Friday, February 29, 2008

Are IBD Follow Through Days After Day 10 Less Reliable?

One of the interesting claims that William O’Neil make about Follow Through Days is that they are less likely to work if they come more than 10 days from the potential market bottom. As part of the study on Follow through Days, I decided to test this. Those who missed the first several installments of this study may want to click on the “IBD Follow Through Day” label lower down on the right hand side of the page. This will be the 9th installment in the series.

Using the original basic assumptions of an 8% decline needed and a 1% up move on the Follow Through Day (as opposed to the current 1.7% requirement that I found to be less effective), I reviewed all FTD’s listed in the study.

A Follow Through Day actually occurring after day 10 was a fairly unusual occurrence. Downtrends and bottom formations typically carry significant volatility, so a strong, high-volume move off a low normally occurs before day 10.

Of the 65 FTD’s listed in the study, only 8 of them occurred on day 10 or later. They are listed below along with the FTD Day # and whether they were “successful” or not. (Success was defined in Part 1 - Are They Predictive?)

Seven of the eight FTD’s that came after day 10 were successful according the study. While the sample size may be too small to claim significance, there certainly seems to be no credence to the claim that FTD’s after Day 10 are LESS reliable. In fact, the opposite appears true. Seven out of eight seems especially impressive considering the fact that only 55% of the FTD's in the study were successful. I suspect one reason for this may be that the delayed FTD allows stocks more time to carve out proper basing formations before the market attempts to launch higher. In light of the facts, it seems a curious claim for IBD to make.
The takeaway here is: next time a follow through day doesn’t come immediately, traders shouldn’t fret. The chance of success is likely higher.

Thursday, February 28, 2008

VIX System Discussion Follow-up

After my VIX post a couple of days ago, “Frank” made some interesting comments in the comment section. He duplicated part of the system I discussed, but his results were very different. Using a 15% VIX stretch (a close 15% above the 10ma) he found 17 of 18 short trades since 2004 to be winners if you wait for a reversion to the 10ma to exit.

The big differences sparked some good conversation over at the Daily Options Report, which has picked up on this study. So I felt I should clear up a few things.

1) I mistyped in my post. Options were not used in my testing - futures were.

2) My "system" looked at several measures of overbought/oversold - it generally entered if more than one triggered and exited when the triggers were removed. The 15% VIX stretch was one example.

3) I did not scale in. If a trigger occurred, the system was basically “all in”. I’m sure scaling as Frank did, would help.

4) I don’t think 1, 2, or 3 above really matter much. Looking back to the beginning of 2004 I did a quick count of the number of days that the VIX closed at least 15% above its 10ma. I counted about 70 times. Frank only found 18 times. I suspect Frank was looking for a 15% stretch in the futures. I was looking for a 15% stretch in the actual VIX.

My point was to show that the futures and options could not be easily traded based on the action of the VIX alone. A stretch in the futures successfully reverting to its mean doesn’t surprise me. I’ll bet in many of those 18 cases the VIX index would have put the system into the trade even earlier and at a worse price – turning several of Frank’s winners into losers.

To sum up – the VIX is not tradable. Buying calls or puts based on VIX action as I’ve seen suggested in the media looks good on the surface but is a horrible strategy. Frank has generously shown that astute traders CAN successfully trade VIX futures by focusing on VIX futures action.
Below is a chart of the VIX(green) and front month VIX futures (orange). The chart is a few months old because I don’t have my futures data fully updated (I don’t trade it). The takeaway from this chart is that sharp moves in the VIX “cash” index are dulled in the futures. Note how while they generally move in the same direction, the cash VIX will oscillate around the futures. If you look at futures further out than front month, this “dulling” of the VIX moves by the futures will be even more pronounced.

Wednesday, February 27, 2008

The Intermediate Term Significance Of A Lagging Nasdaq

One indicator I look at comes from Gerald Appel’s book “Technical Analysis – Power Tools For Active Investors”. It is a relative strength measure of the NYSE vs. the Nasdaq looked at on a weekly chart. Without going into great detail, the premise behind the indicator is that the market tends to perform better when the appetite for Nasdaq stocks is greater than the appetite for NYSE stocks.

Part of this is due to the higher volatility of the Nasdaq, and part of it is due to investors willingness to speculate more aggressively when their outlook is positive. Whatever the reasons behind it, the indicator has been a pretty good barometer over the years. Mr. Appel suggests using a 10-week relative strength indicator to measure this phenomenon. This is what I’ve done in the chart below. The way the indicator works is as follows: When the red line is above the yellow line, the Nasdaq is leading the NYSE. When the red line is below the yellow line, the Nasdaq is lagging the NYSE. (Click to enlarge).

Since 1971, close to 100% of the market’s gains have occurred when the Nasdaq is leading rather than lagging. As you can see from the chart above, the Nasdaq has begun to lag badly. I decided to look and see how the market has performed under similar conditions in the past.

Using the NYSE composite as the “tradable” vehicle, I set up the following rules: 1) The NYSE must make a new 5-week high this week. 2) The current NYSE/Nasdaq ratio must be at least 3.0% below the 10-week EMA. (The red line must be 3.0% or more below the yellow line.) 3) The difference between the current NYSE/Nasdaq ratio and the 10-week EMA must be at it’s widest point in the last 5 weeks. (The red line must be farther below the yellow line than it has been in at least 5 weeks.)

If all three conditions are met, sell the market short on the close. Cover X weeks later. Results below:

As you can see, a Nasdaq lagging as badly as it is right now has been quite bearish historically. The bearish tendency carries through over a significant period of time as well (10 weeks.) If the Nasdaq could begin to assert a leadership role, that could help the current rally attempt greatly. If not, bulls better hope it’s different this time.

