2007: The Return of the Bear
By Frank Barbera CMT. January 9, 2007
Well, Happy New Year one and all, and what a complicated outlook we have headed into 2007. On the surface, the Year 2007 would seem to hold a distinctly bullish bias for the stock market, as positive (if not even outright robust). Pre-Election Years have been one of the great constants in the market for as long as most of us can remember. Since 1950, the S&P has had only one down year in a Pre-Election year, 1987, with a loss of 3.10%. Even more remarkably, over that period of time, 14 of the last 15 Pre-Election years have seen the S&P gain 10% or more; again, going back to 1950. Of those strongly positive years, an amazing 10 of 14 have seen the S&P gain 15% or more, with 5 of 14 gaining more than 20%. In fact,--and get this--the average gain for the S&P in the 14 positive Pre-Election Years since 1950 is a whopping 18.58%! With such a one sided statistical history, a ‘would be’ market ‘bear’ might as well find a nice cave to hibernate in over the next 12 months, right?
Ah, but not so fast. In the chart below, we show another “seasonal” pattern that has an interesting history and suggests 2007 could just as equally be a down year. The so-called “Dicennial” pattern makes reference to what has been a consistent 10 Year Cycle rhythm running through the stock market with at least some level of consistency for the balance of the last few decades. In the bar chart, we show the Dow percentage, gain or loss, for all years ending in ZERO---1900, 1910, 1920, 1930, 1940, 1950, 1960, 1970, 1980, 1990, 2000 over the last 100 years.
In bar number two from the left, we show the Dow percentage for all years ending in “ONE,” 1901, 1911, 1921, etc. From the preceding chart, we note that years ending in Zero, One, and Seven (columns 1, 2 and 8) have all had a negative tendency with year’s ending in “Seven,” by far the worst performers. For the S&P 500, this decennial cycle has yielded important lows approximately every 530 weeks, with major lows in the Asia Crisis of 1997-1998, the 1987 Stock Market Crash, the 1977 25% Bear Market, a soft year in 1967 and a selling panic in 1957. Will there be another decline, and another important low in 2007?
Of course, no one has the answer to such a question, and in all likelihood, there is no one seasonal or cyclical factor that can ever be counted on to mechanically deliver truly rock-solid market guidance. In fact, for both the Presidential Cycle and the Decennial Cycle on a purely statistical basis, 100 years or 1,200 months is a relatively small data sample, with the results of either cycle statistically debatable. However, one element of the market that is less debatable is the over-extended nature of the stock market and the fact that it is entering 2007 with a decidedly vulnerable technical posture.
On a major trend basis, I track several long term Momentum gauges which for the S&P have soared back up to the high end of their historical range. This tells me that the market is clearly no longer on the first floor of risk, but currently resides closer to the technical penthouse, where any downside reversal could represent quite the hard fall. In fact, rarely in stock market history has a market been this overbought, this over-extended, and displayed this type of uniformly one-sided sentiment — all hallmarks of an ultra high risk environment.
In our view, contrary to the overwhelming majority, I believe that 2007 is likely to be a Bear Market year for the S&P, Dow and NASDAQ, with all three averages retaining the potential to fall 20% or more. So why do I feel the way I do? Let's start by looking at a few “Time Spans” and see if this picture doesn’t start to uncloak the hidden wall of towering risks. A few years ago, I pioneered the use of Time Span Analysis which is a way of looking over one’s statistical shoulder to ask the question, "Is this environment statistically normal, or do unusual circumstances prevail?" In Time Span Analysis, we start by acknowledging the market's own internal “pendulum” principle, namely, that there are certain events that repeat year in, year out. The market becomes overbought, and then market becomes oversold; the grand market pendulum swings back and forth. Cycles turn up, and then they turn down. Anyone who has followed even one technical indicator knows there is rhythm to the movements of the crowd, and a sine wave affect at work within the market.
In Time Span Analysis, we want to focus on what we know to be regular, indisputable recurring events and ask “Has this happened recently?” By the time we are done with this analysis, we think you will be and should be a lot more concerned about your stock market investments, as the current outlook for the major averages is NOT promising. That said, let's get right to it. We are going to start with the S&P 500, where we have daily and weekly data back to 1950. Over that time, the S&P has regularly fluctuated back and forth between its upper and lower 50 week (essentially one year) Bollinger Bands.
