In spite of the S&P 500’s consistent declines in the past week as it failed to take out its late June highs, investor sentiment has turned around (relatively speaking) with this week’s reading from the AAII showing 26.9% of respondents reporting as bullish for the first time since early June.  The 7.5 percentage point increase in the percentage of bullish responses this week was a large week-over-week increase by historical standards, although there have been multiple even bigger larger weekly increases over the past few months.

With the increase in bullish sentiment, over a quarter of respondents reported as bullish for the first time in five weeks.  Such extended streaks with as depressed readings have been few and far between with the last five-week streak occurring all the way back in the summer of 1993.  Overall, there have now only been six streaks in which bullish sentiment remained below 25% for at least 5 consecutive weeks.  The longest of these was in December 1990 when it went on for 9 weeks in a row.  Albeit a small sample size, historically the end of these streaks have not been raging buy signals for the S&P 500 in the short term with inline performance versus all periods and somewhat weak returns one month out.  However, three, six, and twelve months later the S&P 500 has been higher almost every time with slightly stronger than normal performance (six months out from the March 1990 occurrence was the only decline).

The increase in bullish sentiment was met by bears falling back below 50% to 46.5%.  Mirroring bullish sentiment, that made for the lowest reading since the first week of June.

As a result of those moves, the bull-bear spread remains firmly in favor of bears. With the percentage of bearish responses outnumbering bulls by 19.6 points, for the 15th week in a row the bull-bear spread remains negative. That steak has grown to be the third largest on record behind a 22-week streak ending in late 1990 and a 34-week streak ending in October 2020.

Neutral sentiment has managed to avoid major shifts in sentiment in recent weeks and this week was no exception.  This reading fell modestly from 27.8% to 26.6%.  That is well within the range of the past couple of years’ readings and is only the lowest since three weeks ago. Click here to learn more about Bespoke’s premium stock market research service.

Jobless claims have continued to rise hitting the highest level since November this week. Claims rose from an unrevised 235K up to 244K this week.  With consistent increases in claims over the past few months, the reading has gone from multi-decade lows to levels that would have been the highest since late 2017/early 2018 pre-pandemic.

As we have noted each week recently, the current week of the year has historically marked a seasonal peak in claims.  As such, the current week of the year has also seen a sequential rise in claims 80% of the time including a rise this week.  NSA claims now are at 241.3K which is inline with the comparable readings for the three years prior to the pandemic. Although part of the recent increase in claims is likely seasonal in nature, given evidence of a slowing economy including weakness in other labor market data like under the hood of Friday’s nonfarm payrolls report, how much that seasonal uptick unwinds in the weeks ahead will be something to watch.

As for continuing claims, there has not been as notable of an increase recently when compared to initial claims. Continuing claims have only risen marginally from the low of 1.306 million in the week of May 20th to 1.331 million in the most recent reading as of July 1st.  That most recent reading also saw a sizable improvement from the prior week with claims falling from 1.372 million.  Click here to learn more about Bespoke’s premium stock market research service.

Over the past few days, equities, including the most heavily shorted stocks, have reversed a good amount of the gains seen since the mid-June lows.  A basket of the 100 most heavily shorted stocks in the Russell 3,000 is currently little changed versus its levels at the start of 2020 compared to a roughly 16% gain for the Russell 3000. Whereas heavily shorted stocks saw massive outperformance versus the broader market in the second half of 2020 and early 2021, that outperformance has unwound since last fall.  More recently over the past two months, though, highly shorted stocks and the Russell 3000 have been performing more or less in line with each other as the relative strength line has trended sideways and is near similar levels to the COVID Crash.

The most recent short interest data as of the end of June was also updated in the past 24 hours.  For the whole of the Russell 3000, the average reading on short interest as a percentage of float currently stands at 6.24%. That is up 20 basis points since the prior update for mid-June and a full percentage point higher versus the start of the year.

Across nearly all industry groups, there have been major shifts in short interest levels since the start of the year.  While most have made considerable moves higher, there are others that have actually fallen significantly like Food & Staples Retail and Telecommunication Services.  Given the big increases this year, Retail continues to have the highest average reading on short interest followed by the Pharmaceuticals, Biotech &Life Sciences, and the automobile industries. Although their levels of short interest are not as elevated, Software & Services and Energy stocks saw the largest increases from the last report while only the Media & Entertainment industry and Food & Staples Retailing stocks saw declines in average short interest between the two most recent reports. That being said, none of those sequential moves were particularly large with the biggest absolute move being the 0.45 percentage point increase in Software & Services.

Switching over to the individual stocks with the highest levels of short interest, Dillard’s (DDS) currently tops the list with a little more than half of its shares sold short.  That reading is only half of a percentage point higher versus mid-June but has nearly doubled since the start of the year. Although it continues to sit near the top of the list, Redbox Entertainment (RDBX), has perhaps seen the most notable shift in levels of short interest versus the last bi-weekly report.  Throughout the year, the recent SPAC has seen large swings in its reported short interest levels. Click here to learn more about Bespoke’s premium stock market research service.

For most of the session yesterday, the S&P 500 was having trouble choosing a direction, but come the final hours of trading with a CPI release looming, the index took a decisive turn lower. Given rampant hot inflation over the past couple of years and its implications for changes to monetary policy, that sort of late day selling ahead of CPI prints has been the norm as shown below.  On average for the past two years, the day before CPI releases has typically seen the S&P 500 trade higher for most of the session before turning lower and erasing its gains in the afternoon; just as we saw yesterday.

