SP500: A four part review of historical prices.
History Part 1. Pattern Recognition.
Did you ever consider the influence of pattern recognition? What is next year's outlook from the viewpoint of history repeating itself? It looks as though you will not be missing much next year. Prices historically chopped around for the first half of the year, but the second half of the year doesn't have the same predictability. Years ending in 4 show flat prices while the subsequent years chart show a strong second half. This divergence in historical fact will be somewhat problematic until the common factors of 1989-90, 1995-96, 1997-98, 2003-04 can be identified. What's the common denominator? What if any tax laws changed?
The graphs below used the daily highs of the SPX from 1982. These were then converted into percentages based on the previous year's last price. The horizontal axis represents the number of trading days in a year.
|
|
|
|
History Part 2. Annual Analysis.
Now for part 2 of your history lesson. Study the following graph. How many times did the SP500 go down and close in the lower half of the trading range and then reverse the following year and close high while posting at least 30% gains from one year to the next (low year 1 to high year 2)? This chart has been on this site all year, but how many of you saw it and really reviewed it carefully. If you had, you would have seen that 1957-58, 1962-63, 1966-67, 1974-75, 1981-82, and 1990-91 have a lot in common. Notice that 1970-71 wasn't included because it closed near the high of its range in 1970.
What else does this chart tell us? Well, since 1950 the market averaged 9.1% year over year gains. That's different from the compounded annualized gain of 7.5%, and these figures do not include dividends, and do not reflect the total market returns. These only pertain to capital appreciation. Next, the SPX went higher on average by 16% from its Dec. 31 close and it went lower on average by 9%. So the average range of the SPX was 25% per year. The key is to buy more when prices are near the bottom of the range and sell near the top of the range each year if you're a short term trader. And if you're a long term investor, then simply buy when prices are lower. The problem is you don't know when that will occur. So buy a portion at the beginning of the year and then wait to buy another portion when the market dips because it will. As a matter of fact, the market decreased in value by 5% or more 30 out of the last 53 years, or 56.7% of the time. The other 23 out of 53 years it didn't. That's why if you divide your expected investment dollars into thirds, or invest monthly, then you can invest in the beginning of the year and then have 2/3 of your money to invest when prices drop the expected 5% to 10%.
Another fact from this chart is that 35 out of 53 years resulted in positive gains. This by itself is good news for investors, but there's a dark side to this statistic. In the world of statistics, this means that there is 95% probability that this didn't occur by chance, which is good, but it failed the 99% probability test, which is bad. This means that there is still the possibility that these gains are the result of chance. This is little disconcerting given that we all would like to think that investing is based on causality. Positive factors should cause prices to rise and negative factors should cause prices to fall. However, the fact that chance, or chaos, may be the governing force makes sense when you consider the fact that if investing were easy and predictable, everyone would be a winner.
Now let's ask if there are there any patterns that can help us to identify large gains and small gains. Well below is a graph that plots the highs separately from the lows. This is the same data as shown above except that these data are connected differently. Now you can see a pattern.
Plus, in all cases when the SPX posted a 20% gain or better, the subsequent year nearly always posted a gain. There were 17 years out of 53 that posted annual highs above 20% and only 1 out of these 17 posted greater gains created by the high of the subsequent year. So expecting the market to produce gains greater in the subsequent year is a "long-shot" (poor odds). This means, we should not expect the market to make a high greater than 22% from the close of 2003 in 2004. The market can go higher, but it isn't likely that it will go 23% higher from the last trading day of 2003.
Second, 6 out of 17 posted gains for the subsequent year of 0% - 5% and 7 out of 17 posted gains of 5%-10%. So 13 out of 17 posted gains from 0% -10%. Only one out of the 17 posted 0%. So, 1:17 posted greater gains than the previous year; 1:17 didn't go up at all; and 16:17 posted gains. These are terrific odds for 2004. So will the market go higher? History tells us that there's a 94.1% chance that it will go higher. But history also tells that the odds favor 1%-10% higher. It also tells us not to expect another year with the high being 20% greater than the close of 2003.
As for the bears, history only shows us one occasion, which is comparable to 2002-2003, and that is 1973-74. The market rallied during the subsequent two years and the lows were only 1% below the close of the previous year. So history is telling the bears to hibernate. Will history repeat itself ?
Now let's return to the Normalized Annual Range chart and examine those years mentioned that are similar. If those years were studied closely, this is what your history lesson would have produced. The correlation to 2002-03 is remarkable and there are quite a few market reactions that were similar. Will the markets repeat themselves one more time? Maybe we should learn to study history more often
Now let's view some of those years displayed above with lows lasting several months and align them. Basically, the graph below shifts the data horizontally without disturbing the price changes. So the graph above differs in that it represents three whole years starting with January 1 and the graph below doesn't start with January 1 nor does it last precisely three years.
The graph below now shows the market reactions both in terms of price and time. As for price, you can see that the SP500 went down 15% to 35% in value and then rose 10% to 30% above the starting values, or 35% to 45% above the lows. Plus you can see that the relative price changes show that the reaction can still go higher from Dec. 2003's current position in the pattern. As for time, you can see that the lows came quickly and the highs followed. The highs appeared somewhere between 500 and 600 days later, or 300 to 400 days after the lows.