Have a look at the chart above and just note the regular cycling action back and forth every few years between upper band and lower band, and then vise-versa. Now, let's have a look at the same chart through slightly different eyes.
In the chart above, we still see the S&P 500 with its 50 week trading bands on the top clip, but this time on the bottom, I put in place the Barbera Time Span Counter, which tabulates the number of weeks it has been since the S&P last tagged its lower 50 Week Bollinger Band. Since 1950, there have been 11 especially long dated cycles lasting 100 weeks (two years) or more. Of those 11, five lasted 150 weeks or more, with only two lasting 200 weeks or more. The delimiters of 100 weeks, 150 weeks and 200 weeks are shown by the horizontal lines on the bottom chart. Of the first ten long dated time spans, the average duration of these longest dated cycles was 163.60 market weeks. The longest ever was the 303 week Time Span between 12/16/94 and 10/06/00, the Great Technology Boom, which was a once in a generation affair. Yet, we are closing in on that Time Span once again, with the current Time Span ultra-mature at an amazing 222 weeks through 1/05/07.
Still feeling lucky about 2007? The chart below shows the current positioning of the 50 Week Trading Bands with an arrow pointing to the current value of the 50 Week Lower Band. Guess what? It closed last week at a reading of 1208.95 with the S&P 500 at 1412.85. That’s over 200 S&P points of potential downside risk IF this market decides it is ready to whip around and starting making for the lower band. In percentage terms, we would be looking at a least a 14% decline from current levels--hardly a positive market environment.
Alright, so you’re not convinced there is any downside risk in the stock market? After all, a daunting Bull could argue well, maybe the S&P will run out the string a little longer. It did 303 weeks one time, and we are “only” at 222 weeks right now. What if it got closer to the 300 week market? That could mean it stays up for nearly another 80 some odd weeks, right? Wrong. One must always allow for time spent moving down as more often than not, declines to go from Point A (the upper band) to Point B (the lower band) chew up considerable amounts of time, all of which is included in the ongoing cumulative tally which does not stop until the S&P closes below the lower band, on a Friday close. So on that level, allowing 40 to 50 weeks for the market to decline, the current advance would hypothetically already be even more mature at between 262 to 272 weeks. Still, a bull can argue that market could go higher and possibly blow off a bit more, or simply hang out at these levels awhile longer.
What about other markers? What else happens regularly in the stock market, but hasn’t happened in a long time? Let’s have a look at some other Time Spans and see if we can further present the real picture of “risk” or “no risk.” In the next chart, we are sticking with the S&P 500, this time using our old friend the 200 day average. On a weekly chart, the equivalent for a 200 day moving average is 39 weeks, adjusting for holidays. In the next study, we have the question, “how long has the S&P 500 200 day moving average been in a rising configuration?” As we know at some point all markets turn down, and when they turn down long enough, into a big enough and deep enough decline, eventually the 200 day moving average turns down. Usually to bring this about, a minimal 10% decline is required. So, without wasting time, let's see how long the 200 day average has been rising.
Answer: It has been rising non-stop since November 5, 2004, or what is now a duration of 115 uninterrupted weeks.
For perspective on that number, let's look at the chart above. The lower horizontal line shows the historical average of the prior 20 longest time spans wherein the 200 day moving average advanced at least 40 weeks or more. That average duration figure is 87.30 using a long term average, and 81.82 using a long term mean. The upper horizontal line shows us where we are right now. At present, the current time span is already the fifth longest time span seen in the last 56 years and is closing fast on the time span which ended at 125 weeks on June 18, 1965. What’s more, let’s look at what happened once these prior long dated Time Spans ended. Starting with the first big bulge over 87 weeks in 1952, we saw a Time Span end at 106 weeks on 10/10/52. In this instance, back in the heart of America’s growth heyday, the market went down to sideways into January 1954, a period of no progress for 14 months. Next, we see the Time Span ended 9/21/56, at 142 weeks (the big spike on the left) and that led directly into the 1957 bear market where the S&P declined for one year and fell 20.40%. Next, the Time Span ended on 06/18/65 at 125 weeks, where within a year the market was lower by 13.07%. Then comes the long dated Time Span ended on October 16, 1987, and we all know what ended that one, right? The downward spiraling vortex of 1987 and the stock market crash, the dollar crash, rising long term interest rates, exploding trade deficits, …ah…sound familiar? But I digress. And then finally, we have the 192 week (super—daddy) Time Span ended on 08/28/98--by guess what? Of course, another financial crisis; this time the Asia Crisis, where the S&P fell 22.43% in 12 weeks.
