Wednesday’s selling followed a slow drift higher and wiped out a few days worth of gains. Below I show results of the 30 instances since 1982 where 3+ days of gains were erased after a rally of 5+ days.
These results appear to provide a solid upside edge over the next 1-10 days. In the past such-give backs after a steady rise have not suffered substantial additional selling. Instead they have often been followed by a resumption of the move up.
The last few years I have shown the “Twas 3 Nights Before Christmas” study along with results for the SPX. It continues to do well. But this year I decided to show results for the Nasdaq Composite. As you’ll see, the seasonality there has been even stronger and more reliable.
The stats in the table are stronger across the board, and the reliability shown at the bottom of the table is nothing short of incredible. Not all my studies are currently bullish, but based on this one I would expect to see at least a mild rise in the Nasdaq at some point in the next several days.
After seeing over 90% of the volume come to the downside on Monday, Tuesday registered a 90%+ upside volume day. A 90% down followed by a 90% up day is something that was never seen from 1970 – 2006. But this is the 10th time we’ve seen it happen since the beginning of 2007. Below I have listed the previous 9 and shown how the SPX has fared the following day.
Though instances are low statistics at this point appear solidly bearish, suggesting a possible 1-day downside edge. The high downside risk vs the small upside reward makes the results especially compelling for the bears.
The Quantifiable Edges Capitualtive Breadth Indicator (more commonly known as the “CBI”) was introduced in just the 3rd blog post I ever did on January 6, 2008. The CBI was devised from a system I use to trade individual (primarily S&P 100) stocks and sometimes ETFs. I devised the system in 2005 and refer to it as my “Catapult” system. The Catapult was designed to take advantage of extreme (often capitulative) selling in these securities. The CBI reading is basically a count of the open Catapult triggers at any one time among S&P 100 stocks. What I noticed early on was that broad triggering of this system was frequently a sign that not only were these individual stocks primed for a bounce, but the market as a whole was also very likely to bounce. I have written an awful lot on the blog about edges that could be found by entering the market when the CBI reached certain levels. Traders that would like to examine any of that research more closely are always welcome to use the “CBI” label at the right hand side of the blog. It will pull up all associated posts.
In February 2008 I began the Quantifiable Edges Subscriber Letter. As part of the letter I included signals for my Catapult System. The success of the Catapult trades over the years has made it an attractive and somewhat popular feature for subscribers. I publish any signals that occur that night in the letter. Then using a limit order the next day I track them in the letter as well. When the exit signal occurs I also note this. The standard exit for a Catapult is at the open the day following the exit trigger. (On occasion I will send intraday updates alerting subscribers that the exit will trigger at a certain level and I am going to exit the trade at the close rather than waiting for the next day’s open.) While many subscribers have profited from the Catapult trades, there has been one common complaint.
The complaint is that since the Catapult is the one system on the site in which I do not reveal the entry and exit criteria it makes the trades uncomfortable for some traders. Not knowing when (or exactly how) the exit will come has prohibited some subscribers from ever entering these trades. They have instead watched many go by and found themselves a bit frustrated that they couldn’t take advantage without understanding the exit criteria. So I set about to solve this issue. Recently I released to subscribers the “Catapult Exit Designer” for Tradestation. The Catapult Exit Designer is open code that triggers entries as listed in the Subscriber letter over the last 4 years (over 250 trades) and allows the users to test their own exit criteria. It generates files for them with results of all trades. And for those subscribers that don’t use Tradestation, I show all the code and explain all the logic behind it so that you may easily transfer it to you preferred platform. In addition I provide sample exit strategies that would have produced results very similar to the actual Catapult exits.
The Catpult trades have done very well over the years. Of those that received fills and were tracked in the letter over 72% were winners, the average trade made 3.35% and the profit factor has been an impressive 3.32. What has most attracted me to this strategy is not just the fact that it has made money, but WHEN. Catapults generally trigger when the market gets scary. They happen when other systems and techniques I employ have sometimes struggled. So they have not only made me money, but they’ve done so when other methods were losing. This was greatly useful in helping me to limit or avoid drawdowns. The timing of the trades can be seen in the chart below. The top of the chart shows the S&P 500. The indicator on the bottom is the CBI. One way I track catapult trades is by “clusters”. A cluster of trades begins when the CBI moves above zero and it ends when the CBI returns back to zero. Historically I found over 90% of clusters would have been net winners had you taken all the trades in the cluster with equal size. (The last cluster that had a net loss was about 2 ½ years ago.) Above each cluster in the chart is a number. That number shows the net additive gains that would’ve been generated by that particular cluster.
(As an example to clarify, the latest cluster closed out in October. That cluster saw 8 Catapults get triggered and filled. They all were winners (not typical). The % gain for each was the following: 2.8%, 6.2%, 5.0%, 11.1%, 5.2%, 8.9%, 0.2%, and 0.3%. Add those numbers up and you get the 39.7% shown above the October cluster. That is what I mean by “additive gain”. It is NOT a portfolio return.)