In addition, notice that the 2003 reaction has less strength than in the other years. First, in terms of percentages, 2003 went lower than in the other cycles. Second, it stayed near the lows longer than any of the other reactions. Third, 2003 has a triple bottom, which doesn't appear in the other reactions. Fourth, notice how the relative price changes are less than the other cycles. In other words, the upward rebound in 2003 is below the other upward reactions. In summary, it's taking longer in 2003 to climb back to break-even.
These are all signs of weakness that the other years didn't exhibit. So despite these pattern's similarities, they also exhibit differences that preclude forecasting the future accurately. However, these historical patterns are compelling.
History Part 3. Correlational Analysis
Here's another attempt at discovering how often history repeats itself. In this method the monthly closes were converted into absolute terms of up or down. Then history was scanned to match the 10 ten months of 2003 from the March lows. It looks like history is telling us that the market still has energy to go up, but not much. Prices look to be range bound after a strong ten month period.
History Part 4. The Presidential Cycle.
There will be many rumors that President Bush and Fed Chairman Greenspan will "push the market: higher. And the manipulation theory will surface as it always does during the last year of the Presidency. So here are the facts as history has recorded them so that you can make an informed decision rather than an emotional one.
The following two charts present the SP500's prices relative to election day (or the day before election day during those years when the market was closed on election day); not Dec. 31 or Jan. 1 as most analysts provide. Notice that this cycle maybe only 1 of 2 cycles in which the market will not be profitable. Of course, there's a whole year ahead of us, but as of Dec. 2003, the market still hasn't reached the break-even point.
|
|
|
The above graph shows a mid-term dip, but what about the last year of the presidential cycle? The graph below presents the presidential cycle in days from election day, so how have previous election cycles performed starting with the 780th day (Mid-December 2003 to re-election) of the Presidency. As the graph depicts, only one cycle lost money from this starting point of December of the third year. Three broke even and 7 cycles earned 10% or more. So the odds are 1 out of 12 that you'll lose. However, there were 5 years in which the market went down 5% before going higher. So those who use 5% stops lost 5% 5 out of 12 times, or nearly 50% of the time.
So let's get this straight. During the last 13 months of the presidency, your odds of losing 10% are 1 out of 12. Your odds of losing if you set your stop at 5% are nearly 50%. This means that you have 50% chance of making 10% to 20%. And if you're thinking of using a 10% stop instead, then at least you odds of losing are only 1/12 rather than 5/12. Given the risk of only 5% using 10% stops seems to be a better choice. Of course if you didn't use stops, you'd have only lost 6% 1 out of 12 cycles. So ask yourself who wins by using stops? Basically, the last year of the Presidency has yet to go down more than 10%. It hasn't happened in the last 12 cycles.
So how can you use this information to your advantage? First, realize that the market will fluctuate. Second, 50% of the time during the last year of the Presidency the market will go negative within 40 days or by the end of January. Third, divide your investment dollars by 4 and purchase stock when prices dip below current levels. This gives you 4 times to buy stock in case prices go lower. Basically when prices get 5% to 10% lower, that's as low as the market should go. If it breaks lower than that get out or don't buy stock, otherwise, the market should rebound from -5% to -10% to 0% to +20%. Then when prices return to break-even, you'll be in profits because you purchased stock below the break even point. Then evaluate the market's position and strength by following the highs and lows. If the market can't trend higher by making higher highs and higher lows, then get out.
By the way, your odds of making 20% this next year are 17% because there were 2 out of 12 cycles that made 20%; making 15% is 50%; making 10% is 75%. So these are great odds considering the alternative is to park your money in a 1 year Treasury bill that will only produce 1%. So history tells us that the stock market is the place to be ... but waiting for lower prices is the hard part.
By the way, if you combine the odds mentioned below in History part 3, which demonstrated that of those years subsequent to the years that posted highs greater than 20%, they weren't likely to post another high greater than 20%. The combined odds of making 20% in 2004 are then only 3% rather than 17%. One last tidbit, if you study the graph you'll see that 3 out of 12 cycles went straight up. So there's a 25% that you won't see lower prices. Think about that.
So in summary, the odds for 2004 are only 3% that it will go higher than 20%. This means that it is possible but not likely. The odds of prices going lower than close of 2003 in 2004 are 40% (79.2% on any given year, 42 out of 53 went lower than close, times the Presidential cycle odds of 50% in the last year of the Presidency). And the odds of making 5% or more are 40% (32 out of 53 years times 8 out of 12 Presidential cycles). The remaining percentage of 17% is for -5% to +5%. So your odds of losing are the same as winning, but given that interest rates are more likely to rise than fall, your more likely to lose in the bond market than in the stock market. Just buy low and sell high as always and the way to do this is not to invest all of it at once because you don't know when the lows will occur. Just be systematic at buying the dips in 2004 because the odds are good historically. Of course you could just stay on the sidelines this year because with even money odds, like 50:50, why invest? Are you feeling lucky?
Below is a chart showing you how prices fluctuated with 6 months remaining in the Presidential Term.
Below is a chart showing you how prices fluctuated with 2 months remaining in the Presidential Term.
created 12/12/03, last updated 9/17/04 ©2004 The Small Investors Software Co.
