Bonds have had their worst year ever by many measures. And September has the potential to be the worst month ever for several bond ETFs and indices. Below I am going to show the worst months of all time for a few bond ETFs.
First, IEF, (iShares 7-10 year Treasury Bond ETF).
Four of the worst ten months ever for IEF occur in 2022 (so far).
Next, AGG (iShares Core US Aggregate Bond ETF)
Five of the worst ten months have come in 2022, including the 4 worst ever.
Lastly, let’s check out the US Treasury 1-3 yr Total Return Index going back to 1990.
Six of the worst 12 months ever have occurred this year. No other year has more than 1 month on the list.
If you feel like your bond investments have been beating you up month after month this year…you’re right.
Wednesday is a Fed Day. And a decent amount of the recent selling may be attributable to anxiety over the Fed announcement. I’ve shown Fed Days to be bullish in the past, especially when there has been a selloff heading into the Fed Day. The study below is from this past weekend’s subscriber letter. It looks at other times SPX closed at a 5-day low and below the 200ma on the day before a Fed Day.
The numbers are strongly positive, and as you can see, the setup really picked up steam starting in 2008. Below are the stats and list of instances from 2008 onward.
SPX would need to close below 3873 today in order to trigger this setup. For many more studies regarding Fed Days, check out the Fed Study category here on the blog.
While SPX and NDX posted small gains last Thursday, the Russell 2000 is now down 7 days in a row. I decided to examine past instances where RUT closed down 7 days in a row and under the 200ma.
If you were hoping that 7 consecutive down days put the market so oversold that it “had” to bounce, well, these results would be disappointing. Winners to losers has been basically 50/50 but the losers have been substantially larger. Below is a look at all the instances and their 4-day results.
Quite a few sizable declines in the mix here. The average drawdown over the 4-day period for all 18 instances was 4.6%, and the average run-up was 2.7%. In other words, volatility was high. Seven down days for RUT does not seem to assure us of anything, other than perhaps some additional volatility.
A couple of weeks ago on Twitter I shared a breadth thrust study that looked at times the % of SPX stock above their 50ma quickly rose from <= 15% to >= 90% within 50 days. The results I shared can be found below.
I noted the following in the letter after examining the results…but this appears to be yet another breadth study that we can add to the growing list suggesting an upside edge over the intermediate-term. Of the 17 instances that triggered since 1957, all of them saw the SPX higher 3, 6, and 12 months later.
I have received some questions and seen some discussion on Twitter about whether the recent tumble the market has taken meant the end of the breadth thrust signal. So I looked at it a couple of ways. I noted that Friday’s selloff caused the SPX to close at a 20-day low. I then looked at other times the market quickly reversed and posted a 20-day closing low after seeing a breadth thrust like those above.
There were only 3 instances that saw such a quick dive to a 20-day low after posting a similar breadth thrust. I never like to read anything into just 3 instances. So I expanded the time period to note any instance that closed at a 20-day low within 50 days. Those results are below.
Still just 12 instances, but also still no indication that the pullback would suggest the original study has been invalidated by the recent drop. While there are bearish studies and indicators out there, I see no good reason to throw out the original breadth study. Other instances where there have been pullbacks to 20-day lows after similar breadth thrusts have all been followed by gains 2, 3, 6, and 12 months later.
I will note that I did see a “breadth collapse” study that triggered last week suggesting more downside over the next month. I wrote about that in the subscriber letter. (You may take a free trial here if you wish to see it.)
So while the 50ma breadth thrust study still appears to be a positive looking out over the longer time-frames, there is contrasting evidence for the intermediate-term. Overall, that leaves breadth readings somewhat questionable. Perhaps we will see more clarity in the weeks to come.
The market had a pretty awful 1st half of the year. The SPX lost over 20% through June, which is something it had not done since 1970 during the 1st 6th months of a year. Out of curiosity, I decided to see how other strongly negative 1st halves performed during the 2nd half of the year.
Those are some impressive 2nd half results. I will note that 1970 is the only one in the bunch that managed to close the year positive, and it only saw a gain of 0.1% at the end of the year. (Remember, a 20% drawdown needs a 25% rebound just to get back to even.) But before we get too excited about these results, let’s also look at instances that did not quite meet the 15% criteria, but were down at least 10% in the 1st half of the year.
Results here are much different, with 5 of the 6 instances posting losses in the 2nd half of the year. Of course, this year fits in with the 1st group, so perhaps we get a nice rebound. But instances are low and its been 52 years since the last instance. So I am not incorporating this into my bias at all. Still, I thought the results were interesting enough to share.
The last two weeks have been quite a wild ride. The week ending June 17th was the worst 1-week decline for SPX (-5.79%) since March of 2020. This past week completely made up for that loss (despite it only being a 4-day week) with a gain of 6.45%. Below is a list of the 4 other times since 1930 that SPX followed the largest 1-week loss in a year with a complete reversal the following week.
1966, which is shown in the chart above, was the only other instance that the losing week was the worst in over 2 years. Results were overwhelmingly positive 1, 2, 3 and 4 quarters out for all 4 instances.
I don’t put a lot of faith in something that has only happened 4 times before, but if you are looking for a possible hopeful sign…well…this might be one.
For years I have published research about the Quantifiable Edges Capitulative Breadth Indicator (CBI). In fact, the 3rd blog post ever, back in Jan 2008, was about the CBI. On Friday, the CBI jumped from 7 to 11. Over the years I have generally regarded any reading of 10 or higher to be a bullish indication. The study below is one I have shown many times. It looks a performance following a spike to 10 or higher. I also added a 200ma filter.
We see that most often during established downtrends the high reading has marked a point where the selling is nearly over (at least temporarily), and bounce is likely to begin. The 17-day curve can be seen below.
That is solid consistency. It is a wild market we are in. Based on the spike in the CBI to over 10, it appears that there is a good chance that we’ll see a bounce, and perhaps some intermediate-term relief fairly soon. Plenty more CBI research can be found here.
A more detailed version of this study appeared in the Quantifiable Edges Subscriber Letter this past weekend, along with several others. Take a free trial here if you want to see it.
In last night’s subscriber letter I decided to take a similar look, but only examining instances during long-term downtrends. Below is a look at how the SPY has performed from 2pm to 4pm on Fed Day’s where SPY closed below the 200ma the day before.
Choppy and inconsistent with net results slightly negative. Not an appealing curve for bears or bulls. But now let’s look at performance from the close the day before the Fed Day until 2pm of the Fed Day during downtrends. (Announcements are typically at 2pm EST.)
Numbers here are very strong and the curve is incredibly consistent. I’ll also note that the last 11 instances, dating back to November 2011, all saw a gain.
This all reinforces my view that it is generally better to exit Fed Day-specific trades prior to the announcement. Of course there is a risk of missing out on a bigger Fed Day rally, but over time that has been the more consistent and more profitable approach.
A few years back I wrote about Zweig Breadth Thrusts in some detail. The Zweig Thrust takes a 10-day exponential moving average of the NYSE Up Issues %. It looks for a move from < 40% to over 61.5% within a 10-day period to signal there has been a breadth thrust.
Over the last 3 days we have essentially what could be considered the inverse setup trigger. The NYSE Up Issues % 10ema has fallen from above 61.5% to under 40% in 10 trading days. Rather than a breadth thrust, we have seen a “breadth collapse”. I looked back at other times where such an “Inverse Zweig Breadth Collapse” has occurred. I found the setup to be rare. The full list of instances can be found below.
There have only been 10 instances dating back to 1928, and the current instance is the 1st one since 1943, when the S&P 90 was the primary S&P index. The modern S&P 500 inception occurred in 1957. So it is fairly amazing to see a breadth collapse so unusual that it has not happened in nearly 80 years. In looking at all the instances between 1931 and 1943, results were not encouraging for the bulls. On average the market was down 2.5% five days later, down 4.3% ten days later, down 7.1% one month later, down 7.3% three months later, and down 7.7% six months later. I will not put a whole lot of faith in a signal that has not triggered in nearly 80 years. But the fact that we have seen such an extreme breadth collapse the last few days does seem notable. And it sure doesn’t appear to be market positive for the next few months.
The strong breadth readings over the last few days triggered one of my oldest and most favorite breadth studies.
There are a lot of positive numbers and the edge appears to be to the upside. Results between 70 and 90 days appear especially strong and consistent. Below is the full list of 80-day results.
That is quite the winning record. Of course every time is different, and we are facing an unfriendly Fed at the moment. But similar breadth thrusts have a strong track record of seeing more gains over the intermediate-term.
In this weekend’s subscriber letter I showed several studies suggesting the Friday selling could be an overreaction and that a bounce early this week appears likely. I noted that Fridays are often viewed as the day most likely to have a downside overreaction. This is partly because people may be looking to bail out ahead of the weekend and avoid getting hurt by further bad news. This concept is something I have discussed in the past here on the blog, mostly recently last November.
In recent years the bounce has not always occurred on Monday. Sometimes we saw more selling on Monday before the rebound started on “Turnaround” Tuesday. A couple of interesting charts from the letter are shown below. They look at performance following Fridays that close down over 2.5%. The 1st is a profit curve assuming a 1-day holding period. The 2nd chart assumes a 2-day holding period.
So in recent years the bounce has not as reliably begun on Monday, but it almost always did by Tuesday afternoon. We’ll see if form holds this week and we get a bounce in the next few days.
Stock market performance leading up to and around many holidays has often been bullish. This is something I have written about several times over the years. Holy Thursday is one such day that has done quite well. I have shown Holy Thursday stats a few times in the past. The chart and statistics below are all updated and zoomed in from previous studies to just show the last 24 years.
Strong numbers and an impressive move from lower left to upper right suggests the market could have a seasonal wind at its back on Thursday.
As I have documented numerous times, “Turnaround Tuesday” has generally been the best day of the week under many circumstances for the market to begin to mount a bounce. Below is a simple look at 2-day pullbacks when SPX is below the 200ma and we are heading into Tuesday.
The numbers appear quite bullish and the curve has seen an upward acceleration. Turnaround Tuesday’s have provided many edges over the years. Traders may want to keep this in mind on Tuesday.
There has been a lot of talk recently about yield curve inversions and whether that means a recession is on the way, and how soon? And if there is a recession, will there also be a bear market? I decided to forget about economic forecast and just look at how the SPX did after a curve inversion. I looked at both the 2yr/10yr and the 3mo/10yr combinations. For the study I used Norgate Data, and looked back as far as my database went, which was 1976 for the 2yr rate and 1981 for the 3mo. Results can be found below.
Note that 21 trading days is approximately 1 month. So 42 days is two months, 126 days is 6 months, 252 days is a year…you get it.
Not many instances to build out a case here. Some good and some bad numbers. More bullish than bearish. Overall, the initial inversion does not seem to be a great timing signal. Academics can argue and tv talking heads can blather about potential consequences, but traders should probably look to better timing devices to make their market judgements. I don’t see myself factoring this into any trading decisions.
