Macro Monday 20 ~ The Philly Fed IndexMacro Monday 20
Philadelphia Fed Manufacturing Index
While the Philly Fed Manufacturing Index (PFMI) is a regional report generated from surveys in Philadelphia, New Jersey, and Delaware by the Federal Reserve Bank, it is particularly useful as it provides an advance indication of the Purchasing Managers’ Index (PMI) report which is released up to a week after the PFMI (the PMI surveys the entire US whilst the PFMI only surveys the regions mentioned above).
The Philly Fed Index is released this Thursday 16th November 2023 and will provide an advance indication as to what to expect from the PMI released Friday 24th November 2023. Both are a review the prior months survey data, October 2023.
The PFMI index dates back to 1968 and is similar to the PMI, the Federal Reserve completes surveys and asks businesses about new orders, shipments, employment, inventories and general business activity, prices paid, prices received, capital expenditures as well as future expectations for business.
A reading= 0 is stagnation
<0 = contraction
>0 = expansion
The current reading is -9 so we are contractionary territory. We did fall as low as -31.3 on the April 2023 release.
The Chart
The main indications from the chart are as follows:
The Orange Zone
▫️ When the PFMI remains in the orange zone for >10 months it has always coincided with a Recession
- We are in presently in this zone 16 months with 2 brief monthly jumps out of it. I think its safe to say we are 10 months+ in the orange zone which historically has always coincided with a recession.
The Red Zone
▫️All Recessions confirmed a reading below -22 on the PFMI (this is below the red line into the red zone on the chart)
- In April 2023 we hit a low of -31.3 which is well into the red zone (sub -22). We have since risen above the neutral 0 level to high of +12 in Aug 2023 however we have since fallen back down into the -13.5 (Sept) and -9 (Oct). The Nov Release is due this Thursday 16th Nov (and is actually the reading for Oct - released in Nov)
Are we already in a mild Recession?
You can see that in March 1970 we reached a similar PFMI level of -31.3, the same level as in April 2023 (there is a dashed red line to illustrate this on the chart). March 1970 was the middle of the 1969-70 Recession which was a mild recession that ran for 11 months from Dec 1969 – Nov 1970. Whilst it was a mild recession as to its impact on the general economy, there was till a 34% decline in the S&P500.
The 1969-70 Recession has many similarities to some of our current economic predicaments, with the main factors leading to the 1970 recession being tighter monetary policy, rising oil prices, rising inflation, and slowing growth in Europe and Asia. Sound familiar?
From Jan – Apr 2023 the Unemployment Rate was at the lowest levels seen since back in 1969 (at 3.4%). For 8 months (Sept 1968 – May 1969) the unemployment rate was down at 3.4%. We reached this level in January 2023 and oscillated there until April 2023 (only 4 months). Since then the Unemployment rate has risen sharply from 3.5% to 3.9% (July – Oct 2023). Interestingly, this move in the unemployment rate from 3.5% to 3.9% also happened from Dec 1969 to Jan 1970 and marked the start of the recession. Could this be an indicator that we stepped into a recession In July 2023? The orange zone and red zone on the chart are triggering a confirmation nod of a recession. During the recession of 1969-70 the unemployment rate topped at 6.08% in Nov 1970, this is something we have not seen yet however we seem to be trending upwards in that direction. Queue the 8th Dec 2023, the next Bureau of Labor Statistics Unemployment date release.
The 1968-70 period was also burdened with high inflation with YOY CPI increasing from 2% - 6% in the 26 month period from Oct 1967 – Dec 1969. Similarly over a 25 month period from May 2020 – June 2022 CPI increased from 0% to 9.08%. The timeline of the 1969-70 inflation is quite similar, not exactly the same rate increase or timeline but similar all the same. Since June 2022 the CPI has come down to 3.7% as of Sept 2023.