This year all the trades have come in 4 big clusters. And they all came during sharp selloffs when other methods may have been under stress. I’d be happy to show years 2008 – 2010 but this post is already way too long. If you want to see those results you may use the link below. They are shown there.
https://quantifiableedges.blogspot.com/2011/02/using-qe-to-your-advantage-subscriber.html
The Catapult System has been a favorite of mine and of subscribers over the years. Hopefully with the new Catapult Exit Designer more people can begin to take advantage of the opportunities the Catapult System identifies. If you think there might be some tough times in 2012 and would like to implement a method that can possibly take advantage of them, then the Catapult System may be one option. Both the Catapult System triggers and the new Exit Designer are completely included in a Quantifiable Edges Gold Membership.
One sector that is especially sensitive to Fed Days is the banking sector (BKX). BKX closed down on Monday about 2.5%. The study below looks at drops of 2% or more on the day before a Fed Day, and how the BKX responded on the Fed Day.
Instances are a little low here but the stats are overwhelming. Below I have listed all instances.
Not shown in all the above stats is that the average run-up was 4.2% and the average drawdown just -0.6% during the Fed Day. Overall risk/reward appears strongly favorable for BKX based on the limited sample size.
Over several time horizons op-ex week in December has been the most bullish week of the year for the SPX. The positive seasonality actually has persisted for up to 3 weeks. I demonstrated this most recently in the the blog exactly a year ago. I’ve updated that study below to include last year’s stats.
The stats still appear extremely strong. I normally only show equity curves in my subscriber letter, but I’m in a good mood this morning, so below is a curve using a 5-day holding period.
Aside from the big pullback in 2000, it has been a strong steady move higher. This would seem to confirm the bullish edge.
Many people are aware that several higher closes during a downtrend will often suggest a downside edge. Few people realize how strong this tendency has been in the Nasdaq over the last decade, especially once the index has closed up 4 days in a row. With the Nasdaq now up exactly 4 days in a row as of Monday’s close, the study below is triggering.
Not only has the bearish tendency been strong and consistent, but it has persisted for as long as 2 weeks in most cases.
Are you wondering why the 50-day low filter is included? Here’s why:
https://quantifiableedges.blogspot.com/2011/10/why-shorting-in-this-environment-can-be.html
A quick study for you the morning…
Not much to go by, but perhaps some caution is warranted intraday. I also looked out over the next few days. They appear to be a tossup.
With SPY looking to gap up over 2.5% this morning below is a quick look at all other instances since 2003 when SPY gapped up over 2% immediately following a 20-day low.
Low number of instances, but the suggestion is obvious. I would note, though that although they all finished higher, they also saw some substantial pullbacks from the opening price.
It appears the market is going to buck historical tendencies today and close down. This will be the 11th time since 1960 that the SPX has closed lower on the Wednesday before Thanksgiving. Below I have listed the other 10 instances along with the SPX performance on Friday (the day after Thanksgiving).
The 2 instances that I have circled are the only 2 where the SPX closed the Wednesday before Thanksgiving down over 1%.
For what it’s worth I would note that statistics associated with these 10 instances (win %, profit factor, avg. gain, etc) are pretty much in line with all Fridays after Thanksgiving. (It has generally been bullish.)
Thanksgiving has typically shown some pretty consistent seasonality. Both the Wednesday before and the Friday after have exhibited bullish tendencies while the Monday after has been slightly bearish. Last year I showed a table breaking it all down by day. Today I decided to show a profit curve that represents simply owning the SPX from Tuesday’s close through Wednesday’s close.
Futures are down quite a bit this morning, so it may be difficult to keep this curve moving higher today, but over the years the next couple of days has been a pretty good bullish bet. And when seasonality is this consistent it is often worth keeping in mind.
The SPY finished at a 20-day low on Friday, and now it is gapping down large this morning. I ran a test to see what kind of intraday edge this might suggest.
The raw numbers appears to suggest a mild intraday upside edge. But take a look at the profit curve.
This curve does not get me excited about buying into this gap down. Instead I will need to see more compelling evidence in order to anticipate an intraday move higher.
I am at the excellent Las Vegas Traders Expo for the next few days. I plan on seeing several presentations and will be speaking myself on Saturday. I’d love to meet some blog readers, so please come introduce yourself if you see me.
More information here:
https://www.moneyshow.com/tradeshow/las_vegas/traders_expo/main.asp
And here’s the info on my presentation:
Monday’s extremely low volume could be a short-term bearish sign. The 0.96% drop did not quite qualify for this 2008 study, but below is another past study that does exemplify the low-volume issue. Rather than index volume, it uses SPY volume. Either one would yield similar results in this case.
While not the most overwhelming edge we’ve ever seen, it does seem to strongly suggest that caution is warranted.
Veterans Day is one of the few US holidays when the bond market is closed and the stock market is open. Columbus Day is another. But while Columbus Day has exhibited some quantifiable edges, Veterans Day has not. Veterans Day is celebrated on November 11th (or the closest weekday to November 11th) each year. For a brief period from 1971 – 1977 it was celebrated on the 4th Friday of October. Below is a profit curve that shows how Veterans Day has performed over the years.
While for a good long while it appeared Veterans Day may provide an upside edge, the curve topped out in 1992. Since then it has been mostly lower. Happy Veterans Day and thanks to all veterans!