Thursday, May 12, 2005

2005 Stock Market Forecast

THE U.S. STOCK MARKET IN 2005

December 31, 2004


I have made year-ahead stock market forecasts based on George Lindsay's methods twice before, the first time on January 2, 2003 and then again on January 5, 2004. To put my 2005 forecast in perspective I first quote from the summaries of the 2003 and 2004 forecasts.

From the January 2, 2003 forecast:

"The 20 year cycle and Lindsay's 12 and 15 year periods all suggest that a bear market low occurred in 2002 and that a bull market top is due possibly as early as late 2004 but more likely sometime in 2005.

"The drop from the market's December 2002 top will probably end at a low above the October 2002 low and terminate the basic decline which began from the march 18, 2002 top. Counting forward a long basic advance of 26 to 32 months from the upcoming secondary low also projects a bull market top for 2005. Finally, I suspect that the first stage of this bull market will take the form of Lindsay's three peaks and a domed house formation."

From the March 21, 2003 update to this forecast:

" [the] low established on March 12 [2003] ended the drop from the 954 top [of December 2002]. Counting forward 26 to 32 months (a long or extended basic advance) from March 12 we find a bull market high likely sometime between May and November 2005...If this foldback pattern continues to develop the market should now rally to the level of the top of the big rally which preceeded the drop into the July, 2002 low. This is the 1178 level. After 1178 is reached [in late 2003 or in 2004] the foldback pattern would then call for a drop to the 940 level (2004?) and then a rally to 1320 (mid 2005)."

From the January 5, 2004 forecast:

"These calculations all point to the same general conclusion. The first half of 2004 should be bullish although not as strong as the last 9 months of 2003. A good part of the year's first 9 months will probably be spent in an 80 point trading range [in the S&P]. A top should develop around 1178 and be followed by a substantial break of 120-180 S&P points and this break will probably end in the fall of 2004. After that low a fast advance lasting 7 to 8 months should culminate at the peak of the domed house and a bull market top around 1340 in 2005."

What can Lindsay's methods tell us about 2005?

First let's consider the 20 year cycle and Lindsay's long term time periods. The years 1985, 1965, 1945, 1925 and 1905 were all bullish years for U.S. stocks. Lindsay's 15 year 3 month period from bear market lows to bull market highs reinforces this 20 year cycle, bullish prognosis. Adding 15 years 3 months to the October 1990 low predicts a bull market top for January 31, 2006. Moreover, adding 12 years 10 months (Lindsay's time period from bull market tops to bear market lows) to the January 31, 1994 top (which started a year long sideway's period) predicts a bear market low for December 1, 2006. The predicted 10 month interval from 2006 high to 2006 low is the length of one of Lindsay's basic declines. This internal consistency reinforces our confidence in the forecast of a bullish 2005.

The next step in Lindsay's forecast technique is to consider the status of the market in terms of his theory of basic advances. These are time intervals, measured in calendar days, which typically start at bear market lows or at secondary lows near the bear market low and end at or very near the subsequent bull market top. In my 2003 and 2004 forecasts I noted that the 1998-2000 basic advance was abnormally short. Lindsay's theory of alternation would therefore lead me to expect an abnormally long basic advance for the 2002- 2005 bull market. After analyzing the basic declines during the 2000-2002 bear market I concluded that the March 12, 2003 secondary low probably marked the start of a long or extended basic advance. This in turn implies that a bull market top should probably develop sometime between May and November of 2005.

The fact that Lindsay's theory of basic advances predicts a bull market top before his long term time periods do also has some implications. Lindsay often observed that when these two forecast methods are out of sync by a few months the market does its best to make both "come true" for all practical purposes. In this instance I would therefore expect either a top very near the average forecast date (i.e. around September- October 2005) or alternately (see below) a top in July followed by a sideways trading range that terminate in January 2006. In either case 2006 should be a bearish year.

At this juncture we also have two additional pieces of information that were not available a year ago.