Tuesday, February 26, 2008

Do Reliable Oscillations In The VIX Make VIX Options An Easy Profit Vehicle?

I’ve seen some articles in the press over the last few months suggesting that one way to profit in volatile markets is by trading VIX options. They typically make it sound easy. “You don’t even need to know the direction of the market. You just need to determine whether volatility is likely to rise or fall. If you think volatility is going higher, you can buy VIX call options. If you think its going lower you can buy VIX put options.” The problem with this logic is that VIX option prices do not follow the VIX index. They follow VIX futures prices. A couple of months ago I decided to quantify how much this really matters.

It is well known by traders that the VIX has a strong tendency to oscillate. Therefore, when people consider trading the VIX, they many times think mean-reverting strategies will work best. I took some simple mean-reverting strategies and applied them to the index to see what kind of returns I would get. Two examples were: 1) Short the VIX if it closes 15% or more above its 10-day moving average. Cover when it closes below its 10-day moving average. 2) Short the VIX if it closes at a 10-day high. Cover when it closes below its 10-day moving average. In both cases the opposite stretch would apply for purchases. Not surprisingly, they worked. What was intriguing was HOW WELL they worked. I then combined these strategies with a few others to create an indicator which would signal to me when the VIX was stretched and due for a reversal.

Assuming you treated the VIX as a security and allocated a certain dollar amount whenever you bought/shorted it, over the last 3 years my simple system would have returned about 170% per year based on raw returns (no commissions or slippage).

Now, to see the effect that trading futures would have on the system, I downloaded all the historical data from CBOE and ran the trades through using front month options. I performed rollovers those times when the future expired before the trade closed. Note that the entry and exit triggers were based on the action in the VIX – not in the futures. The purpose of the study was to see whether someone could trade futures/options based on the action in the VIX index. The results? Instead of returning 170%/yr over the last 3 years, the system now returned 5% total!! Factor in some commissions and slippage and my incredible system is now a money-loser.

Moral of the story: Be careful when trading VIX options / futures. Simple systems which look spectacular on the VIX cash index simply do not translate.

Below are some informative links which also discuss this issue.

Monday, February 25, 2008

Surge Leads To Breakout - Is That Good?

More bailout talk today led to another surge of buying. The major averages closed strongly positive on good breadth and increased volume. Technically notable is the fact that the triangle pattern in the S&P was broken to the upside. An argument could be made that the triangle broke to the downside on Friday and that break quickly led to an upside reversal. Based on the definitions in the studies I laid out last week, there was no break on Friday. The first breakout came today. In this case I’ll stick with my definitions because those are the ones that were used to develop the statistics. The minimum target based on those studies' triangle measurements would be 1446.57. Of note is that over 70% of these patterns have failed to reach their target before dropping below the lower triangle line. In this case a failure would mean a move below 1327.04. It appears looking for an entry to fade this breakout may provide the greatest edge.

The surge the last two days which broke this market out has certainly been impressive. I looked back to see how the market had reacted following similar boughts of strong buying. Below is a table displaying how the market has performed near-term following two strong days of buying during a long-term downtrend.

Historically, fading these surges has provided a decent edge. As can be seen in the picture above, not all breakouts are good news. Between the 2-day surge and the triangle breakout study it appears the edge over the next few days is to the downside.

Sunday, February 24, 2008

Late Day Surge Doesn't Help To Establish Direction

While the move down Friday morning may have violated trendlines some technicians drew on their charts, it did not signal a break for either of the triangle testing methods I laid out last week. For violations to be official I currently would need to see one of the following: 1316.75 broken on the downside for method #1, 1369.23 on the upside for method #1, or 1367.94 to the upside or 1327.04 to the downside for method #2. Therefore, as per the studies, we are still waiting for a break to look to fade. The studies on Thursday night suggested fading the breakout could be the highest-odds play.

While it failed to move the S&P 500 out of its range, the action Friday certainly wasn’t mundane. A late-day reversal led to a sizable upside surge – moving the indices from squarely negative to squarely positive. I decided to look and see whether past late-day surges had led to a directional edge over the days following. Looking back 30 years I was only able to find 22 other times where the S&P 500 gained over 1.5% in the last hour of trading. Results following those instances were mixed. Over the next one and two days the market closed higher 10 of 22 times. Over 3 and 5 day periods the market increased 14 times. Looking our further (10, 15, 20 days) results were close to a coin flip. The win/loss ratio in all these instances was close to 1 as well.

In other words – the late day surge on Friday seems to have no predictive value when looking out over the next few days. All it managed to do was put the market squarely back in the middle of its range. Traders with time-frames longer than intraday may want to continue to wait for a range resolution before getting aggressive.

Friday, February 22, 2008


The formation that everyone seems to be focusing on at the moment is “triangles”. The major indices are all making them and the media and blogging community have taken notice. The standard line is that triangle formations show a contraction of volatility. A break out of the formation can lead to a sharp move in the direction of the breakout. So the big question being asked is “Which way is it going to break?” To me a better question seems to be “What is likely to happen after it breaks?”

The profit potential of a triangle breakout is normally determined by subtracting the low of the triangle from the high and adding that to the breakout point. Triangles are one of those formations that are easier to spot on a chart than they are to program, but I attempted it two different ways.

I first looked at formations where the most recent swing low was higher than the previous swing low and the most recent swing high was lower than the previous swing high. A breakout was defined as a move outside the nearest swing high or low. The profit target was defined as the distance from the entry point plus (or minus for shorts) the highest swing high minus the lowest swing low. A “failure” was defined as a reversal through the opposite swing point in the triangle.

The second way I defined a triangle was a bit simpler. I looked at weekly bars and defined any two consecutive “inside bars” as a triangle. Here again I set a profit target using the biggest bar and defined a stop point as a move through the opposite end of the most recent inside bar.

In both cases the success rates were highly disappointing. Looking at all S&P 100 stocks over the last 15 years, method 1 posted a success rate of about 38% and method 2 a success rate of 30%. The success rate in the S&P 500 Index over the last 30 years was 30% for method 1 and 27% for method 2.

Some traders may still be tempted to play the breakout because the potential reward is higher than the risk. Even so, method 1 was only slightly profitable as gains outsized losses by a mere 1.03 to 1 without factoring in commissions or slippage. Method 2 showed net losses.

It seems to me that the best way to trade these kind of triangle formations is not to play the breakout. Since about 2/3 of the breakouts I tested eventually failed, I’d rather wait for the breakout to occur and then evaluate possible reversal areas that could offer a more substantial edge.

Thursday, February 21, 2008

Will The Reversal That Reversed The Reversal Reverse?

Tuesday night’s study suggested upside was likely over the next 1-5 days following Tuesday’s gap up and reverse lower. In my Subscriber Letter Tuesday night I suggested a gap down Wednesday morning to $134.50 or lower should offer compelling risk/reward for the long side. As luck would have it, the market actually gapped down and reversed up. I was glad to hear some traders were able to take advantage of this. So now the market has gapped and reversed two days in a row – leaving some participants scratching their heads and suggesting this market is acting crazy. In reality, it’s played out according to historical norms.

So is there any historical edge after the market gaps and reverses 2 days in a row as it now has? Assuming Day 1 is a gap up and reversal lower and Day 2 is a gap down that reverses higher I ran some tests. I tried running the tests with a number of different parameters, including adjusting the size of the gap and the size of the reversal. Generally this two-day pattern has led to short-term weakness – in effect reversing the reversal that reversed the original reversal. (Huh?)

As an example let’s look at the NDX. If you require a gap of at least 0.25% each day and a reversal of at least 1% from the opening price, the NDX has closed lower between 57.5% - 61% of the time over the next 1-3 days. Losses generally outsized gains as well. The pattern has occurred 41 times since 1986. The next day the market lost 0.8% on average. It lost 0.9% on average over the next 2 days.

The downside edge here isn’t huge but it appears overly optimistic to think today’s rally is going to send the market off to the races. In fact, the market seems more likely to lose a little over the next day or so and find itself stuck in the same range it has been.

Tuesday, February 19, 2008

An Ugly Gap Failure?

The market gapped up big today but couldn’t hold on to those gains as the S&P 500, Dow 30, and Nasdaq all finished in the red. Reading financial columns tonight you’ll be told that the inability of the market to hold on to its gains was a bad thing. The large black candlesticks are ugly. Market participants are selling into strength. All of which means there is likely more downside to come. But is any of it true?

I ran a study on the Nasdaq 100 tonight to find instances where the index gapped higher by 1% or more and then closed lower on the day. I wanted to see whether the negativity tended to carry through to the next day or next several days. Here’s what I found.

Going back to 1986, the NDX has had 52 occurrences where the market gapped up over 1% and then finished negative on the day. It closed higher the next day 54% of the time. The chance of the NDX being up on any day over the period was also 54%. Over the next 3 and 5 days the chances of the market being higher were 62% and 65% after a large failed gap up. The NDX has closed higher between 55.5% and 56.5% of the time over all 3 and 5 days periods since 1986. So looking out over three and five days the supposedly negative reversal actually seems to be a positive for the market.

Even more compelling is the average gains realized over the period. For the 1-day following the occurrence, the market gained on average 0.4%. An average day in the NDX since 1986 saw a gain of 0.06% - much lower. An average week for the NDX was about a 0.3% gain. The average week after the failed gap up? A 1.3% gain.

If you thought today looked bad and you were worried about the market going forward because of it – you shouldn’t be.

Monday, February 18, 2008

Quantifiable Edges Subscriber Letter

I'm pleased to announce the Quantifiable Edges Subscriber Letter is now available. Those who signed up to recieve the first week's issues free of charge should have already have them in their mailbox. I decided to make the first issue available to anyone. Just click here. Trial subscribers will also be receiving a primer on "How To Get The Most Out Of The Quantifiable Edges Subscriber Letter". Should you wish to receive the primer and the rest of this week's issues for free, just drop an email to with your name and email address. Any questions or comments about the Subscriber Letter may also be sent there.

Subscription information is available here.

I hope you all enjoy it.


Friday, February 15, 2008

Follow Through Days - Better After Small Or Large Declines?

As someone pointed out in the comments section today, Investors Business Daily finally declared a Follow Through Day (FTD) on Wednesday. Therefore, I thought I should post the next study in my series on Follow Through Days. When we first started I had planned on this being Part 4, but the study has evolved as the market moved in interesting ways. For those who missed the first several parts of the series, you may click here for the full shebang. This will be the 8th post on the subject. When I am finished I hope to have the most complete and accurate information available anywhere on Follow Through Days.

Today I will try and assert whether Follow Through Days are more reliable after small or large declines. As you may recall, in the standard test I set up initially, I chose a decline of 8% to be required before a FTD would be looked for. To my knowledge, how deep or long a decline must be before someone can begin looking for a FTD has never clearly been defined. In September of 2005 IBD suggested that FTD’s can be useful even after pullbacks as small as 5%. I decided to look at how FTD’s performed after pullbacks of varying degrees of market declines. Below are the results from December of 1971 through today. Success parameters are laid out exactly as they were in Part 1.

After a decline of 5% or more – 115 FTD’s, 63.5% successful.
After a decline of 8% or more – 64 FTD’s, 54.7% successful.**
After a decline of 10% or more – 48 FTD’s, 48% successful.
After a decline of 12% or more – 36 FTD’s, 44.4% successful.

Generally, the deeper the decline, the less likely it is that a FTD is going to be successful. It is reasonable that more serious selloffs have more difficulty reversing. It is a bit disappointing though that a tool which is billed to signal the end of a market decline seems to fair worse when it’s needed most.

** This was the base test looked at previously.

Would you like detailed market analysis, CBI sector coverage and quantified actionable trade ideas from me several times per week? The Quantifiable Edges Subscriber Letter is coming next week! Just send your name and email address to to receive the first week’s issues free of charge. More details should be available this weekend.

Thursday, February 14, 2008

Count To Three

In Larry Connors’ book “How Markets Really Work” he devotes a chapter to “Up Days In A Row vs. Down Days In A Row”. He looks at how the market has performed after it has moved in one direction for 1, 2, or 3 days in a row. Because of the long-term upward drift of the stock market, very few of the studies published in the book show much of a negative tendency. One situation that showed a negative tendency over the following 1, 2, and 5 day periods was when the market was trading below its 200-day moving average and had posted three higher closes. (This happened today.)

The book looked at data from 1989-2003. Over that period the following results were shown:

Since the end of 2003 it has only happened 4 times until today. Those dates were 7/29/04, 8/17/04, 8/26/04, and 11/29/07. So Larry’s data for this study remains pretty current.

I tested back 35 years and found the results to be similar. About a coin flip on direction, but downside risks were higher than upside rewards giving the next 1-5 days a negative expectancy.

This study on its own is not a reason to sell short. It should serve as a caution though for those holding long positions. Rallies during longer-term downtrends are not normally very persistent. Long-side traders will many times shorten their time frames while short-sellers use the bounces as an opportunity to add more short exposure - all causing a lot of backing and filling. Also note that two of the highest volume days of the last two and a half weeks were on January 31st and February 1st. A lot of people bought a lot of stock on those two days. They’ve been underwater ever since. After three up days in a row, the market has moved back into the middle of the range of those two days. We may see a lot of folks here that are relieved to be able to get out near breakeven on their purchases. This should create some additional selling pressure.

Seems to me like a good time to tighten things up on the long side. Aggressive traders could consider taking a shot short here soon if they’re able to find a favorable entrypoint.

Rob Hanna


Would you like detailed market analysis, CBI sector coverage and quantified actionable trade ideas from me several times per week? The Quantifiable Edges Subscriber Letter is coming soon! Just send your name and email address to to receive the first week’s issues free of charge. More details should be available this weekend and the first issue will likely go out some time next week.

Tuesday, February 12, 2008

The Double Support Trailing Stop

Today saw some big swings but did little to change my moderately bullish bias. Rather than post a study tonight, I thought I’d change things up a bit.

In some comments on January 23rd, “Tim” mentioned he remembered an article I’d written a long time ago in which I discussed “double support stops”. (It was actually the first column I ever published). Since then I’ve seen a large number of readers have done searches on “Double Support Stop Rob Hanna” or something along those lines. Anyway, the column can’t be found anywhere anymore. But you want it – you got it. From way back in 2002 (along with a 2002 chart and ticker symbol - the ticker has since changed to HW if you want to pull it on your own software), I give you “The Double Support Trailing Stop”.

The Double Support Trailing Stop

The Concept
Like most traders, I have a whole checklist of criteria that must be met before I will consider purchasing a stock. I spend a substantial amount of time researching potential candidates so that I may find a few stocks that meet my criteria. One of the most frustrating and difficult things for me is finding a great stock, buying it at the right time, and then being stopped out of it just before it reverses and goes on to huge gains. Having traded in multiple time frames over the years I have experienced this frustration with day, swing, and intermediate term trades. To keep myself in winning trades longer and avoid being shaken out early, I developed my “Double Support Trailing Stop Technique”. I will show an example of this technique on an intermediate term trade, but have successfully employed it in many different time frames, and originally created it for my day trades.

The idea is simple. If I require a large numbers of reasons to enter a trade, I should require more than just one reason to exit it if I want to catch as much of the move as possible. I will therefore require two levels of support to be broken before I will PERMANENTLY & COMPLETELY exit a trade. How you define support does not really matter. You can use pullbacks, consolidation lows (or highs), fibonacci’s, moving averages, trend lines, or whatever fits your trading style.

An Example
Being that the below example is an intermediate term trade, I simply used consolidations & pullbacks, along with the 50 day moving average (shown as the rising blue line) for my support levels. I define a pullback to be any bar (or series of bars) with a lower low AND lower high after a stock has made new highs. The stock shown is Headwaters Inc (ticker: HDWR). I purchased this stock as it burst out of a flat base in the middle of March. My initial stop on the breakout would have been 8% from my purchase price, or the low of the base. After breaking out, it moved up sharply for 5 days before starting to pull back. It pulled back for seven days before reversing and moving higher. The low of this pullback is near the 1st green arrow labeled “Support”. It became support in my eyes after the stock move higher from there by making a higher high and higher low on the next bar. When that happened I was able to move my initial stop up to the breakout point. (I consider the high of a flat base or high tight flag to be a level of support also.)

Next, the stock moved up again for several days before making a second pullback down to the green arrow labeled “Broken Support”. The “Broken Support” bar was initially just the next area of support after the stock had pulled back and moved higher. When that happened, I was able to move my stop up under the 1st green support arrow. It became “broken” 5 days later when the next bar labeled “Support” moved below the low of “Broken Support”. If I had exited at the “broken support” level here, I would have been shaken out of the stock and missed the rest of the move. Instead, the stock continued to move up and create higher and higher support levels.

