http://www.sentimentrader.com

Thursday, August 7, 2003  9:15 PM EST

For the twelfth week in a row, the bull ratio in the Investor’s Intelligence sentiment survey has been above 70%.  The bull ratio is calculated by dividing the number of bulls by the number of bulls plus the number of bears, so it gives a truer reflection of bullish opinion among those who have an outlook one way or the other.  This is just a bit misleading, because those who are neutral are essentially bullish and are just expecting a near-term correction, but it’s usually safe to exclude the “neutral” camp from these calculations. 

This 12-week streak over 70% is the eighth-longest such streak in the survey’s history (the all-time record is 26 weeks, tied by the periods ending in October 1976 and July 1983).  Since 1987, we hadn’t even seen a streak of more than 3 weeks in a row with such a high bull ratio.  The average reading during the past three months has been 75%, which once again is the first time since 1987 that we have seen a three-month (12 week) average at 75% or above.  To put it another way, since mid-May an average of 6 out of every 10 newsletter writers participating in the I.I. survey have been bullish on the stock market, 2 have been bearish and 2 have been neutral.  An average of three bulls for every bear over a three month period is indicative of what I would think is excessive optimism. 

I took a look at every instance in the history of the survey where bulls outnumbered bears 3-to-1 over a three month period (i.e. a 12-week average bull ratio of 75% or more).  There were 8 distinct occurrences in the past 33 years.  Once the 12-week bull ratio initially reached 75% or greater, it stayed above 75% for an average of 17 weeks.  However, this is greatly skewed by a 66-week streak beginning in January 1976, so if we take out that outlier, the average consecutive number of weeks above 75% drops down to 10. 

So how did the market do in the face of so much sustained optimism?  In turns out pretty well, or at least better than one would suspect.  After the 12-week bull ratio first reached 75% or greater, the S&P 500 didn’t suffer a meaningful decline (let’s say of 8% or greater) for an average of 46 days, or over two full trading months.  The shortest lead time was 5 days and the longest was 94.  Obviously, then, we wouldn’t want to sell the market in general as soon as the three-month bull ratio reached 75%.  How about if we wait until the bull ratio peaks, and then begins to come back down?  That’s better, as the average lead time before the market itself peaks drops down to 7 days.  However, the downside to that is that the market peaked before the bull ratio did four out of the eight times, by an average of 25 days (although once again the figure is skewed by one instance of 79 days – without that, the average drops to 7 days).  Also, once again there was one very long lead time of 89 days of the market peaking after the bull ratio did. 

Interestingly, the market finally peaked in April four out of the eight times, July twice, and June and January each once.  I don’t think that necessarily means anything, but statistically it would be quite rare to see one month get 50% of the market peaks just by chance. 

The statistics bear out the assumption that if we wait until the bull ratio peaks, and optimism begins to subside, then we would do much better by selling the market at that point than we would by selling immediately when the 12-week average reaches 75% or more.  There was an average of 39 days between the ratio first reaching 75% and it finally peaking.  Taking out an outlier of 141 days in 1976, the average drops to 25 days.  

This is usually easier to see by looking at the charts and tables themselves, so here they are.  First, this chart shows how the S&P 500 performed in the months after the 12-week average bull ratio first reached 75%: 

We can see that the average return up to 6 months later was still positive, and the market was higher more than 60% of the time across every time frame. 

Now let’s look at the other scenario, where we wait until the 12-week average bull ratio peaks and begins to decline.  From the week where the average finally peaked, here is how the S&P 500 performed: 

Statistically, it looks like it would make more sense to wait until the ratio finally peaked before selling.  If one would have done that, historically the market performed meagerly afterward, with the average return below zero even up to six months later, and the market was negative more than 60% of the time 1, 3 and 4 months later. 

For reference, the tables below go into much more detail about how the S&P 500 fared after the two scenarios mentioned above – first, the 12-week bull ratio initially reaching 75% or higher; second, the bull ratio finally peaking and beginning to fall back. 

In order to clarify what the tables mean, let’s go over the first month after the 12-week bull ratio makes it to 75% or above: 

The “R” column shows the average return of the S&P 21 days after the bull ratio first hit 75%, which we can see here was 1.7%.  The market was higher 75% of the time, which we see in the “POS” row.  The maximum return out of the eight occurrences on a closing basis was 5.5%, and the greatest loss was 4.5%.  The S&P gained a maximum of 8.0% during the initial month (from the “+” column), and lost a maximum of 5.4% (from the “(-)” column).  The average amount the S&P rallied between the time the bull ratio first hit 75% and 21 days later was 4.2%, and it declined an average of 1.8% during that time.  You can see that this would have proved to have been a better buy signal than sell signal. 

