It’s hard to believe that sentiment can change so fast in the market that one day investors and traders are bidding up stocks to record highs, but then the next day sell them so much that it takes the market down over 2%.  That’s exactly what happened not only in the last two days but also two weeks ago.  While the 5% pullback from a record high back on June 10th took the Nasdaq back below its February high, this time around, the Nasdaq has been able to hold above those February highs.

In the entire history of the Nasdaq, there have only been 12 periods prior to this week where the Nasdaq closed at an all-time high on one day but dropped more than 2% the next day.  Those occurrences are highlighted in the table below along with the index’s performance over the following week, month, three months, six months, and one year.  We have also highlighted each occurrence that followed a prior one by less than three months in gray.  What immediately stands out in the table is how much gray shading there is.  In other words, these types of events tend to happen in bunches, and if you count the original occurrence in each of the bunches, the only two occurrences that didn’t come within three months of another occurrence (either before or after) were July 1986 and May 2017.

In terms of market performance following prior occurrences, the Nasdaq’s average and median returns were generally below average, but there is a pretty big caveat.  While the average one-year performance was a gain of 1.0% and a decline of 23.6% on a median basis, the six occurrences that came between December 1999 and March 2000 all essentially cover the same period (which was very bad) and skew the results.  Likewise, the three occurrences in the two-month stretch from late November 1998 through January 1999 where the Nasdaq saw strong gains also involves a degree of double-counting.  As a result of these performances at either end of the extreme, it’s hard to draw any trends from the prior occurrences except to say that they are typically followed by big moves in either direction.  The only time the Nasdaq wasn’t either 20% higher or lower one year later was in 1986. Like What you see? Click here to view Bespoke’s premium membership options for our best research available.

This week’s readings on sentiment through the American Association of Individual Investors‘ weekly survey was little changed following choppy price action over the past week.  Bullish sentiment fell just 0.23 percentage points to 24.14% marking a second straight week with less than a quarter of respondents reporting as bullish and the lowest reading since May 14th’s 23.31%. That small drop this week on top of the declines the prior two weeks also marks the first back to back to back declines in bullish sentiment since the three weeks ending January 9th. That’s right, not even during the bear market did bullish sentiment fall for three consecutive weeks. While there may not have been another period since January with consistent declines like the current stretch, there were other three week spans with larger declines. From April 16th to May 5th there was an 11.19 percentage point decline and from February 20th to March 12th there was a 10.86 percentage point decline compared to the current 10.41 percentage point decline from the June 4th high to today.

Likewise, neutral sentiment moved slightly lower falling from 27.85% to 26.96%.

Meanwhile, bearish sentiment remains the predominant position among survey respondents, rising for a second straight week to 48.9%.  Like bullish sentiment, this week did not see a particularly large move with the bearish camp only rising 1.12 percentage points.  The current reading is now the highest since May 14th when more than half of respondents reported as bearish.

That has sent the bull-bear spread further into negative territory.  Now at -24.76, the bull-bear spread is now at its widest level in favor of bears since May 14th.

While AAII’s survey holds a bearish bias at the moment, the Investors Intelligence survey is much more bullish. Bullish sentiment in this survey rose from 54.9% last week to 57.3% this week which is the highest level since January 22nd. That is also at the upper end of the past decade’s range in the 86th percentile. Bearish sentiment on the other hand fell for a 13th straight week which is its longest stretch of consecutive declines since at least 1997. Now at 18.4%, bearish sentiment is at its lowest level since January 22nd. Click here to view Bespoke’s premium membership options for our best research available.

For a record 12th straight week, jobless claims declined this week.  While they’re moving in the right direction, over those same 12 weeks, claims have actually exceeded consensus forecasts nine times. Seasonally adjusted claims totaled 1.48 million which was down 60K from a revised 1.54 million last week (the original release was 1.508 million). While that 60K decline was larger than the 26K decline the previous week, the rate of improvement in jobless claims remains toned down from what was observed over the past few months.

As previously mentioned, jobless claims have been consistently coming in above estimates.  In the three months that jobless claims have declined, there have only been three weeks (April 10th, April 17th, and June 5th) that claims have come in better than expectations. In the chart below, we show the count of weeks in rolling a 12-week span that claims were above forecasts. The current reading of 9 weeks is actually just off the recent peak of 10 weeks thanks to the recent beat on June 5th. Prior to that, at the end of May, 10 of the 12 weeks had seen claims miss estimates.  Before that, you would have to go all the way back to 2011 a stretch of weaker than expected reports that were this weak.  The weakest 3-month span relative to estimates was back in the 12 weeks ending November 14th, 2008 when claims had missed estimates for 11 of 12 weeks.

To put it briefly, while any drop in jobless claims is welcomed and that has been observed over the past three months, claims have seemed to have hit a bit of a plateau in terms of improvement. Non-seasonally adjusted claims embody that dynamic of shrinking improvements.  Non-seasonally adjusted claims only fell by 6K to 1.457 million this week.  That 6K decline was actually the smallest weekly move in absolute terms since a 5.1K decline back in early February- back before the surges of the COVID era.

Unlike initial claims, continuing claims actually beat estimates this week falling below 20 million for the first time since mid-April.  Continuing claims have now fallen for four of the past five weeks since the peak of 24.912 million claims on May 8th. Also unlike initial claims, this was actually a larger improvement than what has been observed in recent weeks. The first reading after the aforementioned peak in claims was the largest one week drop on record totaling 4.071 million claims. This week’s 767K decline to 19.522 million continuing claims was the second-largest on record. It was also more than double the prior week’s decline.  Granted, even with those improvements more than a tenth of the US workforce is currently unemployed. Click here to view Bespoke’s premium membership options for our best research available.

After eight straight days of gains, the Nasdaq looks like it’s ready for a breather today as the index is down over 1%.  The current streak represented the longest string of back to back gains for the Nasdaq of 2020 and the longest streak since the 11-day streak that came to an end last December.  Throughout the Nasdaq’s history going back to 1971, there have been over 80 streaks of at least eight trading days, but since the tech boom really took off in the late 1990s, these streaks haven’t been as frequent with just 26 now since 1995.

Once a winning streak of eight or more days comes to an end, one would think that short-term returns going forward would be below average.  However, looking back at how the Nasdaq performed following prior eight-day winning streaks, more often than not, the index has actually tended to have better than average short-term forward performance.  In the week that followed these prior streaks, the Nasdaq averaged a gain of 0.55 (median: 0.35%) with positive returns 68% of the time.  One month later, the average gain increased to 1.79% (median: 2.15%) with positive returns 84% of the time.  Three months later, the effect of an eight-day winning streak probably didn’t mean much, but even here the Nasdaq averaged a gain of 4.14% (median: 3.96%) with gains 72% of the time.  What’s notable about all of these performance numbers is that they are all better and more consistent to the upside than the Nasdaq’s historical average one week, one month, and three-month returns.  Today’s performance is ending the Nasdaq’s winning streak with a thud, but history shows that prior streaks have typically been part of longer periods of outperformance.

One item worth noting about today’s decline for the Nasdaq, though, is that with a decline of over 2%, the only other time where an eight-day winning streak ended with a decline of more than 2% was back in July 1986. Like What you see? Click here to view all of Bespoke’s membership options for access to the best research available.