A look at the February selloff and the Quantifiable Edges CBI

As the early February volatility explosion unfolded, it was difficult to anticipate when the selling would reach a level that the market would find a bottom (at least temporarily). The selloff exceeded historical levels based on % changes in range and volatility increases. One indicator that once again demonstrated its worth was the Quantifiable Edges Capitualtive Breadth Indicator (CBI). The CBI quickly spiked to 25 and then continued up as high as 31 in the ensuing days. I noted the initial spike here on the blog, and tweeted (@QuantEdges) updates over the following days and weeks.

 

One CBI-based strategy I have shown in the past involves going long the S&P 500 when the CBI spikes to 10 or higher, and then exiting the position on a return to a “neutral” CBI level of 3 or lower. The chart below shows how this approach would have worked out during the February market.

2018-03-04-1

In this case, the selling was not over, and another brief leg down would have had to be endured to take advantage of the strategy. But those that utilized the edge and showed the fortitude to hold until the CBI again turned neutral were again rewarded.

 

Below are updated hypothetical stats for the strategy.  ($0 commissions assumed.)

2018-03-04-2

Numbers here are quite impressive and suggest a strong bullish edge. Lastly, the profit curve:

2018-03-04-3

The strong, steady upslope serves as some confirmation of the bullish edge suggested by the numbers.

 

Even before starting Quantifiable Edges in 2008, the CBI was one of my favorite indicators for identifying selling capitulation that is highly likely to be followed by short to intermediate-term market rallies. And it has always been a staple at Quantifiable Edges. In fact, the 3rd post I ever did was about the CBI.

 

Later this week I will be releasing the most detailed analysis I have ever done of the CBI. I will show charts of market action during and after every spike since 1995. I will share a detailed spreadsheet with breakdown and stats of all such CBI spikes. And I will show many additional studies as well as strategies traders can incorporate into their own methods. And it will all be free of charge. I’d encourage traders to keep an eye out for it.

 

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

SPX Performance After Three 1% Down Days

Last night I looked at 3-day pullbacks a number of ways in relation to current market conditions. I thought blog readers might find the following interesting.

 

I noted that SPX closed lower by greater than 1% for the 3rd day in a row on Thursday. In the past, that has often been followed by gains the next day. But times where is wasn’t…well, take a look at the chart below.

2018-03-02-1

Good chance we bounce in the next few days. Some chance things get a whole lot worse.

 

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

A Two-Day SPY Pattern Suggesting A Bullish Edge For Wednesday

SPY gapped up and closed lower Tuesday after leaving an unfilled up gap on Monday. This triggered a simple study that I have examined a number of times over the years in the subscriber letter. The study can be found below.

2018-02-28

The numbers here all look solidly bullish, suggesting a potential 1-day upside edge.  Traders may want to keep this in mind today.

Of course if you read the blog regularly, you may realize this is in conflict with a study I showed over the weekend that suggested possible bearish implications through Wednesday.  That’s ok.  Not all evidence will always point in the same direction.  My favorite tool to help establish my bias when examining conflicting evidence is the Aggregator.  In any case, it will be interesting to see how Wednesday plays out, and how that sets up the next few days.

 

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

The Negative Impact Of Friday’s Low Volume

I mentioned in a Tweet on Friday that the low volume on Friday’s rally was a bit concerning. The study below is one I featured in the subscriber letter this weekend. It examined other times substantial rallies occurred during uptrends on very light volume.

2018-02-25

Stats here suggest a downside edge. Perhaps not a huge edge, but in my view one that appears strong enough to warrant some consideration when establishing my short-term bias. So traders may want to keep this in mind as we begin a new week. I will also note that I ran the same test, but switched the volume requirement to “NOT the lightest in 20 days”. Of course there were many more instances. With volume not coming in extremely low, the average trade flipped to moderately positive across the board. This suggests the low volume is a factor.

 

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

Sunday Marks the Quantifiable Edges Subscriber Letter’s 10th Anniversary!

Sunday Feb 18th marks the 10th anniversary of the Quantifiable Edges Subscriber Letter.  I can hardly believe I have been writing it for 10 years, but it is true.  A few highlights and anecdotes from the last 10 years…

 

·         When the letter began, there was not even a website – just a blogspot blog and a Paypal button.

·         My 1st subscriber was Mr. Norwood.  I remember distinctly, because my wife grew up in a town called Norwood and I thought perhaps that was a good sign.  It was, and Mr. Norwood is still a subscriber today!

·         I’ve been honored to be able to speak at several conferences, contribute to books, and be interviewed by magazines.  But my favorite interview was with Ben Powers for Your Trading Edge magazine.  What made that one special was that they put my name on the cover in big bold letters.  My daughter was 6 years old at the time, and after she saw her dad’s name on a magazine cover, she told several teachers and camp counselors that I was famous.  (Despite my fame, in 10 years I have yet to ever have someone stop me out in public and ask if I am Rob Hanna of Quantifiable Edges.)