The time came for me to sell this stock at the area marked “Double Support (Sold When Broken)”. It was at this time that the stock moved below two levels of support. The first was the 50 Day Moving Average, and the second was the pullback low just above the “Double Support” green arrow. As you can see, this turned out to be a good time to take profits and exit this trade. Three days later it was trading nearly 30% lower.

Additional Considerations
Earlier I used the term “permanently & completely exit”. Due to the fact that the 1st and 2nd levels of support may be very far away from each other in some instances, I may consider exiting the position partially, or exiting and then looking for an opportunity to re-enter the position, when it breaks the 1st level of support. The main reason I would do this is not to give back too much of my profits. Once two levels of support are broken, I want to be completely out of the position and will not consider re-entering without a whole new base setting up.

Since the “Double Support Trailing Stop” will keep you in positions longer, it is most effective when you are trading with the overall trend of the market. In other words, long positions in an uptrending market, and short positions in a downtrending market.

Can Contracting Range Hint At Direction?

After the negative inside day on Friday, the S&P refused to sell off hard and even posted a decent gain today – breaking its recent streak. Perhaps this may signal a change in character from the downtrend since the October highs. Traders may not be looking to leave the party every time there’s a pause in the conversation.

Speaking of pauses, have you noticed how the price range has been tightening? The range over the last 3 days has been the tightest of any three days since December. Even more interesting to me is that over the last 13 days, the high-low range the S&P 500 has traveled has barely exceeded the range it traveled in the one day prior to that. The SPY on January 23rd had a range of $7.35. Since then the total range has been $7.88.

Looking at the S&P 500 cash index I went back to see other times the market traded in a range nearly as tight relative to one bar over a 13-day span. The parameter I used was that the range of the last 13 days had to be within 115% of the range of the 14th day back. I found 31 other occurrences going back 30 years. Twenty-five of those occurrences saw the market higher a month later. Only three times that I found was the range contraction followed by a break to the downside which led to significantly more selling. Those occurrences were August 1985, September 1990 and August 1998.

This study seems to be another example of what I’ve been seeing lately. Over the next several weeks risk/reward appears to favor the upside. Even this contracted range has some serious volatility, though, so risk should not be underestimated. Good timing and proper (reduced) position sizing appear key.

On another note, I was pleased to have my recent post on leadership breadth appear in this week’s Festival of Stocks along with many interesting columns from other bloggers.

Sunday, February 10, 2008

Why Inside Days Get Me Down

Friday was an inside day. For me that meant laying in bed most of the day trying not to vomit. For the market that meant a lower high and a higher low than Thursday causing Friday’s range to be completely “inside” Thursday's range. The indices also closed lower on the day. Below is a chart of the SPY going back to October 8th, which was the day it closed at its highest level. Every inside day with a down close is marked with a blue dot.

The return following each of the inside days with a down close is as follows: -0.48%, -2.62%, -1.27%, -2.45%, -2.64%. That is an average loss of 1.9% the following day. Five for five losers. The pattern isn’t pretty.

Even before the recent market top, this pattern has had bearish tendencies. Looking back to the beginning of 2004, there have been 53 inside days with down closes for the SPY. Sixty-four percent of them were followed by a selloff the following day. The average loss was about 0.7% and the average gain 0.5%. Total losses outsized total gains by 2.3 to 1.

Looking back even further, since the 2000 market top, the SPY has closed down 58% of the time following an inside day with a down close. There were 104 occurrences. The average loss the next day was 0.9% while the average gain was 0.6%. In all total losses outsized total gains by about 2.1 to 1.

Friday’s pattern may have been an ideal one for me while trying to deal with the flu. For the market, the pattern has historically signaled short-term trouble.

Two other quick notes:

After conducting the study I did a search on “stock market inside day”. Dr. Steenbarger had an interesting study regarding them a couple of years ago.

In the comments section after my study of leadership breadth at market bottoms on Thursday night there was some discussion of the importance of looking at new low figures near market bottoms. Dr. Steenbarger also saved me some time this weekend and wrote a nice post on new lows near market bottoms. Thank you Dr.! (Now how about a little something for the nausea?)

Thursday, February 7, 2008

Is Leadership Breadth Important For A Successful Bottom?

One reason cited by IBD recently for their lack of confidence in the current bottom attempt is the lack of stocks with sound basing formations. While doing a historical study of the number of basing formations seems near impossible to me, one reader suggested looking at new highs. I thought this was a good idea since it should give a reasonable estimate of leadership breadth.

I looked at every Follow Through Day (FTD) identified in Part 1 of the IBD Follow Through Day Study and calculated the percent of New York Stock Exchange stocks that hit new 52 week highs on the day of the FTD. I broke the results down into “Successful” and “Unsuccessful” FTD’s.

This is what I found:

The average percentage of NYSE stocks making new highs on “successful” FTD’s – 2.1%
The average percentage of NYSE stocks making new highs on “unsuccessful” FTD’s – 2.0%

The median percentage of NYSE stocks making new highs on “successful” FTD’s – 1.2%
The median percentage of NYSE stocks making new highs on “unsuccessful” FTD’s – 1.4%

The minimum percentage of NYSE stocks making new highs on “successful” FTD’s – 0.1%
The minimum percentage of NYSE stocks making new highs on “unsuccessful” FTD’s – 0.1%

It appears that leadership breadth has no predictive value when assessing the likelihood that a FTD will succeed or fail.

On January 31st there were 18 new highs out of 3272 issues traded on the NYSE according to my database. This equates to 0.55% and has hereby been deemed a useless fact.

