http://www.sentimentrader.com

Sunday, October 26, 2003

SPECULATIVE ACTIVITY

Bottom Line - Bearish

We’ve seen a prolonged rush into the most speculative stocks, which has resulted in negative future market performance in the past when it has reached this degree. 

One of the hallmarks of the recent rally has been the speculative nature of many of the leaders.  This is not an original insight, for sure, as it has been discussed ad nauseam for months on end now.  The part of this that could lead to something useful, however, is trying to determine if the speculation is excessive or not, which is the crux of sentiment analysis in the first place. 

One widely-available, though not commonly-known, index that tracks speculative investing is the NYSE Beta Index (ticker symbol NHB, also known as the High Beta Index).  It is difficult to find information on this index, but it was created to track the 100 stocks on the NYSE with the highest beta, or correlation to broad market moves.  The “broader market” can be defined as the NYSE Composite Index, or S&P 500, or any other index that is representative of most of the market capitalization, and by definition it has a beta of 1.0.  A stock with a beta of 2.0, for instance, should increase by 20% for every 10% rise in the broader market, on average.  A stock with a beta of 0.5, on the other hand, would typically only increase about 5%.  A stock with a beta of -1.0 would decline by 10% for every rise of 10% in the broader market.  I’ve discussed the concept of beta several times in the past, as it is an important part of several of our Rydex indicators.

When the index was first created, its “charter” was that it would only include stocks trading above $10 per share, and have a float of at least 7 million shares.  The index would be re-balanced every January and July so that the highest-beta issues continue to make up the 100 stocks.  Frankly, I have no idea if this is still being followed, but my assumption is that it is.  I ran a scan to re-create the makeup of the index myself, but it undoubtedly deviates somewhat from the stocks that actually make up the index right now.  In looking at the stocks that fit the criteria of the index, it is clear that many of them are extremely speculative.  Over 65% of the equities do not pay a dividend, and 37% showed negative earnings over the past 12 months.  The average beta of the 100 stocks was 2.4.

The reason I bring this up is because the Beta Index continues to outperform the broader market, showing that investors have no qualms about continuing to pile into the most speculative of issues.  The chart below shows the performance of the S&P 500 versus the Beta Index, with March 11^th, 2003 being the starting point.  The Beta Index has outperformed the broader market from the get-go, and the spread continues to widen.  As of Friday, the S&P was ahead by 28% while the Beta Index has enjoyed a 55% spurt.  The action has been particularly striking over the past couple of weeks. 

Is this excessive?  Indeed it is, and the ramifications are interesting.  I determined this by looking at how stretched the two indexes were from their respective long-term moving averages (252 trading days - equivalent to one year).  For example, as of Friday the S&P 500 is 10% above its 252-day average.  The Beta Index, on the other hand, is 25% above its yearly average.  This gives us a “speculative spread” between the two of 15%, meaning the Beta Index is stretched 15% more than is the S&P 500. 

Historically, this is quite extreme as the following chart shows.  The chart plots this “speculative spread” between the two indexes, as calculated above, versus the S&P 500 since 1983.  As I said, the current spread is at 15%, which as you can see from the chart below is one of the highest readings in 20 years. 

There are many telling conclusions that come of this.  First, we can very clearly see that when investors flee beta and migrate to safety, it has corresponded with excellent buying opportunities without fail once they begin to dip their toes back in the water.  The most obvious example of this is 1998, when the Beta Index lost 34% of its value over six months while the S&P lost “only” 13%.  Once investors began taking on more risk after the panic, the market shot higher.

If we look at how the S&P 500 performed after each of the extremes, a clear bias emerges.  The table below shows how the S&P performed within various intervals following periods of extreme risk-aversion, defined as any day where the speculative spread was at -10% or below.  264 days out of the 5,232 days studied qualified as showing extreme risk-aversion. 

The table shows the average and maximum gain the S&P put in within the given number of days, as well as the average and largest loss.  For example, within 90 days of a period of extreme risk-aversion, the largest gain the S&P 500 enjoyed was 33%.  The largest loss (i.e. maximum drawdown if one was in a long trade) out of any period was 13%.  If we average all of the maximum gains out of all 264 days, we come up with an average maximum gain of 14% as opposed to an average maximum loss of only 5%.  Both measures show that the upside reward nearly tripled the downside risk.  Perhaps most telling, 155 out of the 264 days showed maximum gains of 10% or more within those 90 days.  The number of days showing losses of 10% or more?  11. 

Now let’s look at the other side of the coin, which is where we are today.  This table shows the S&P 500 performance after periods of extreme risk-taking, defined as a speculative spread between the NYSE Beta Index and the S&P 500 of +10% or more.  488 days out of the 5,232 studied fit the definition of showing extreme risk-taking. 