The point here is that on virtually every occasion we have seen this type of elongated period of one sided stock market movement, the demise has not been pleasant. Why might this be so? Markets seem to go through cycles where they price in varying degrees of risk, at times when confidence is high, ignoring even the greatest risks with a dismissive version of “talk to the hand.” At other times when confidence begins to deteriorate, the focus begins to sharpen with a razor blade edge, and at those times, all of the mis-priced risks, all of the inefficiencies in the system are purged as are all willing and unwilling sellers. Who says that there is anything new under the sun in the stock market?
For time and memorial these cycles have reigned, periods of low volatility followed by periods of high volatility. In the chart above, we see the S&P since 1970 with a Time Span Counter on the bottom clip. In this Time Span Counter, we are checking in on the number of days the S&P has gone without a 2% single day decline. Guess what? We are in record all time territory right now with a string of 917 trading days without even a 2% single day decline! Now, if that doesn’t tend to blur the memory regarding what downside risk is, then I don’t know what will. Anyone who needs a refresher on coping with risk is invited to spend a few weeks in the Energy pits or the Gold pits, where scarcely a day goers buy without heart stopping volatility.
But low volatility have been seen before, and they always end. When they do, watch out, because the close cousin of no-volatility is high-volatility, and the stock market is great at going from one gear right into the next. Over the last 100 plus years, the Dow and S&P have had a few instances when they have been able to string together a few positive years in a row. For example, in the 1990’s the S&P gained from 1995 to 2000, a period of six years. In the 1920’s, the DJIA gained 6 years in a row from 1924 to 1929. In both cases, six years has only been done twice, and on those occasions. In addition, in both circumstances, the markets crashed into violent and extended bear markets which subsequently erased all of the preceding gains in the prior six years. Not pleasant.
Outside of the two six year anomalies, there have been one or two five year streaks, the DJIA from 1949 to 1953 as one of them, but by far, the majority have seen three year winning streaks end and run into trouble in year four. Coming into 2007, the S&P 500 is up four years in a row, with a gain of 26.38% in 2003, +8.99% in 2004, +3.00% in 2005, and 13.61% in 2006. Pushing for a fifth year of gain against this backdrop seems to be pushing the statistical envelope. The only time in history when the market entered a pre-election year up four years in a row on a major index was in 1999, and it did manage to press even higher peaking in 2000. Of course, we all know what happened in 2000. In addition, it is worth noting that back then, that market had taken a huge breather in 1998 with the averages having a 20% correction during the Q3 1998 Asia Crisis. The averages recovered to end 1998 in positive territory, but at least within the 5 year streak there was a healthy washout — correction of 20%.
Never, and I mean NEVER, have we seen a four year bull run extend outward without at least a 10% decline somewhere in the mix. That is the challenge we face in 2007, as I will end with this final chart showing 100 years of the DJIA and a final Time Span chart--this time, a Time Span that tracks 10% corrections in the Dow. In the past 100 years, we have now gone the longest period of time — thru today, a total of 959 trading days in which the DJIA has not seen a pull back of at least 10%. That has never happened before with this current streak easily dwarfing the prior long dated runs on 839 days in 3/30/1994, 806 days ending on 10/13/87, and 772 days ending on 6/09/53.
As I said earlier in this piece, this is the most over-extended stock market, and probably one of the most over-owned and uncorrected stock markets history has ever delivered. For those of you who like Presidential Cycles and are feeling lucky (and I wish you good luck in 2007), something tells me that against this type of stacked deck, you’re gonna need it. Seldom, if ever, has risk premium been this low, and perceived risk, this obfuscated. History says 2007 should favor a Return of the Bear.
© 2007 Frank Barbera