There are some broader similarities between the late 1960’s and early 1970’s to present day, the Vietnam war was raging and was receiving significant funding from the US government with many bills passed in support of the war effort. There was also significant poverty issues in the states as the war dragged on, and the awareness of money being spent on it was creating social discourse on the topic. Whilst the current situation of funding towards the Ukraine and Palestine conflicts is obviously very different, a similar awareness and disapproval is present as many domestic states are suffering with poverty. US President Johnson summarised the late 60’s quiet well in a 1966 speech stating that the nation could afford to spend heavily on both national security and social welfare — “both guns and butter”, as the old saying goes. Only in today’s circumstances only one of these seem to still be taking priority and it isn’t butter.
I believe todays chart and post demonstrates a few things, that there is a high probability that we are already in a recession as of July 2023, however on a positive note the period we find ourselves in has many similarities to 1969-70 period, where the recession was a very bearable and mild one. With some luck, unemployment might top at 6.08% within 9 or 10 months like in 1970 and we will see a correction no greater than -34% on the S&P500 eventually. We already survived a 25% S&P500 decline from Dec 21 – Sept 2022. Minus 34% from our recent $4,580 high would put the S&P500 at approx. $3,000.
Obviously there are no guarantees of any of these scenarios playing out, but at present we are certainly playing to the same tune as the 1969-70 period.
PUKA
Recessionincoming
Macro Monday 19~Nonfarm Payrolls Macro Monday 19
Total Non-Farm Payrolls: Pre-Recession Observations
What is Non-Farm Payroll?
The nonfarm payroll measures the number of workers in the U.S. includes 80% of US workers. The figures exclude farm workers (Nonfarm) and workers in several other job classifications such as military and non-profit employees.
Data on nonfarm payrolls is collected by the Bureau of Labor Statistics (BLS) and it is included in the monthly Employment Situation report (the “Employment Report”) which includes two surveys, the Household Survey, and the Establishment Survey. Nonfarm Payroll is included in the latter the Establishment Survey.
The Establishment Survey gathers data from approximately 122,000 nonfarm businesses and government agencies for some 666,000 work sites and about one-third of all payroll workers. Anyone on the payroll of a surveyed business during that reference week, including part-time workers and those on paid leave, is included in the count used to produce an estimate of total U.S. nonfarm payrolls
The Full Employment Report is released by the BLS on the first Friday of each month at 8:30 AM ET and reflects the previous month's data.
The Chart
▫️ The Chart highlights the last four recessions (red shaded areas)
▫️ The aim of the chart is to identify what Non-Farm Payroll movement occurred prior to each recession (in the blue shaded areas) so that we create a gauge that identifies the early warning signals of such recessions.
▫️ From reviewing the data (illustrated in the data chart), prior to each recession there was a either a confirmed decline in Non-Farm Payrolls prior to recession or an increase of less than 0.0300 mln in Non-Farm Payrolls prior to recession (a tapering off or sideways move). This was evident prior to all four recessions reviewed.
Main Findings:
1. The four most recent recessions all seen a decline in Non-Farm Payrolls prior to recession or an increase of less than 0.030 mln in Non-Farm Payrolls prior to recession (the “Signal”). Advance notice of recession was 1 to 12 months depending on recession (final column)
2. Currently we do not have a decline or an increase of less than 0.030 mln in Non-Farm Payrolls thus suggesting we do not have an advance recession warning triggering at present.
3. From a review of the data chart we are now aware that a pre-recession signal can trigger and provide us with 1 months advance notice or 12 months advance notice. In the event the parameters of number 1 above are met to provide a Signal, we can then add this chart/metric as a recession warning chart.
Breakdown of Each Recession Signal
(signal defined in 1 above):
▫️ The 1990 recession gave us a 1 month advance warning of recession.
▫️ The 2000 recession provided 2 advance warnings (2 & 3 in the chart), one signal gave us a 9 month heads up and the other a 3 month advance notice.
▫️ Similarly, the GFC 2007 recession provided 2 advance warnings (4 & 5 in the chart), one gave us 5 month advance warning, and the second was the signal the recession had started.
▫️ COVID-19 provided a 12 month advance warning with a decline registered from Jan – Feb in 2019.
Side Note: Interestingly this has some alignment with last week’s chart on Durable goods. In Feb 2019 one year before the COVID-19 Crash the Durable Goods Moving Average provided an advanced sell/recession signal, and whilst the S&P500 did rally c.13.5% after the signal over the subsequent 12 months, the S&P500 ultimately fell 23% thereafter in a matter of months taking back all those gains and more. Durable Goods is also included in the Establishment Survey so maybe it should come as no surprise that we have synchronicity between both charts on the COVID Crash. The Durable goods chart is also not presently signaling a recession similar to this Nonfarm payroll chart. Both charts appear to demonstrate some resiliency in the employment market (echoing Jerome Powell's sentiment that Employment is tight).
False Signals
▫️ Unfortunately there are a number of false signals throughout the chart whereby a decline in payrolls or an increase of less than 0.0300 mln is observed with no follow up recession however most of these false signals are either 1 month in duration or happened in the direct follow up years after the recession slump (when a recession is no longer of concern). Regardless, for this reason the Non-Farm Payrolls Recession Signal cannot be utilized as a standalone indicator, we need other charts and data to help identify the risk of recession.
▫️ Other data should be utilized in conjunction with Non-Farm Payrolls such as the following closely aligned charts all of which are show concerning pre-recession patterns in one way or another;
1. Total Non-Farm Layoffs and Discharges
2. Total Nonfarm Job Openings
3. US Continuing Jobless Claims
1. Total Non-Farm Layoffs and Discharges is signaling a similar trend to the 2007 Great Financial Crisis were there was an initial increase of c.450k (up to the first peak) and eventually a total increase of c.885k from lows to peak recession high.
- At present we are trending upwards and had an initial peak of c.507k (it could be the only peak or the initial peak, time will tell).
2. Total Nonfarm Job Openings is signaling a significant decline in job openings much larger than the prior two instances where job opening declines led to recession.
- A quick glance at the chart and you can see that we have exceeded the typically level required for recession and exceeded the typical timeframe (using GFC and COVID as reference points).
3. US Continuing Jobless Claims -Prior to the last 8 recessions the average increase in cont. claims was a 424k increase over an average timeframe of 11 months.
- Since Sept 2022 Cont. Claims have increased from c.1.3m to 1.818m (an increase of c.518k over a 13.5 month period). We are above both pre-recession averages number of increase and time.
In summary:
▫️ Last week’s Durable Goods Chart and this week’s Nonfarm payrolls chart are not triggering a recession warning at present. Both charts appear to emphasize a resilient labor market.
▫️ In stark contrast all three of the additional charts I provided above are incredibly concerning on the recession probability front. In particular Cont. claims , the most concerning of the bunch, is surpassing all pre-recession averages, highlighting that people are finding it harder to recover from a job loss and find a new job. This chart alone would suggest that the labor market is beginning to significantly soften.
▫️ Over the past week we have also had an update to the Purchaser Managers Index which declined further into contractionary territory from 49.0 to 46.7 (est. 49.0). Another signal towards a softening labor market.
▫️ It would be remiss of me not mention that I have seen a Month Over Month (MoM) Chart of the Nonfarm payrolls doing the rounds and it appears to illustrate a softening and slowing of labor conditions (will share in the comments). Such a trend could translate to a gradual tapering and/or decline on our monthly Nonfarm chart over time.
When you consider all of the above, you would have to expect a market decline is around the corner but also expect some continued lag before we see it due to those few charts that are not even showing the pre-recession signals, never mind an actual recession signal. The charts holding out are Durable Goods, Nonfarm Payrolls and ill throw in Major Market Index TVC:XMI as a complimentary chart that has not lost its support as of yet. We are also aware that the Dow Theory has confirmed a bear market and has been expecting a market rally before bear trend continuation (the sell into rally). All the same these moving parts can change and pivot so we have to keep an open mind but its hard not to lean very cautiously as it stands. We can keep an eye on these final charts that remain defiant as they may be the final strongholds and provide us with the final warnings in the event of....
As always folks stay nimble out there
PUKA