·         I have made some good friends through Quantifiable Edges – some who I ended up doing work with and spending lots of time with, and others that I still have never met in person!  One friend I spoke with for 8 years before meeting.  Then on a family trip a couple of summers ago I finally got the pleasure.  Jeff took me all over Orcas Island, and we went out on his boat, and my family and I got to see whales in the wild for the 1st time!  It was the thrill of the entire summer!

·         Another friend I have made along the way is Tom McClellan.  Tom’s father Sherman (along  with Tom’s mother) created the McClellan Oscillator.  I mention them here because several years ago I spoke at an event in California.  (I live in Massachusetts.)  Sherman lives in California.  And when Tom heard I was going to speak at this particular conference, he told his dad he should go see me.  So I knew Sherman was coming to see my speech, and I was a bit nervous to meet him and have to speak in front of him and a big crowd.  The function room where I was speaking had about 250 chairs set up.  Sherman was one of the 1st people to arrive, and he came up and introduced himself and we spoke briefly before I had to talk.  My talk was about Fed Days, and I guess I chose a topic that was too narrow for many people, because of those 250 chairs, 238 of them went unused!  I flew all the way across the country to speak in front of 12 people!  And Sherman McClellan, whose work I greatly admired, and who I had hoped to impress was one of those 12.  I was mortified.  He couldn’t have been nicer.  But I was still mortified.

 

So thanks to Quantifiable Edges I have had some wonderful experiences, and some trying ones.   I’ve met many terrific people.  I have also traded through a lot of ups and downs – both professional and personal.  Many of my market calls and trade ideas have turned out very well, and a few have not.  But when it comes to QE, perhaps the stat I am most proud of is this one:  in 10 years there has never been an NYSE opening bell that was not preceded by a Quantifiable Edges Subscriber Letter.  Through ups and downs, deaths, surgeries, illnesses, vacations and more, I have managed to get a letter out every single night for the last 10 years.  (I didn’t always WANT to write it, but I always have.)  And at this point I am as energized as I have ever been to continue writing.  But I will say this…some time in the next 10 years I intend to take a day off…maybe even a few days off.  Not yet, though.  Sunday will mark the 10th anniversary, and I’ll once again be writing the letter, looking to find Quantifiable Edges that will help me and my subscribers.

 

Thanks to all the readers and subscribers who have supported the Quantifiable Edges Subscriber Letter over the years.  And if you don’t have a subscription yet, then there is no time like the present.  Don’t wait another 10 years before trying it!

Follow Through Days That Occur With Moderate Breadth & Moderate Volume Have Struggled Historically

One notable bit of evidence that emerged on Wednesday was the fact that it qualified as an IBD Follow Through Day (FTD). I have done a lot of research on FTDs over the years. Much of that research can be found on the blog. Here is a link.

2018-02-15

The failure rate here is substantial no matter how you look at it. A short-term downside edge is suggested which largely plays out in the 1st 2 days. Every instance closed below the entry price over the next few days. And these FTDs have demonstrated a paltry 20% success rate. All these stats are impressive and point to a downside inclination over the next few days.

(Definitions for “successful” rallies as well as FTD determination criteria can be found in this post from 2008.)
 

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

Quantifiable Proof That Chicks Don’t Dig Me

Google Analytics is a tool that is used by most websites that allows you to evaluate activity on your site. You can see how much traffic there is, where that traffic is coming from, and more. With it being Valentine’s Day, I thought I would see how popular Quantifiable Edges is with the ladies. Below is the gender breakdown of recent site visitors from Google Analytics.

2018-02-14

My wife would be happy to see this…except she’ll probably never see it. Because like most women, she doesn’t ever visit Quantifiable Edges.

Happy Valentine’s Day!

 

 

Want quantitative market research delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

What SPY’s Gap Up, Reverse Down & Rebound Back Up From Friday Suggest For This Week

The sizable gap up, pullback, and then move back higher on Friday triggered an old Quantifinder study for the 1st time in a long time. Below is the full list of trades with a 5-day holding period.

2018-02-11

All 8 instances saw run-ups of at least 1%, and they all closed positive. While instances are low, the initial inclination appears quite bullish. This study may be worth some consideration when thinking about this upcoming week.

 

 

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

Thoughts On Dealing With Historically Abnormal Markets

I have discussed some lately that the market is acting outside of historical norms. Thursday’s action reinforced that. The pullback has come so fast and been so extreme that it is going beyond even many of the most extreme moves in similar situations. For instance, I looked back to 1960 with the SPX for to find other times SPX closed down > 10% from a 250-day closing high within 2 weeks of that high. Thursday marked the 1st time this has ever happened. If I loosened the criteria to a 9% drop from a 250-day high within a 2 week period, then there were 2 instances. They occurred in 1980 and 1986. Below are charts of them.

