Expectations coming into 2005 were extremely high with the vast majority looking for a strong January to kick of the new year. This was not only evident by the position of most of our sentiment indicators but also from the tone of comments in the financial press. January was to be the kick off to the fifth year of the decade, which over the past 120 years has not only not been been negative but historically it has had the best average returns of any year of a decade. Of course the market, as it usually does turned right while everyone was looking left selling off sharply to open the year, and tracing out a key negative reversal the first day of 2005.
There went the January effect as the decline was lead by the stocks that had performed the best in 2004, the small caps as measured by the Russell 2000 and also the transportation stocks. This brings to mind an old wall street expression "Last years winners, this years losers". This may have been borne from the fact that a lot of investors chase performance and when one theme gets overly crowded with a one way decision this usually marks a peak or trough. It is hard to argue with history. As such the track record of the decennial pattern is one tough cookie. Twelve straight wins is as strong as it gets. However, this decennial pattern and year five has not only become common place in market forecasts this year but has also been accepted by Main Street. And not to a small degree but to what may be best described as an obsession. It seems as though nearly everyone is not only talking about the strong year five but are approaching it as their inalienable right to have a big year in 2005. We have been in this business for nearly 22 years. We have seen two other fifth year of a decade periods, 1985, and 1995. We do not recall in either period much talk if any in regards to the decennial pattern and the historical performance of year five in either of the prior instances. Yes there was some talk of it but that stayed mostly in the technical and cycle circles and certainly did not become mainstream.
In all of our years in the business we have never found an indicator that has been 100%. It would be great to do so but frankly it does not exist whether it be a momentum indicator, a sentiment indicator or what have you. A good example of this was seen in 2000 to 2002 as it relates to the seasonal periods. The May-September/October period is historically negative while the September/October to April period has been historically positive. In 2000 this pattern failed consistently. The market rallied strongly from late April 2000 to September 2000 and declined from a September peak into March of 2001. It peaked in May on a counter trend rally and declined into September. In 2003 from May to September, the negative seasonal period, the S&P gained over 15%. This is based on the May 1 print low to the October 31 print high. There are no doubt many reasons why over the past few years this pattern actually was 180% opposite of its historical behavior. We do not think that we will ever truly know why this occurred but one thing we did notice is that the seasonal pattern had become considerably more mainstream over the past several years than it had previously. The point is that there will be a time when the decennial pattern will fail as all indicators eventually do. The fact that it is so widely talked about and so widely expected suggests the possibility that just maybe it will fail this year. This is not a forecast nor are we trying to play devils advocate but the main job of the market is to make the majority wrong the majority of the time. The higher the degree of acceptance the bigger the potential failure or success.
As we mentioned, coming into 2005 expectations were not only high but bordering on extreme. A number of our key sentiment indicators had moved to levels last seen in February-march of 2004. Some such as the Rydex ratio for the S&P moved well above last years peak. We also saw the volatility indexes move to nine year lows. To be fair, the SA&P also surpassed its March 2004 peak so the move higher by the Rydex ratio is not as extreme as it would have been if the S&P had failed. The decline the past two weeks has produced some easing in a few of the indicators. The most notable is the put to call ratio as the 10-day average for the total CBOE is at the high end of neutral. However, the majority of these indicators although off their peak to a degree are still strongly bearish. A prime example can be seen by the sentiment composite as it has fallen to a very low +3 (it ranges from zero to +24).
The rally in late December left a number of bearish divergences in place from a number of momentum indicators such as the McClellan oscillator, our own volume oscillator and such indicators as 13-day RSI. We also had divergences from the daily new highs, which peaked on December 1. However, as important as these divergences were (are) to the short-term there is a much larger series of divergences in place that have medium and long-term implications for the market. For example, not only did the new highs fail in late December to move above their early December peak but they failed by an even wider margin to move above their late 2003-early 2004 peak in spite of the higher high in the S&P. We see the same failures from a number of indicators. Thirteen week RSI, the weekly new highs, and the percentage of stocks above their 200-day moving average are but a few of the multitude of medium and long-term divergences that are popping up. Others include the McClellan summation index, which has made three lower peaks since its June 2003 peak. At the same time, the monthly trend oscillators remain negative while the weekly measures are beginning to roll over. in fact the more sensitive component of the weekly trend measures has turned down as of last week. For the record, this indicator is also showing multiple negative divergences from its peak in June 2003, a pattern last seen from January 1999 to September 2000. That also occurred with a declining monthly trend oscillator.
The late December short-term divergences were confirmed in early January with a number of these same indicators moving to oversold levels very soon after the price peak. For example, the McClellan oscillator moved to -217 only three days after a new multi month/multi year high in the S&P. This type of move is more often than not an initiation signal not an ending one. The rally the past week has to a degree relived the deep oversold reading on a few indicators but others such as our breadth and volume oscillator are at or still very near oversold levels as are both the 10-day and open 10 Arms. In essence we have a mixed picture from the momentum indicators but on balance they are slightly supportive of a counter trend rally. In looking at a sixty minute chart of the S&P from purely a chart pattern we do see the potential of a small base developing that would also support a counter trend move. We are leaning in favor of this, a strong bounce over the very short-term. We do not see this potential bounce as anything more than a correction of the post January 3 decline but it does have the potential to be a fairly robust bounce. We remain bearish short-term with a stop of 1188. Medium-term the evidence is building that what we saw on January 3 was more than just a short-term top and decline. Medium-term we are beginning to lean in this direction but do not have enough evidence to support this although it is building quite rapidly. For now we are going to remain neutral on the medium-term. we want to see how any rally unfolds from current levels. Long-term we remain bearish as the technical evidence continues to mount that the market is in the latter stages of its bear market rally if that rally has not already ended.
