnings of All Employees” is only +0.27%, which is exactly the pre-pandemic norm… see chart 2. Average Hourly Earnings of All Employees increased only +0.21% in October.
US unemployment rate
3-Nov-2023. Employment Situation. October 2023. Unemployment Rate 0.5 percentage point above the cycle low - this is a big deal! Unemployment Rate U3 in October increased to 3.90%. This is above the 12-month moving average which was at 3.61%. Historically, this has been a strong signal of an impending recession. See chart 1 and 2. Chart 3 shows the cycle of the unemployment rate and the S&P500 cycle simultaneously. Chart 4 shows various measures of the unemployment rate and their increase since recent lows. The Table 1 shows accurate calculations of the unemployment rate and more.
labor recession leading series
3-Nov-2023. Employment Situation. October 2023. Selection of leading recession data series from the labor market: 1) Unemployment rate, up 0.5 from the cycle low, strong signal, chart 1 2) Initial claims, up only 20k from the cycle low, no signal right now, chart 2 3) Continued claims, up 462k (+36%) from the cycle low, strong signal, chart 3 4) Average Weekly Overtime Hours Manufacturing, significantly down, strong signal, chart 4 5) Average Weekly Hours Manufacturing, significantly down, strong signal, chart 5 6) Index of Aggregate Weekly Hours Manufacturing, down 1,2% from cycle high, weak signal, chart 6 7) Temporary Help Services Employees, some signal, down from cycle high, but only to pre-covid level, chart 7 8) ISM Manufacturing Employment Index, at 46.8 it’s quite some signal, chart 8.
US unemployment rate august
5-Sep-2023. Why has the US unemployment rate (UR) increased so much? Rising UR is a potential "big deal" for the markets. Especially if it is confirmed in the coming months. Of course, it does not have to have an immediate impact on the markets, but it could be another “macro straw” to complete the current cycle. UR increased in August by 0.29 percentage points from 3.50% to 3.79%. UR is calculated very simply by dividing the number of unemployed by the size of the Labor Force. The number of unemployed increased by 514k in August (this is the largest monthly increase since April 2020 and as a rule such monthly increases do not happen outside of a recession - see charts). The Labor Force grew by 736k persons in August. It is therefore easy to calculate that the increase in the number of unemployed resulted in an increase in UR by 0.31 pp, and the increase of the Labor Force resulted in a decrease in UR by 0.02 pp (details in the table). In practice, a change in the number of unemployed will always be entirely responsible for a change in the UR (since there are 6m of Unemployed and the Labor Force is 168m). Unemployment Level consists of Job Losers, Job Leavers, Reentrants to Labor Force and New Entrants (see chart). In August, Job Losers increased by 294k, Job Leavers decreased by 51k, Reentrants +77k, New Entrants +94k. These numbers are not that essential, their distribution may be more important. Before the recession, one could expect a decrease in the share of Reentrants and Leavers, and an increase in the share of Losers. But about this and whether you can get any “leading” signals from it in the next post…
5-Sep-2023. Why has the US Unemployment Rate increased so much? Part 2. All categories of unemployed (Job Losers, Job Leavers, Reentrants and New Entrants) move more or less together during the cycle. But before and during recessions, the share of Losers (in total unemployed) increases and the share of Leavers and Reentrants decreases (see graphs). Looking at today's situation, one can say that such trends have already started (see next chart). So we are closer to another recession. Job leavers are unemployed people who quit or otherwise voluntarily left their previous job and immediately began looking for new employment. Reentrants are unemployed people who have past work experience but were not in the labor force for a period of time prior to beginning their current job search. New entrants are unemployed people looking for their first job. They have no previous work experience. Job losers consist of the following subgroups: - People on temporary layoff (These are people who have been given a date to return to work or who expect to return to work within 6 months. Unlike the other unemployed subgroups, those on temporary layoff do not need to be looking for work to be classified as unemployed). And People not on temporary layoff (data available since 1967), i.e.: - Permanent job losers (people whose employment ended involuntarily) - data available from 1994, and - People who completed a temporary job - data available since 1994. The last chart shows the breakdown of Job Losers. However, for this data series, the categories generally move similarly throughout the cycle, and it's hard to squeeze any leading signals out of it.
US labor market Mar-2024
7-Apr-2024. Strong at face value, but not so much under the hood. US March Payrolls. The market usually pays more attention to the month to month change in employment data, but several leading series indicate lower both wage and new jobs growth in H2 2024... My key takeaways from recent data: 1) Challenger layoffs are quite elevated recently… third month in a row around 80-90k – see Figure 1, 2) Despite high employment growth, the annual change in wages is the lowest in 3 years, and the Quits rate (from the Jolts report) indicates a further decline in wage inflation in the following months - see Figure 2, 3) NFIB Small Business Hiring Plans Continue to Decline, now at 11% net – the lowest level in 4 years – Figure 3 i Figure 3a, 4) The unemployment rate dropped in the US from 3.9% to 3.8% (exactly by only 3bps from 3.86% to 3.83%), and in Canada it's the other way around... - Figure 4. Since the bottom of the cycle, the unemployment rate in Canada has increased by 110 bps, and in the USA only 40 bps. Figure 5 shows a high correlation in the unemployment rates between the US and Canada.
US labor market Apr-2024
5-May-2024. US Labor Key Charts – update. My bunch of leading charts from a month ago on the US labor market hit the mark! Link to that post: https://www.jamkaglobal.com/post/strong-at-face-value-but-not-so-much-under-the-hood Challenger layoffs, Quits rate, NFIB Small Business Hiring Plans, Canada unemployment rate - all of them came really leading... And in the subsequent month we have a higher unemployment rate, lower wage inflation, and weaker job growth… It's time to update those charts - since they seem to work... The most important data, i.e. wage inflation... we have a drop to 3.92% YoY in the case of AHE (average hourly earnings, total private), in addition the Quits rate dropped even lower in the next month (March 2024, JOLTs report) - see Figure 1. Bonus chart: Quits rate vs Effective Fed Funds Rate… it looks very interesting - Figure 2. NFIB Small Business Hiring Plans has good leading properties... in April we have a drop in payrolls to 167k (total private), NFIB rebounds to 12% in April - see Figure 3. The U3 unemployment rate increased in the US to 3.9% in April - which is consistent with the strongly rising unemployment rate in Canada - Figure 4. Data for April in Canada will be available on May 10. However, other data do not fully confirm the rising unemployment rate: Challenger Cuts decreased in April (Figure 5), and Initial Claims do not increase at all, similarly to Continued Claims.. Figure 6. All-in-all, this is good news for risky assets - we are moving a bit towards the Goldilocks scenario... falling wage inflation and slightly slower labor market growth... Bonus Chart - Historically, rising unemployment rates have indicated an impending recession... see Figure 7.
