http://www.sentimentrader.com

Thursday, March 20, 2003  9:00 PM EST

With today's up market, we've now closed higher 7 days in a row on the S&P.  This is an extremely unusual occurrence, as we can see from the table below:

For the purposes of this study, the Dow Jones Industrial Average is used as a proxy for the "market" up until January 1950.  From 1950 on, the S&P 500 is used.

Since 1900 (over 28,000 trading days), the market has been up 7 days in a row 174 distinct times.  We can see that it has gone up the next day about half of the time, as there were 84 instances of 8 up days in a row.

These types of streaks have become considerably more rare in recent history, as there have been only eleven streaks of 7 consecutive up days since 1990.  This is the least number of 7-day streaks in any given decade in history.  Even the 1930's had thirteen, and the 1970's had sixteen.  Perhaps with the increasing use of program trading, arbitrage strategies and the increasing failure of trend-following strategies in the equities market, it is a trend that will continue.

In any event, how can we use this information?  Let's take a look at how the market performed AFTER each of these streaks.  The charts below show the average return for several different time frames after each streak.  It also shows the percentage of time the market was positive the given number of days later.  I'll go over some examples below to clear up any confusion about what the charts may show.

We can see that after seven consecutive up days, the market showed an average return 5 days later of just under 0.50%, with a 67% probability of being higher.  The market exhibited an average return and probability of being positive that was greater than random for every time period up to 60 days after the 7-day streak, except for a slightly less than random chance of being higher the next day.  That same thing can generally be said after each of the other streaks as well.  If we look at another example, say after 11 consecutive up days, we can see that 60 days after such an event, the market showed an average return of nearly 5%, with a greater than 80% chance of being higher.  Of course, the greater the number of days in the streak, the less occurrences there are - meaning the sample size gets smaller and our confidence in the results should be tempered accordingly.  That's the reason you see the wild swings in the 12-day chart.

Let's take a look at a more recent time period, say just the past decade:

There have been nine streaks of 7 consecutive up days since 1993, but only two of 8 or more.  This is why you see such a low probability of the market being higher in the short-term.  However, and this is something that has held true over the past century, these streaks have typically NOT signaled the top of a speculative extreme.  In fact, just the opposite could be said - they are a sign that a rising market is more likely than a declining one over the next few months.  When we couple this information with the TRIN thrust information presented on Tuesday, it suggests that the low seen last Wednesday could be significant.  As much as I regret missing the move in our model portfolios (and most of it in my personal one), I think it's important to not develop a revenge mentality and look for chances to "get even" at the market by HOPING it goes down.  Objectively, I'm finding it more and more difficult to find reasons why last Wednesday could not prove to be even more significant than the 10% move we've already seen.

About the only issue I can find with this rally is that it is being widely accepted by marginal traders, including those who trade equity options.  You will find a new indicator posted to the site tonight, which is a de-trend equity put/call ratio.  You will recall that the equity ratio has shown a secular rise since the bear market began.  In fact, the average equity p/c ratio since 2000 is 38% higher than the average during the late 1990's.  This presents a problem when trying to analyze what constitutes an extreme.  For example, a reading over 1.0 used to be significant, now it almost seems like the normal course of business.  In an attempt to eliminate this problem, I have begun posting the de-trended ratio, which simply looks at the difference between the 10-day average and the 26-week average.  This puts the most recent two-week period into the context of the last six months, and is much more conducive to our normal way of looking at things in standard deviations from a mean value.  Since this indicator regularly gyrates around a mean value close to zero, when the indicator is stretched more than a standard deviation above or below zero, it means more than what the non-de-trended ratio would mean.  In essence, this indicator should be consistently more effective at calling either extreme than the regular ratio.  We are now approaching the lower band on this indicator, which means that the 10-day ratio is becoming stretched under the six-month average.  This suggests that call volume has been relatively heavy compared to put volume.  Since most of these options are bought to open, and calls that are bought to open are normally part of a strategy that relies on a rising market, then this means that equity options traders are becoming quite optimistic that this rally will continue.  These traders are consistently more wrong than right, and this is a troubling sign for the short-term.