I found the results somewhat surprising as I thought leadership breadth would provide at least some advantage.

I still feel leadership is important to sustain a rally. It appears many times the real bull leaders may not emerge immediately. While the FTD typically comes 4-10 days after the bottom, leadership may take 3 weeks or more to establish itself.

So will this rally attempt succeed? I don’t know. I do know if it fails it won’t be because leadership breadth was too weak.

After all this recent testing of "conventional market wisdom" I’m starting to feel like Adam and Jamie from the Discovery Channel. Myth: Leadership Breadth Is Important At Market Bottoms – BUSTED! Think I could land my own tv show?

Good For A Day - But Back To No-Man's Land

The market gapped lower as I was hoping and then after much back and forth it finished on an up note. As I indicated last night, I was interested in buying a large gap down and have held on to some of the position for a swing trade. This one has worked out so far, but I’m not terribly excited with today’s action.

Over the last few days there have been some fairly extreme readings. Two examples were the Arms Index mentioned two nights ago and the sharp three day selloff followed by a sizable gap down this morning. The reversals generated from these conditions the last two days have not been inspiring. Today’s action was wildly up and down. Bulls have not been able to take control for more than a couple of hours at a time.

It has now been 5 days since the January 31st follow through day. Last Friday I did a study on the short-term implications of follow through days. It showed that success or failure of the new bully rally attempt is about 67% predictable based on the action of the first 5 days after the follow through day. That study therefore now predicts a continuation of the bear.

The good news is the market has managed to hold support so far. Readings around the January lows were severe enough that it remains plausible a significant bottom could have been made. I’ve obviously not yet given up on that idea. The market is in no-man’s land right now. Tests run tonight based on current action are showing no significant edge. Technically, a move above 1400 would signal the bulls are taking charge while a move below 1270 would signify the bears are still in control. That range is very wide but with the recent and expected volatility, I’d be surprised if it wasn’t broken fairly soon. I’m holding a small long position leftover from today but not looking to aggressively add to it.

Retest approaching

Just time for a quick note tonight…

The study I posted last night showed a good possibility of a one-day rally in the S&P. The one day rally lasted only half a day. After that the market fell apart again.

I’ve gone over a few things tonight, but I’m seeing pretty much what I expected. Price is becoming stretched, but the price drop alone isn’t providing any huge edge. An example would be the current Nasdaq price action. The Nasdaq has dropped over 1.25% three days in a row. Looking back to 1985 this has happened 48 other times. Over the next 1-2 days the Nasdaq has rebounded about 2/3 of the time. Not a bad ratio, but the average losing trade dropped about 4%. That’s more downside risk than I’d like to take on.

Many of the oscillators and other market gauges I follow are simply not stretched the way they were two weeks ago. My Capitulative Breadth Indicator (CBI) -click CBI label for details- for instance still remains at 0. Once we get closer to retest area a case could be made for an entry on some kind of reversal since a reasonable stop should be nearby. We aren’t there yet and I’m not comfortable looking for even a swing trade at this point.

One thing that may interest me is a gap down tomorrow morning. Cisco (CSCO) is trying to help create that. I showed recently that large gaps down in downtrending markets offer significantly more reward than large gaps down in uptrending markets. A gap lower therefore could be playable. I would only consider keeping part of the position beyond a day if it was strongly in my favor.

Tuesday, February 5, 2008

These Are Some Big Arms!

There was some serious selling pressure today. Selling swamped buying on both the NYSE and the NASDAQ. The Dow, S&P 500 and Nasdaq all finished down close to 3%. Volume rose across the board.

The Arms Index for both the NYSE and the NASDAQ was very big. (Although as you can plainly see from the picture on the right, my arms are bigger.) The NASDAQ Arms Index closed over 3 and the NYSE Arms Index closed at about 2.7. Many people view large spikes in the Arms Index as a contrary indicator. It reflects panic selling on the part of market participants. This panic is viewed as a short-term oversold condition that leads to a reflexive bounce.

I ran some tests to measure today’s action against history. All tests were run back to 1/1/92. I found that a 1-day spike in the Arms Index of either the NYSE or Nasdaq without confirmation from the other showed very little edge.

A Nasdaq Arms Index greater than 3 has occurred 79 times since 1992. The next day the Nasdaq finished higher 57% of the time. The average winning trade was 1.7% but the average losing trade was 1.9%. The total expected value was a scant 0.2% rise the next day in the Nasdaq.

The NYSE Arms Index has posted readings over 2.5 fifty-seven times. The next day the S&P 500 rose 63% of the time. The average winning trade was 1.2% and the average loser 0.8%. On average the S&P gained just over 0.4% the following day.

Neither of these stats really gets me juiced about jumping in front of this train. (Did you see the photo of me looking juiced?) When I look at the events on a combined basis, though, results improve dramatically. Since 1992, there have been 13 days where the Nasdaq Arms finished over 3 and the NYSE Arms finished over 2.5. The S&P 500 rose 11 of the 13 times the following day. The average gain in the S&P 500 was 1.6% and the average loss was 0.15%. The average trade the next day was good for 1.3% and the mean was 1.2%.

After one or two days the results fell more in line with random. Any edge here appears to be very short-term.

Looking out a bit longer, the recent bottom is now looking very much in question. The positive feelings the bounce brought about were nearly all washed away yesterday and today. As I’ve been pointing out, the action immediately after a Follow Through Day is normally telling. So far it doesn’t look good.

Lastly – some house cleaning. Tradestation apparently did a data update on their advance/decline data last night. The Appel Daily Breadth signal I’d mentioned a few nights ago is now no longer appearing. As you might imagine, this aggravates me to no end. I doubled checked using other databases and found a signal should not have been generated. (It came very close.) I’ve imported advance/decline and volume data from one of my other data sources and will be using that from now on. Tonight’s tests were run using QP3 data imported to Tradestation (and confirmed with a 3rd database). I apologize for the false information generated by bad data and have already taken steps to minimize the chance of it happening again.

I.B. Ranting

After persistently running higher for almost two weeks, the market finally began its pullback today. I am not a participant in the pullback, but rather an observer. While the overbought readings indicated there would likely be one soon, my studies indicated risk/reward was unfavorable in trying to short it.

As far as I’m concerned the action during this pullback becomes very important. I demonstrated last week that the first several days following a Follow Through Day are a pretty good predictor of success or failure. After 2 days we are still holding up. I’ll continue to monitor the action carefully.

Warning…rant coming…

One issue that has been raised is whether last Thursday was in fact a Follow Through Day. IBD apparently failed to label it one. In Friday’s Big Picture column they wrote:

“Given the market's volatility since late December, you'd need to see bigger gains than Thursday's to signal a fundamental shift in the market's trend.
As noted in Thursday's Big Picture, it's almost a moot point even if the market did manage to assemble a follow-through session of powerful gains in heavier volume. The reason? There are virtually no stocks close to proper buying positions right now.”

From my perspective, Thursday satisfied the requirements IBD previously laid out for Follow Through Days. Due to their perception of volatility and breadth they remain negative on the market's prospects. As they should, they are deriving their outlook from multiple readings of market health. Unfortunately it seems they would like to advertise the Follow Through Day as a magic indicator that never fails. Therefore, either they make excuses or change the definition of it so that it appears not to fail. Should this rally take hold I have no doubt they will refer to last Thursday as a Follow Through Day in a future publication. It is a tool that uses volume and price action. Breadth, while important, is a separate matter.

Whether Follow Through Days always work is not important to traders. What is important is whether they can be utilized as an effective tool for helping to identify market bottoms. So far we’ve seen that they have been about 55% accurate and risk outweighs reward. (part 1 and part 3) I’d say they have some utility. Constant refinement of the definition to retrofit recent market conditions destroys much of that utility in my eyes.

The McClellan Oscillator is a terrific tool for measuring the market’s health. Should the McClellans decide to constantly adjust the way it should be calculated, it would fail to be as useful.

The Follow Through Day study I laid out and have been discussing the last few weeks has identified Thursday as a Follow Through Day based on the original definition offered by William O’Neil. For purposes of my study and my trading it shall remain one whether it works or not.

…end of rant.

Monday, February 4, 2008

Breadth Follow Up & CBI and Bounce Candidates Closure

Lowry's Breadth
First, a quick follow up to my post on breadth yesterday. It was pointed out to me that Thursday and Friday may have qualified as back to back 80% Days. In Lowry’s report, back to back 80% days can substitute for one 90% day at some bottoms. I said “may” qualify, because according to my data (which could be wrong)**, Thursday was a 79.61% upside volume day. This could qualify as 80% by human standards, but perhaps not by computer standards. Anyway, certainly seems close enough to quell some of the concerns I’d previously expressed about the lack of a 90% day…

Bounce Candidates Closure
Back on the night of the 21st as the futures were down about 5% I wrote a post showing that market bounces from extremely oversold and capitulative conditions typically see the most beat up stocks bounce the best – and by a substantial amount. I looked at Dow stocks to illustrate this. I also listed the 5 worst performing and best performing Dow stocks from December 26th through January 18th. The time frame I looked at for the intial move off the lows in that study was 6-8 days. It has now been 8 trading days since the market bottomed, so I thought it might be interesting to see how things played out this time. Below are the stocks I listed in the previous column along with the results:

Once again the Lowest RS stocks trounced the Highest RS stocks by more than the Dow itself bounced. In fact 4 of the top 6 performing Dow stocks during the bounce were among the 5 “Lowest RS” stocks I listed.

Capitulative Breadth Indicator (CBI) Returns to Zero
For those that haven’t been paying close attention, the CBI closed at 2 on Thursday and 0 on Friday. While I suggested taking at least partial profits earlier, Thursday’s move below 4 would have signaled the exit to the standard CBI trade I’ve discussed in the past. Below is a graph showing the action in the CBI going back to October. Buying the close on the 22nd and holding until Thursday’s close would have been worth about 5% in the S&P 500. Of course with some stock selection such as described in the Dow example above, traders may have done significantly better.

Incidentally, a zero CBI is not a signal to short. The CBI only measures "oversold" - not "overbought".
**9:55 am - Edit - I've now been told my data is in conflict with others who show 1/31 to be about 82% upside volume.

Sunday, February 3, 2008

Whose Breadth Stinks?

Breadth indicators seem to be at odds here. Is it bad, good, or so good it’s bad? Here are some notes to help confuse you:

Leadership Breadth
Investors Business Daily in their Big Picture column on Friday noted that “There are virtually no stocks close to proper buying positions right now.” This is a measure of breadth and potential leadership. The basic idea is that the market will struggle to rally strongly without leadership.

90% Days and Herding
A few weeks ago it was pointed out to me that from July 2007 until now there has been a higher concentration of 90% volume days on the NYSE than at any time since the 1940’s. (A “90% volume day” is a day where either up volume or down volume swamps the other by at least 9 to 1.) Lowry’s created the 90% day concept and has used them to identify market bottoms in the past. Detailed information can be found in their recently updated 2001 study. So far while this market rally has produced some strong up days with both price and volume it has yet to produce a 90% upside volume day. This would seem especially concerning to me given the fact that they have become more frequent recently.