Here are the stats for how the S&P 500 performed for the first six months after the 12-week bull ratio first reached 75% or higher: 

FIRST WEEK ABOVE 75%...FIRST THREE MONTHS

FIRST WEEK ABOVE 75%...LAST THREE MONTHS

Now, here are the stats for how the S&P 500 performed for the first six months after the 12-week bull ratio finally formed a peak and the optimism began wearing off: 

AFTER THE BULL RATIO FINALLY PEAKS…FIRST THREE MONTHS

AFTER THE BULL RATO FINALLY PEAKS…LAST THREE MONTHS

For the “what’s the bottom line” personalities among you, the takeaway from this is that three months of unadulterated optimism among the newsletter writers surveyed by Investor’s Intelligence has not in and of itself been a good sell signal.  A better sell signal occurred an average of about two months later, when the bull ratio finally peaked and optimism began to decline (often due to declining market prices, no surprise).  Obviously, each moment in the market is unique, and we only have 8 distinct occurrences from which to gather conclusions, which itself makes them tenuous.  Also, none of the other 8 occurred in a post-bubble environment, which certainly diminishes their applicability.  Still, I think it’s important to keep in mind that historically, this type of bullishness by this particular population was not in and of itself a good sell signal. 

I mentioned this weekend that Rydex traders had not yet shown any undue pessimism in their choice of asset in which to place their money.  Beginning this week, especially yesterday, that changed.  For the last few days, these traders have become considerably more aggressive in their downside bets, which from a contrarian perspective is bullish for the market, as these traders are usually wrong when they get confident.  Our shortest-term Rydex indicator, the RSI Spread, is oversold (though not terribly so) at -70.  Even some of the longer-term indicators are beginning to show pessimism, as the 3-month Rydex Stochastic registered the first “zero” reading since February.  The chart on the site is a 5-day moving average, so you won’t see it hitting zero now, though it has dropped significantly in the past few days, and is down to its lowest level since March.  The Beta Chase Index, which measures the relative asset flow between the high-beta “run and gun” funds and the low-beta “defensive” funds has just hit the lowest level since June 25^th, showing that these traders are pulling in their horns to a relatively large degree.  Some of this pessimism will likely be alleviated after tonight’s numbers are reported, but usually these types of readings will give us at least a two- to three-day respite as the market rallies. 

Many of our shorter-term indicators are now at a similar place to where they were in late March, mid-May and late June – the three retracements since the March low that quickly lead to higher prices.  This time is different due to the fact that the momentum has been sucked from the market, so while higher prices are certainly possible, even probable in the short-term, a rally here should just serve to set up longer-term shorting opportunities.

- Jason Goepfert

Disclosure: long OEX puts

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, August 5, 2003  10:39 PM EST

Over the past five days, the OEX put/call ratio has been quite low, and it has been under 1.0 for nine out of the past ten days.  Going out further, the 21-day average is now at 0.84, which is the lowest reading in 15 years.  You may know that I typically view this indicator in a non-contrarian manner, meaning low OEX put/call ratios tend to be bullish for the market while high ratios tend to be bearish.  OEX traders have a deserved reputation for being some of the better market timers around. 

While this low p/c ratio should be bullish, there is something about it that isn’t quite right, as I pointed out on July 24^th.  That “something” is the open interest configuration of these options.  The relatively heavy OEX call volume we’re seeing is simply not going into new contracts.  Instead, it has been used to close existing contracts, for the most part, on a greater pace than what puts are seeing.  Over the past 10 days, call open interest has increased by 44,429 contracts.  Put open interest has increased by 39,704 contracts.  The logical question from that would be “Call open interest has increased by nearly 5,000 more contracts than put open interest, isn’t that bullish?”.  Well, during that time there have been 297,000 calls traded and 222,000 puts.  So despite there being 75,000 more call contracts than put contracts traded, call open interest has increased by only 5,000 more contracts. 

A couple of weeks ago, I showed a chart that presented this information graphically.  At the time, OEX traders were opening more puts than calls, relative to the volume each was seeing.  Several of you asked for some longer-term evidence that this can be an effective way to look at things, so tonight I’ve included a 42-day average of this information (approximately two trading months). 

The chart below shows this two-month average of how much put volume is actually going into new contracts as opposed to call volume.  When the indicator is above 0.00, then more put volume than call volume is going into new contracts.  It can be thought of as a kind of “efficiency” of trading, showing how much determination OEX traders have to really bet on a market move.   

Since I will likely be referring to this in the future, let’s call it the OEX Determination Index for short. 

We can see quite clearly that usually when 10% or so more put volume than call volume went into new contracts over a sustained period (i.e. the blue line in the chart above approached the red horizontal line), the market had trouble making headway.  Conversely, when more call volume than put volume went into new contracts, especially when it became extreme (i.e. the blue line approached minus 10%, the green horizontal line), the environment was positive for stocks. 

Over the past week, even though the OEX put/call ratio has been low, and the open intere