As I said on Tuesday, the Rydex mutual fund timers are also embracing the rally.  That has continued, as our short-term RSI spread has hit a new all-time record (since 2000) at +100.  This means that the Relative Strength Index of the bull ratio is 100, while the RSI of the bear ratio is 0.  This has never happened in the 3-year history of this indicator, as the previous high spread was around +90.  This tells us that there has been a vicious shift in assets (and psychology) among these traders, and they are scrambling like mad to get positioned for a further rally.  As I said on Tuesday, this type of mad dash to the long side is almost never handsomely rewarded, particularly in the context of a longer-term downtrend. 

This meshes well with the Rydex Beta Chase Index, which measures the momentum flowing into high-beta Rydex funds and compares it to the momentum of the low-beta, or "safe" funds.  Currently, the Index is at 4.20, which is the most extreme since the mid-January market peak.  So not only has money been moving into funds that do well in a rising market, it has been moving into the funds that would do the BEST in that market environment.  This is speculation to the Nth degree, and once again it is a short-term bearish omen.

I mentioned earlier this week that our longer-term indicators in this fund complex were suggesting there was more room to rally, and that remains the case.  Although the extreme asset shift during the past few days has begun to have an impact on these slower-moving measures, it has not been nearly enough to bring them close to extreme territory.  This suggests that while some short-term weakness may be likely, there is more room to go on the upside before we would have seen the type of persistent long-side speculation that usually ends a bull move.

There is evidence that the public is jumping upon this move, which has to be considered a negative for a continued move higher.  This and the fact that the market rarely continues streaks of 7 consecutive up days should have short-term traders at least looking for short setups.  However, as I said in the intraday note, overbought or oversold does not mean much in this market right now.  We are chained to news headlines, and that means more than any sentiment reading, or technical level, or earnings report.  Everything is secondary to the war right now, so that must dictate the amount of risk you expose yourself to at any given moment.  You should realize that there is a very real possibility that you will wake up one of these mornings and see the futures limit up or limit down, and you have to be comfortable with the positions you have should that happen.  Barring any major developments (which is very unlikely, but I'll say it anyway), I think short-term traders can look for good setups in either direction tomorrow and have a decent chance for success.  For longer-term or long-only traders, I would view any significant declines (say to 840 or 850 on the S&P) as a chance to establish or add to long positions.  As I have been saying lately, however, if we trade below 830 all bets are off.

Disclosure:  no positions

This disclosure is not intended as trading advice in any form.  It is meant as a note to subscribers that the author may have a position directly affected by the market outlook reflected in the commentary.  Although the author takes great pains to remain objective in any commentaries, it is only fair that readers should know that the author may have taken positions in accordance with his market outlook.  Positions can and do change at any time, without notice to the reader.

Tuesday, March 18, 2003  8:40 PM EST

Yesterday's closing NYSE TRIN reading of .22 was tied for the third lowest in nearly 40 years (you can find this information yourself in just a few seconds in the Indicator Search & Sort function on the site by clicking on "Asc" next to Breadth, then click on the "TRIN" column heading in the table at the bottom of the page).  As I said in yesterday's intraday note, this is not necessarily a "bad" overbought omen.  In fact, I have found that the TRIN functions best as a contrarian indicator only when we see persistent readings, such as those reflected in the 10-day moving average. 

The chart below shows each occurrence of the TRIN closing under 0.30 from 1965 through today.  The dotted lines are each occurrence, and the thick solid black line is the average. 

This may be a bit confusing, so let's use a couple of examples.  The average return 3 days after a TRIN reading under 0.30 is 1.20%, and the market has a 69% chance of being higher.  After 10 days, the average return climbs to 1.77%, and once again the market has a 69% chance of closing higher than it did the day of the low TRIN reading.  We can see from the dotted lines, which plot each individual occurrence, that after 10 days there were several instances which lead to relatively large gains of greater than 4%.  Conversely, there were only two which lead to losses of approximately 4%.  This tells us that a single day's TRIN reading that would normally signal an overbought condition instead augurs a positive bias for the market, with a relatively high degree of consistency.  A TRIN under 0.30 suggests that there was triple the amount of volume going into advancing issues than declining, so it's a sign of strong buying power.  Until that reaches an extreme (i.e. a low 10-day average), it should be taken as a positive sign.