Brett Steenbarger has also done some great work with the concept of breadth volume. He refers to it as “herding”. Here is an introductory discussion of it from his blog. For more discussion you could simply search on “herding” in his search box on the top left hand corner of his page. I'm curious to see if Dr. Steenbarger has anything to say about this in the near future.

A Rare Breadth Thrust Signal
In Gerald Appel's book "Technical Analysis - Power Tools for Active Investors" he publishes a system called the "Daily Breadth Impulse Signal" (chapter 6, page 142). Essentially it looks to enter the market when the 10-day exponential average of advancers/decliners hits an extreme. The trade closes when the same measure eventually declines to a certain number. The published results were fairly impressive. Since 1970, there were 26 trades with 22 winners and four losers. Avg winner: 5.24%. Avg loser: 2.37%.I recreated his indicator some time ago. Using my data, results were slightly different, but similar.The last signal given before publication was 05/2004.

According to my data, in November of 2004 another signal triggered. This one lasted for close to a month and was good for about a 1.3% gain.

March 21, 2007 was the next trigger I could find. Unfortunately, the signal was cancelled 5 days later on a shakeout, causing a whipsaw loss for the system as it was unable to benefit from the uptrend that followed.

According to my data, Friday 2/1/2008 gave the first buy signal since 3/21/07.

The McClellan Oscillator
A few people have pointed out to me the overbought nature of the McClellan Oscillator. I’ve yet to complete my study of the current situation in regards to the oscillator, but I did find the following quote from a 2004 paper by Tom McClellan very interesting:

“First of all, deeply negative readings tend to indicate conclusion of a down move, whereas extremely high readings tend to show initiation of a new up move…The postings above +200 in September 1998 were a sign that the bulls were going to be coming rushing back in, and they had a lot of money in their pockets to push prices higher for a sustained period of time…So when one sees a very high reading, it may be a sign that a brief pullback is needed, but it is also a sign that higher prices should be expected following that pullback.”

The entire paper may be found here. The quote was taken from pages 24 and 25.

Summary thoughts
Breadth has been extremely strong, but we’ve yet to see a real rush that would look like a “herding” day. The strongest areas in the last week and a half have been those that were beaten down the hardest. Contrary to what some would have you believe, this is NOT unusual at market bottoms. (I’ll review results from my previous post on this later.) I’m continuing to take it day by day. Since exiting my long positions based off my bottom indicators I’ve yet to reallocate heavily either long or short and am waiting for a better edge.

Friday, February 1, 2008

Short-term Implications of Follow Through Days

Time for CANSLIM traders to go back to work. An IBD Follow Through Day triggered today. At Quantifiable Edges I’ve discussed the intermediate-term implications of Follow Through Days in great detail. Catch the entire series so far below:

While intermediate-term traders are rejoicing with today’s Follow Through Day, swing traders are noting how overbought the market has become on a short-term basis. Most of what I’ve read by short-term traders this afternoon and evening has been short-oriented. The prevailing theme is that we are now short-term overbought in a longer-term downtrend (or bear for some). Most people believe these are ideal conditions for shorting.

What swing traders looking to short need to understand is that nearly every Follow Through Day produces short-term overbought conditions in a downtrending market. They frequently arrive within 4-7 days of a bottom. Today was day 7 of the rally. As it was last week, most of the time the reversal off that bottom is violent. This can cause oscillators to become overbought. When the formula calls for a typically violent reversal, a week of gains, and a strong rally on high volume to cap it off – you’re bound to be overbought short-term. Does that mean it’s a good time to try a swing trade short?

To test it I looked at the 1-5 day returns of all 64 Follow Through Days listed in my study. $100,000 per trade. Go long on the close of the Follow Through Day. Exit X days later. Results below:

More often than not the market trades higher over the next 1-5 days. The average win is larger than the average loss. Profits continue to be made on the long-side. Shorting is a losing game in this scenario. Perhaps this disbelief by short-term traders is what helps to continue to fuel the rally as they are constantly forced to cover their losing positions.

Is short-term success or failure indicative of long-term success or failure?
One interesting claim that IBD sometimes makes about Follow Through Days is that those that fail normally do so shortly after the Follow Through Day. I decided to also look at this concept tonight.

I broke the 64 Follow Through Days in my study up into two groups – the successful ones and the unsuccessful ones – to see if their early performance hinted at their chance of longer-term success. Below are the breakdowns – same as above - $100,000/trade, long at the close of the Follow Through Day, and exit X days later.

Early action after Follow Through Days that eventually “succeeded”:

Right off the bat most of these posted nice gains. The short-term winners among the group averaged another 2-3% upside in the first week. The short-term losers suffered 1-1.5% drops on average. Net profits were substantial.

Early action after Follow Through Days that eventually “failed”:

Those that eventually failed tended to show signs of failure right away. Notable here is the average loss was appreciably higher than the average gain.

A basic rule of thumb is that the success or failure was determined with about 67% reliability within the first week after a Follow Through Day. For instance, note there were 41 total trades that were in the black after 5 days. About 2/3 of them went on to “successful” rallies. The same ratio applies for the losers. There were 23 losers after 5 days. 15 of them ended up with “failed” rallies and the other 8 were “successful”. The 2/3 rule holds fairly accurate whether you are looking at winners or losers over any period from 1-5 days after the Follow Through Day. The Unemployment Report tomorrow morning has a chance to set the tone early on in this one.

To summarize the two main points tonight:
1) Don’t be too eager to short. It’s doesn’t have positive expected value just after a Follow Through Day.
2) Watch market action closely over the next week. It should give you a pretty good indication of the intermediate-term.