2018-02-08

I don’t read too much into just two instances. They did both bounce in short order, but more significant to me is that SPX is doing something it has never done before with the 10% drop from a 250-day high within 2 weeks. Running the same test on the Dow I found only 2 instances. They occurred in 1928 and 1933. So you need to look back 85 years to see such a sharp drop from a high level.

Trading based on historical probabilities becomes more difficult when the market does not comply to historical norms. And when action is so extreme (115% 1-day VIX moves, 10% drops from all-time highs within 2 weeks, etc) that there are little or no comparisons, then keeping some powder dry and not blowing up your accounts becomes more important than timing entries perfectly. It is tough to quantify risk in a market like this. While a strong bounce in the next few days appears highly likely, using too much leverage may not allow an account to survive a few days. Stay in the game. Trade well.
 

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

When The Quantifiable Edges Capitulative Breadth Indicator Hits 25

As I Tweeted out (@QuantEdges) as we approached the close yesterday, the CBI spiked on Monday from 6 to 25. Twenty-five is an extremely high number. And while readings that high have been rare, they have also been quickly followed by a bounce in the SPX. Below is the list of the 5 previous instances, dating back to 1995, and their 5-day returns.

2018-02-06

You’ll note when looking at the dates they were all significant historical market events. Long-Term Capital, Post 9/11 attacks, 2002 crash, 2008 crash, and 2015 China-induced crash. Every instance bounced over the next 5 days, but 3 of them saw further declines between 4% and 8.5% before the bounce occurred. The message appears to be that the market is likely within a few days of a capitulative bottom, but price may have to fall further before the big reversal kicks in. And with XIV/SVXY news pending and possible implications not yet known, it could be a scary ride.

 

 

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

A Closer Look At Historical Performance Following New Fed Chairmen

A couple of weeks ago I did a little study that looked at performance following the induction of a new Fed Chairman. With Jerome Powell starting his new job on Monday, I decided to expand on that study below.

2018-02-04

Obviously it appears to be a bit of a mixed bag. The most positive results came in the 1st 3 weeks (15 trading days). Perhaps the market view the new chairman with enthusiasm. The most negative results were 50-day out. These were also greatly skewed by the crash of ’87 and another giant swoon in 1930. Below I have listed the 13 new chairmen and the results 15 and 50 days after their start.

2018-02-03

As I mentioned a few weeks ago, what stands out to me is that the last 10 instances all saw the Dow higher 15 days later. Looking back as far as I am and using such a low sample size, I do not view these results as significant. But as an exercise in curiosity I found it interesting.

 

 

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

SPX at Highs with XIV at Lows

XIV is an inverse-VIX ETN. In other words, it was designed to generally trade inversely to VIX futures on a daily basis. Since VIX and SPX typically trade opposite each other, you would think that XIV and SPX would often close in the same direction. And you would be right. Of course, XIV depends on more than just the movement in the VIX to determine its price. Among other things, it is influenced by short-term VIX futures movement and the term structure of the 1st couple months of VIX futures contracts. This is not the place to get into a deep discussion of XIV price influences. But it is important to understand that it 1) generally trades inverse to VIX movement, and 2) will often trade in the same direction as SPX. XIV has diverged with SPX in recent days. In fact, while SPX closed at an all-time high on Thursday, XIV closed at the lowest level of 2018. This can be seen in the chart below.

2018-01-25qe

The divergence between the 2 is highly unusual. In fact this is the 1st time ever (since 2011 XIV inception) that SPX has closed at even a 10-day high while XIV has closed at a 10-day low. If we loosen the criteria to only require a 6-day SPX high and a 6-day XIV low we can find 7 previous instances. Their 1-day results can be found below.

2018-01-26

The number of instances is low, but early indications suggest a possible 1-day downside edge.

 

 

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

When New Years Begin With A Steady Stream Of Up Days

The start to 2018 has been fairly remarkable. The SPX has only closed down 3 days so far, while closing up 11 days. That is a substantial hot streak, and one might think that such a strong run to start the year would almost certainly have to pullback soon. So I checked.

2018-01-23

The imminent pullback theory certainly does not seem to work here. All 6 previous instances were higher 2, 7, 8, and 9 days higher. And the worst loser over the 1-10 day period was only 0.3%. The kind of early-year strength we are currently seeing has always been followed by more upside in the past. The years where this occurred were: 1965, 1967, 1976, 1979, 1987, and 2012.

With just 6 previous instances, and only one in the past 30 years, this study does not get me enthused about jumping into a long position here. But it does make me a little more wary of trying to short into this strength.

 

 

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

How the Quantifiable Edges Aggregator uses expectations and risk/reward analysis to establish a reliable market bias

When it comes to predicting market movement, there are an endless number of indicators out there. And they are based on an ever-growing number of ideas. The most common indicators are based on things like price action, volume, breadth, sentiment, government policy, or cross market analysis. But people might also look at sun spots, moon cycles, weather, or any number of things where they find a correlation to market movements. No matter what indicators a trader favors, it does not take long to realize that none of them work perfectly. And even if you find one that is incredibly reliable, it may not be providing strong readings often enough to generate consistent returns. Therefore, every trader I know looks at more than one input or indicator to try and find an edge.

Of course the more indicators you use the more likely it is that some of them disagree. It is rare that a trader will see all of their indicators line up perfectly at the same time. Often price action may be suggesting one thing, while breadth, or sentiment, or intermarket action may be suggesting something else. So traders need to determine whether the mix of evidence is suggesting a bullish or bearish indication, and how strong that indication is.

At Quantifiable Edges, I use indicators in a slightly different way. Rather than simply interpret readings or patterns that I am seeing, I generate market research studies to understand how similar situations have performed historically. It is these studies that I publish in the Quantifiable Edges Subscriber Letter (and sometimes the blog or elsewhere) that help me to establish my market bias. But like other traders indicators, my studies don’t always agree. So the tool I use to help me weight my studies and determine a market bias is the Aggregator. A chart with the Aggregator included can be seen below. It shows readings from December 2017 into January 2018.

2018-01-22q

Each day, the Aggregator takes the current studies on the Quantifiable Edges Active List and creates a composite estimate. That estimate looks out over the next few days, and is represented by the green line. If it is above zero, then estimates are looking for the market to rise over the next few days. When the green line is below zero, then estimaanticipating a decline.

The black line I refer to as the Differential. It shows the difference between recent estimates and actual market movement, and it provides an overbought/oversold indication. When it is above zero, that implies the market has failed to meet recent expectations and could be considered oversold. When it is below zero, that means the market has exceeded recent expectations and could be considered overbought.

Combining expectations with overbought/oversold means there are 4 possible formations on the chart.

  1. Both lines above zero
  2. Both lines below zero
  3. The Aggregator expectation line is above zero and the Differential line is below zero.
  4. The Aggregator expectation line is below zero and the Differential line is above zero.

Formation 1 I generally consider to be a bullish setup. Oversold with bullish expectations is often a good place to buy.

Formation 2 I generally consider bearish. Overbought with bearish expectations can mean a good shorting opportunity.

Formations 3 and 4 I generally consider to be neutral. Positive expectations in an overbought market often don’t offer great risk/reward. And neither do negative expectations and an oversold market. So I will not normally be looking to take on new index positions when the Aggregator is in formation 3 or 4. But both of these still provide valuable information. Oversold with negative expectations keeps me from trying to go long a pullback that does not show a high probability of bouncing. And knowing there are positive expectations when the market is overbought helps me to avoid shorting into overbought situations where the historical indications are for further gains.

To help you easily spot where the Aggregator formation has changed to bullish, bearish, or neutral, I have included arrows and signals on the chart. Keep in mind, the signals occur at the end of the day and suggest a bias for the next few days. As you can see, the Aggregator has done a nice job of anticipating short-term market movements.

The bottom line is the studies help me determine whether there is an upside or downside edge. The Differential measurement helps to assess risk/reward. The Aggregator chart combines them and helps me to establish a short-term market bias. The Aggregator chart is published each night in the Quantifiable Edges Subscriber Letter, and it has proven invaluable to me since I began using it in 2008.

 

 

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

A Historical Look At Market Reaction To New Fed Chairmen

Jerome Powell is expected to take over for Janet Yellen as the new Fed chairman on Feb 3rd. A few days ago in the letter I looked at SPX performance after a new chairman takes over. I used the SPX and looked back to 1970. Tonight I decided to take the analysis back to 1923 using my Dow data. Like with the SPX, I found the first few weeks to be the most consistent and interesting data. Once we look out much further, the results become more mixed. So below is a list of past Fed chairman changes along with the 15-day performance of the Dow.

2018-01-21

After 3 rough instances to begin the study, the rest of the chairman have been greeted with enthusiasm by the market in their early days. It did not always last. (Greenspan saw the crash of ’87 after just a few months on the job.) But if the last ten chairmen are any indication, the market rally may continue during February. Or if the market decides it really does not like Mr. Powell, then a Eugene Meyer nosedive would make for a tough few weeks. Personally, this was more an exercise in curiosity, than anything I plan to base a big trade on. Good luck Mr. Powell! We hope you are met with at least as much enthusiasm as Thomas McCabe was.

 

 

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.