Although today's reading was not as extreme, it was still quite low at 0.50, which is in the bottom 4% of all readings in the past 40 years.  This has pushed the 10-day average of the TRIN down to a neutral 1.31 from an oversold 1.60 yesterday.  Going back to 1965, I checked all instances of the TRIN having consecutive closes below 0.50.  There were 17 distinct instances, and once again the conclusion is that it is a positive for the market going forward.  The following two charts show the market reaction after two consecutive days of the NYSE TRIN closing at 0.50 or below:

We can see that 10 days after such occurrences, the S&P 500 was up an average of 2.4%, and it had an 88% chance of being higher.  Interestingly, 11 out of the 17 instances took place in the 1980's, while the 1960's only had 1, and the 1970's and 1990's had two each.

Historically, these occurrences didn't signal anything spectacular over the coming months.  However, I want to touch on both the most recent occurrences, and the one that may most closely approximate our current situation.  Regarding the most recent instances, the first was 4/29/97 (with a TRIN of 0.47) and 4/30/97 (0.42), which kicked off a 19% run in the S&P over the next four months.  The second was 10/8/98 (with a TRIN of 0.49) and 10/9/98 (0.43), which lead to a whopping 43% gain over the ensuing nine months.  The counter argument to this of course is that we were in the midst of a roaring bull market, so of course there was heavy buying pressure.  So let's look at the one that may be closest to our current situation, in 1975.

Granted, the market was already up over 30% off its low, whereas now we're barely 13% higher, but this was the closest we can come.  We can see that after the low was reached in 1974, the S&P rallied strongly into 1975 before suffering a 10% decline into September.  It was at that time that the market closed two consecutive days with a TRIN below 0.50.  Currently, we declined about 17% from the high off the low in October before getting our two TRIN days.  Obviously, the recent retracement was much greater as a percentage of the rise off the low.  I could name a hundred different things that was different about 1975, so I don't think it's especially prudent to suggest it is a template for our current situation, but it's the closest we have as far as the TRIN readings go.

Today's equity put/call reading of 0.46 is the lowest we've seen for a couple of weeks, and it's relatively unusual to see call volume twice that of put volume.  A few months ago, I presented some information which showed that when the market is down going into option expiration, then the put/call ratio tends to be high, as traders gamble on cheap option "lottery tickets".  Since the trend is down going into expiration, they concentrate on puts, which obviously drives the put/call ratio higher.  Conversely, if the market is in an uptrend going into expiration, then the put/call ratio tends to be low, as traders switch their focus to cheap call options, hoping to strike it rich with a continuation of the trend.  So it's not really a surprise that today's quiet session prompted skewed put/call volume, but the troubling thing is that it's helping to drive the 21-day average of the ratio down to a level that could almost be considered overbought.  We're not there yet, but another couple of days of low readings will push the average into dangerous territory.

This newfound optimism on the part of options traders meshes well with the action seen in the Rydex mutual fund complex, as our short-term RSI spread has ballooned to +90, a new three-year record.  This tells us that there has been a massive shift of psychology from an aggressive bet on the short side, to a fear of missing "the big turn".  Since this is a short-term indicator, most effective in a time frame under 5 days or so, it does not suggest that this most recent move is necessarily over.  It only means that there has been a large amount of bets placed on the market moving higher, and the market rarely cooperates so nicely.  Most often, we will see at least some kind of retracement of recent gains after such a psychology shift, just to make sure everyone is uncomfortable.  Our longer-term measures  have not yet approached overbought, so as far as they are concerned, there is more room to rally over the coming weeks.

The volume and volatility we've seen the past week is certainly welcome, and I think necessary in order to be confident in a rally.  Most of our shortest-term measures relieved their overbought condition with today's choppy session, so the bulls have every excuse to take us higher and through the next layer of resistance.  As I have been saying, I am bullish as long as we remain above 830 on the S&P and are not overbought.  We ARE over 830, and have not yet reached overbought, so my inclination continues to be to look for retracements to establish long positions.

- Jason Goepfert

Disclosure:  no positions

This disclosure is not intended as trading advice in any form.  It is meant as a note to subscribers that the author may have a position directly affected by the market outlook reflected in the commentary.  Although the author takes great pains to remain objective in any commentaries, it is only fair that readers should know that the author may have taken positions in accordance with his market outlook.  Positions can and do change at any time, without notice to the reader.

© 2003 Sundial Capital Research, Inc.  All Rights Reserved.