With another hotter than expected print this morning for both headline and core measures of inflation, yesterday’s late day sellers seem to have been on the right side of the trade as the S&P 500 has fallen further and is on pace for the sixth drop on a CPI print in a row. Taking a look at intraday price action of the S&P 500 on CPI release days, historically there has been a downward bias with all releases since 2000 averaging a drop of a few basis points by the close.  Over the past two years it has been far worse with an average decline of 28 bps.  Most of that drop has actually occurred after a bout of midday selling.

Using data from our Economic Indicator Database, filtering out for only times in which CPI has come in above expectations (on a month over month basis), that same pattern is apparent.  The S&P 500 tends to trade in the red for most of the morning but the worst declines occur in the early afternoon before some stabilizing into the close.

In response to today’s hotter than anticipated release, the S&P 500 gapped down 1.2% which ranks as the fourth worst gap down for the index on a CPI day since at least 2000.  Going back over the past two decades, there have only been a total of seven gaps down of 1% or more on a CPI day.  Including today, four of those releases have occurred this year: February (-1.2%), March (-1%), June (-1.64%), and today (-1.23%).  Taking a look at the intraday pattern of the S&P 500 across those 1% or larger gaps, the index has tended to continue to fall throughout the session with the lows of the day tending to occur around 2:30 EST with modest sideways action into the close from there.  Click here to learn more about Bespoke’s premium stock market research service.

With inflation running out of control and markets furiously attempting to re-price Federal Reserve interest rate policy, we’re seeing some truly wild moves in the fixed income markets.  This week, the big moves have come at the short-end of the Treasury yield curve as three-month Treasury yields have surged by 35 basis points this week alone.  While the 2y10y US Treasury yield curve has been inverted for seven trading days now, the 3m10y curve, which is the FOMC’s preferred measure of the yield curve, has remained positively sloped, but the rate of flattening has really picked up in recent days and since early May has gone from around 225 bps to under 75 bps today. Click here to learn more about Bespoke’s premium stock market research service.

As mentioned above, the bulk of the flattening in the 3m10y curve has occurred more recently.  The chart below shows the 20-day rate of change in the 3m10y US Treasury yield curve since 1962.  The red line indicates the threshold for 100 bps of flattening in a four-week span which is where we are at now.  Prior to the current period, there wasn’t a similar flattening of the 3m10y curve in a four-week span since the Financial Crisis. In August 2011, the rate of flattening got close to 100 bps in a four-week span but came up just shy. In other words, these types of moves in the yield curve don’t happen very often.

How often are they? The chart below shows the distribution of four-week changes in the 3m10y yield curve over time grouped into 20 bps increments.  Including the last few days, where we are at in the current period  (-100 to -120 bps of flattening in a four-week span) has only occurred on 77 trading days, and there have only been a total of 223 trading days where the yield curve flattened by 100 bps or more in a four-week period.  On a percentage basis, that works out to less than 1.5% of all trading days. When it comes to markets, it seems as though there’s always something unique happening out there, and these days, you don’t see what’s happening at the short end of the Treasury curve very often.

The only US release on the economic calendar today was the NFIB’s Small Business Optimism Index, and it was a big dud.  Compared to last month’s low reading of 93.1, the index fell to 89.5 versus expectations for a much more modest decline to 92.8.  As a result of that larger than anticipated drop, the index is now below the COVID low and down to the lowest level since January 2013.

As the headline number is now in the 6th percentile of its historical range, several other components are similarly depressed including a record low reading in expectations for the economy to improve.  Although some indices have managed to hold up decently—for example, reported inventory levels and the difficulty in filling openings remain in the top few percentiles since the survey began—breadth in June was very weak with many of the declines being historically large.

As we noted in today’s Morning Lineup, labor market indicators within the report have cooled off in recent months but remain at historically impressive levels.  Meanwhile, inflation readings are also continuing to come in very elevated in spite of coincident declines in sales expectations.

Again, the worst category of the report was the reported outlook for general business conditions.  This index set the bar for a record low even lower as a net 61% of respondents reported that they expect the economy to worsen in the months ahead.  Given this, there is also a historically muted share of respondents reporting now as a good time to expand their businesses.  That reading came in at a meager 3, which alongside April 2020 and March 2010, was the joint lowest level since March 2009.

The NFIB also provides some detail as to why businesses are reporting such a soured outlook regarding expansion.  Half of the responses blamed broader economic weakness as the reason. Ironically, that goes against the fact that only 2% reported weak sales would be a reason while another 3% reported strong sales would be the reason to expand. Additionally, while inflation remains such a major concern, only 4% blamed higher costs as a reason to not expand. So while half blamed economic conditions as the reason not to expand, there were some contradictory cross currents.

While it is a drastically lower share at only 10%, the next biggest reason for not expanding was the political climate. As we have frequently noted in the past, one issue with the NFIB survey is that there has been a sensitivity to politics, namely a bias in favor of Republicans. Given the Democratic administration, it is not exactly surprising that political climate would come in second.

Another interesting point to note on this month’s survey regarded the uncertainty index. This measure essentially tracks the number of times respondents select “uncertain” or “don’t know” as an answer to a survey question.  Here was another index that has reached the lowest level in over a decade. This month’s low reading would imply that firms have a decidedly negative outlook for the economy.

While sales expectations are down near the worst levels since the Global Financial Crisis and the COVID Crash, actual observed changes in sales have not yet cratered in the same way.  That being said, on net there are more firms reporting lower sales than higher.

In other words, small businesses expect sales to get much worse than what current levels of sales would have historically implied.  As shown below, the spread between those two indices has only been lower once on record and that was at the start of the pandemic.  Similarly, the record low in General Business Outlook starkly contrasts with where the headline optimism index has been sitting.  That spread is at a record high meaning again economic outlook is far worse than the overall level of optimism.