In the previous two annual forecasts I said that I expected a Lindsay "three peaks and a domed house" formation to develop during the 2002-2005 bull market. Just such a pattern developed in the Dow Industrials during 2004. The three peaks came in February, June and September and spanned a 7 month interval. This compares favorably with the 6 to 10 month interval Lindsay observed for major examples of this 3P-DH pattern. The subsequent separating decline in the Dow ended on October 25 at 9708. I interpret the December 9 low as Lindsay's base point for measuring forward in time. To this date we add 7 months 10 days to predict July 19, 2005 for the top of the domed house rally and the end of the bull market. Lindsay also observed that after the domed house is completed the subsequent bear market returns at least to the price level at which the three peaks formation began. This in my interpretation is the 958 low of August 2003. Thus 958 is a reasonable target for the bear market low expected late in 2006.

The second piece of information is the development during 2004 of what Lindsay called the "middle section" of the basic advance. In this case it is a declining middle section and lasted from February to October in the Dow and from March to August in the S&P. I shall use the Dow to make my projections to maintain consistency with the 3P-DH analysis above.

Lindsay's "count from the middle section" is a long term tool and in principle can be used only to predict the timing of the next bear market low and of the subsequent bull market high. However, Lindsay himself often described the process of forecasting as similar to the process of assembling a jig-saw puzzle. All of the pieces (forecasts derived from various techniques: long term time periods, basic advances and declines, 3P-DH and counts from the middle section) have to fit together smoothly. This requirement makes the entire Lindsay method much more effective than any one of its techniques used in isolation.

In this instance we know that Lindsay's other methods predict a bull market top for the second half of 2005 and a bear market low late in 2006. Moreover, counting 15 years three months from the March 1994 low brings us to June 2009 as a likely bull market top. Now we can attempt to count from 2004's middle section in the Dow. Lindsay's "point E" for this decending middle section is June 25, 2004 in our interpretation. The first consideration is that the time from this point E to the bull market top should equal the time from the bull market top to the next bear market low. At the moment my best estimate for this low comes from the 12 year 10 month time interval and is December 1, 2006. This is a little more than 29 months from point E and if the count from the middle section were to work exactly this would imply a bull market top 14 쩍 months after June 25, 2004, i.e. September 10, 2005.

Moroever, the duration of the subsequent bull market, in Lindsay's theory of the middle section, should equal the time from point E to the bear market low. My current estimate for this time interval (again based on the 12 year 10 month period) is 29 months and thus I would expect a bull market top in May 2009, almost exactly coincident with the implication of the 15 year 3 month period from low to high.

Let's now summarize the deductions I have drawn from Lindsay's timing methods. First, 2005 should be a generally bullish year. The bull market top could come as early as July 19, 2005 (3P-DH) or as late as January 31, 2006 (15 year 3 month period). My best guess is that in any case the market will trade essentially sideways after July 19, 2005 but that no really bad drop will occur until 2006 begins.

A bear market should be expected for 2006 with a low coming late in the year. The years 2007 and 2008 are expected to be bullish with a bull market top in 2009.

Where might the S&P stand at these highs and lows? Here Lindsay's timing methods are silent but I can make some deductions based on historical averages.

The 2000-2002 bear market dropped the S&P 50%. The last bear market of comparable magnitude was the 1973-1974 bear market. The subsequent 1974-1976 bull market sent prices up 77%. A comparable advance from the 2002 low of 768 predicts a 2005 top at 1350. The 1976-78 bear market dropped prices 28%. A similar drop from a 2005 top at 1350 would give a low in 2006 around 980. This should be compared with the 958 forecast for that low derived from the 3P-DH formation. The 1978-1980 bull market moved the S&P up to 225% of its 1974 low. A repeat performance for the 2007-2009 bull market would predict a top for the S&P in 2009 at 1730.

For those interested in learning more about Lindsay's methods we suggest the booklet "Selected Articles by the late George Lindsay" which is published by Investors Intelligence in New Rochelle, New York.

Carl Futia

Copyright 2004.

No comments: