Following up on a disappointing ADP employment number yesterday, today’s release of weekly jobless claims also indicated a modest deterioration in labor market data. Seasonally adjusted initial claims have risen in back-to-back weeks up to 207K. On the bright side, that was below expectations and is only a minor increase as claims remain below their range from throughout the spring and summer this year. Additionally, in the low 200K range, claims are still at a historically healthy level.

Before seasonal adjustment, claims were actually lower at 172.78K.  In one sense, that lower reading is not exactly surprising as week-over-week declines have been observed roughly 70% of the time historically in the current week of the year.  However, what is now more unusual is that it sets a new low on the year. As we have frequently noted in recent weeks, this time of year typically sees claims put in an annual low, but the new low this week is a bit later than normal. In fact, outside of the pandemic years (2020 and 2021) when claims were historically volatile, the last time an annual low occurred this late in the year was 2014.  Prior to that, 1967, 1980, 2000, and 2011 were the only other years with an annual low in the 39th week or later. In other words, claims have remained strong, and seasonal headwinds haven’t yet seemed to come into play in any impactful way.

Like initial claims, seasonally adjusted continuing claims came in stronger than expected at 1.664 million. That is a tiny drop from 1.665 million the previous week but is still off of the low of 1.658 million from two weeks prior.  In all, that leaves claims at historically strong levels with a modest multi-month downtrend still in place.

The S&P 500 has continued to hit new lows in the past week, and some sentiment readings have reflected that negative tone.  The AAII sentiment survey was not necessarily one of those as this week saw a mixed result.  For starters, bullish sentiment actually ticked up to 30.1% from 27.8%.  That ends a streak of three straight weeks of declines as bullish sentiment was above 40% as recently as the first week of September.

While bullish sentiment went the other way of price action, bears did increase slightly from 40.9% to 41.6%.  That brings the total increase in bears over the past three weeks to 12.4%, the largest three-week increase since late August.

That means that on net, AAII sentiment actually shifted slightly more bullish this week. The bull-bear spread continues to be negative (meaning more investors are reporting bearish than bullish sentiment), but that reading was higher at -11.5 this week.

Other sentiment surveys were not as hopeful.  Both the NAAIM Exposure index and Investors Intelligence survey saw readings shift more bearish in the latest week’s data.  In fact, both of those surveys’ readings have been more bearish week over week in each of the past three weeks.  All plugged into our sentiment composite, that has outweighed the modestly more bullish reading in the AAII survey. As a result, the composite indicates sentiment levels are 0.69 standard deviations more bearish than what has historically been the norm.  While not exactly an extreme, especially in the wake of the past couple of years, that is the most bearish read on sentiment in just over six months.

While maybe not as relentless as the move higher in rates over the last two months, selling in equities has been pretty consistent.  In the year-to-date chart of the S&P 500 tracking ETF (SPY) below, we show the last 50 trading days in gray to point out that there have been an extremely large number of days during this span where the day’s intraday low was lower than the prior day’s low.  We highlighted this trend in a Chart of the Day last week, but it has remained pronounced since then. In total, 33 of the last 50 trading days have seen SPY make a lower low relative to the prior day’s intraday low, and if SPY falls below $420.18 today, it would be a record 34 days in a trailing 50-trading day period where the ETF made a lower low relative to the prior day’s low.

The chart below shows the number of days over a rolling 50-day period where SPY made lower lows, and at a level of 33, the current period is tied with three others (March 2022, October 2008, and March 2008) for the most since SPY’s inception in 1993.  You don’t need us to tell you that none of these periods were positive for the market.  What makes the current period unique is that the magnitude of the decline during this period has been relatively mild at less than 10%.  During each of the three other periods, SPY was down at least 10% from a 52-week high and as much as 45% (October 2008).

This morning we received two of the latest updates on labor market demand with the release of the August JOLTS report in addition to postings data from Indeed through the end of September.  The JOLTS report came in well above expectations (9.61 million versus 8.83 expected) indicating a solid rebound in labor market demand headed out of the summer.  In spite of that positive reading, the overall trend of lower openings remains in place and is echoed by Indeed’s data.  As shown below, the more timely and higher frequency postings data has also been trending lower since the end of 2021. That being said, the summer has seen those declines decelerating with postings only slightly lower over the past three months. Modeling the JOLTS number on the less lagged Indeed data would predict that postings would remain around these levels next month. In tonight’s Closer, we will provide a full rundown of the latest JOLTS report.

In addition to national reads on job postings, the Indeed data also provides geographic breakdowns by US metro, and in the table below, we highlight the 25 MSAs (metropolitan statistical area) that have seen the best and worst postings growth relative to pre-pandemic baselines as well as how far they have fallen from their respective peaks (we highlight when each of those peaks were as well). Many of those with the highest number of openings relative to pre-pandemic are also those with smaller populations.  Conversely, many of the largest metros have seen job postings fall off the most. There have also been a growing number of cities where postings are now below pre-pandemic levels.  San Francisco is the worst of these with postings down nearly 20% from baseline.

After a relatively dismal two months for stocks, the market kicked off what has historically been its strongest period of the year today.  In the post-WWII period, the S&P 500’s average performance in Q4 has been a gain of 4.1%, which is more than double the 2.0% average gains of Q1 and Q2 and ten times the average gain of Q3 (0.4%).

While Q4 has been positive for equities, the month of October has historically been volatile. Since 1945, the spread between the month’s daily closing high and closing low has been 7.1%. While the average spread for every month except October fits within a 1.3 percentage point range of 4.7% to 6.0%, October is all alone at more than a full percentage point from the high end of that range.

With Q4 being the strongest quarter of the year and October being the most volatile month, they don’t call October the month of market bottoms for nothing.  In looking back at every market decline of at least 5% (without a rally of 5%+ in between), market lows have easily been the most prevalent in October.  As shown in the chart below, 33 (14.4%) of the ‘market lows’ since 1945 have occurred in October, and the only two other months that account for even 10% of all market lows were March and September.  Seasonality is on the side of bulls heading into Q4, but that doesn’t mean the road is smoothly paved.

Please click the image below to view our October 2023 market calendar.  This calendar includes the S&P 500’s historical average percentage change and average intraday chart pattern for each trading day during the upcoming month.  It also includes market holidays and options expiration dates plus the dates of key economic indicator releases.  Click here to view Bespoke’s premium membership options.