While not a perfect opposite of the table above, we can still see a clear bias here, only in the other direction.  Within 90 days, we can see that the maximum gain seen was 21%.  This is actually an outlier (meaning a very unusual occurrence), as there were no other gains greater than 14%.  It occurred in July 2002, as the speculative spread was still above +10% as the market bottomed in late July last year.  The maximum loss during this time was a whopping 32%, which is not an outlier.  There were 103 additional days that showed a loss greater than 20%.  The average maximum gain during this time was 4%, as opposed to an average maximum loss of 10%.  I show that 70 days out of the 488 exhibited a maximum gain of 10% or more within 90 days.  In contrast, nearly three times as many days, 192, showed a maximum loss of 10% or more. 

It is important to note that these results were obtained during the largest bull market in history.  If we are headed into a prolonged period of wide swings such as was seen in the 1930/1940 and 1960/1970 periods, then this type of oscillator should show an even clearer bias towards bouts of strong gains following risk-aversion and losses or stagnation following demonstrations of extreme risk-taking. 

The connotations here are negative for the market going forward, as we are currently in a period of extreme risk-taking.  The spread between the “stretch” in the NYSE Beta Index compared to that of the S&P 500 is at 15%, which has only been matched a couple of times in the past 20 years – the Spring of 1983, and May of 2002.  February of 1992 came close, but didn’t quite match the extreme we’re seeing now.  When looking at how the S&P 500 performed after other periods of extreme risk-taking, it is apparent that significant upside was limited, while significant downside was relatively common. 

AAII

Bottom Line – Bearish

I haven’t discussed the sentiment surveys much lately because they really haven’t changed much.  Everyone watches Investor’s Intelligence, so there’s nothing new to add to that.  Market Vane has consistently shown more than 55% bulls for two months now, which is at the very top of its range over the past four years.  The percentage of bulls in the Consensus survey, at least as of October 10^th, was 67%, which again is at the very top of the range since 1999.  

I’ve discussed the weaknesses with the methodology of the AAII survey before, but I still find it useful as a reflection of small-trader sentiment.  That survey has now averaged a bull ratio (bulls / (bulls + bears)) of 73% for the past 12 weeks.  This type of sustained bullishness from this general group of traders was unprecedented throughout the entirety of the great bull run of the 1990’s, and has only been (slightly) surpassed twice.  Those times were January 2000 and September 2000.  There were a few other times, such as January 1993, August 1997 and April 1998 where the bullishness came close to what we’re seeing now, but couldn’t quite match it.  When we look at what happened after January and September 2000, then, well, you know the rest. 

CONCLUSION 

We got the obligatory bounce off 1020 on the S&P that I mentioned on Thursday, and so far this correction appears to be no different than any of the others we’ve seen since March.  From the looks of price action alone, we seem to be in fine condition and it’s difficult to try to bet against that.   

On the negative side, we’re seeing several examples of extreme risk-taking.  That in itself is not necessarily a bad thing, but it has reached extremes that portend future weakness with a high degree of accuracy.  The NYSE Beta Index is one, as outlined above.  Another is the continuing speculation of small traders, as observed in the ROBO put/call ratio^TM.  That ratio stayed at 0.40 again this week, and in fact the smallest of traders (trading 10 contracts or less) bought over 840,000 call contracts to open, which is the greatest amount in over two and a half years.  There was certainly no great rush for put protection, as put buying accounted for only 13% of total opening transactions, which is one of the smallest percentages we’ve seen.  And every time we approach the highs, Rydex traders jump on board and the assets reach bullish extremes.  That pattern has been consistent over the past few months. 

On the plus side, we’re now pretty much through the weakest time of the month, and the market will have the wind of positive seasonality at its back beginning early next week and continuing basically right through the end of the year and into the beginning of next year.  The last few days of October and first few days of November tend to be positive more than 60% of the time on average, and that can be a powerful force.  We also have all of the other studies I’ve shown over the past month showing that the type of overbought readings we’ve seen, and the kind of price persistency the S&P has shown (such as the amount of time it has spent 5% above its 200-day moving average), have historically lead to positive markets over the next few months.  Also, going into next week, the severe overbought conditions that I mentioned last week are gone, and in fact the case could be made that we are even mildly oversold. 

Each time we reach some sort of short-term extreme, the market has backed off, both on the upside and the downside.  With the crosscurrents we’re seeing now, I don’t see any reason to believe that will not continue.  As the market becomes short-term overbought and approaches the upper end of its range (perhaps even breaking to marginal to new highs), I believe it will be difficult to maintain that and we will fall right back.  Should we fall enough to muster some oversold readings, I will be looking long.   

For next week, the short-term signs look positive.  As I said, we have positive seasonality coming on board the closer we get to the end of the month, and we are mildly oversold.  We are still in a longer-term uptrend, and Friday’s reversal certainly could continue to bring in buying if we are able to retake the stair-stepped resistance points created over the past two weeks.  So, I am more likely to try the long side now than I was the past couple of weeks.

 - 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.