Stock Market Cycles, Part 7: Secondary Market Moves

See parts 1, 2, 3, 4, 5 and 6.

We have selected the Dow Jones Industrial Average (DJIA) for analysis because data is available since before 1900. In the seven tables that follow are shown all the market moves of the past 107 plus years that were 10% or larger, based on closing prices (daily). There have been 231 such moves since 1900, an average of 2.14 per year. However, we shall see later in this article that the distribution of secondary market moves over time has not been at all uniform, with the 1930s averaging 5.7 secondary moves per year and the 1990s averaging only 0.8 per year.

A color code in the last column of each table indicates which moves were part of a primary bull market (green shading) and which were in a primary bear market (red shading). Note that all counter-trend changes are less than 20%, but some of the changes in the same direction as the primary trend are 20% or greater. A counter-trend change of 20% in a primary bull or bear market, marks the end of the primary market. For example, if there has been an ongoing primary bull market, a drop of 20% marks the end of that bull market and the start of a new primary bear market. The top of the concluded bull market is the high from which the 20% drop occurred.

Counter-trend moves are commonly called “corrections”. Changes in the same direction as the primary trend are commonly called “legs”. We have called all changes “legs” to simplify table headings. Some entries in the “leg” column in the following tables are actually “corrections”.

For an example, in Table 1 look at the primary bull market that started at the end of the bear market ending on the 6/25/1900 bottom. There are three moves included in the primary bull market. The first is a leg up of 10.0% ending 7/23/1900. The second move is a leg down (a correction) of -10.3% ending 9/24/1900. The third and final move of this primary bull market is a leg up of 43.4% ending with the primary market top 5/1/1901.

The move up of 43.4% is considered a secondary bull market move because it is included in a series of moves that has not been interrupted by a decline of at least 20% that would indicate the end of the primary bull market and the start of a subsequent primary bear market. A move of any size 10% or larger in the same direction as the primary market is considered a secondary market move. This is true even when a bull market consists of only one leg up, without a correction of 10% or more.

The tables in this article include two columns under the heading “Secondary Cycle”. These columns identify secondary market tops and bottoms. Recognize that some of these secondary tops or bottoms are also primary tops or bottoms. The cyclical gains (losses) from one top to the next are also shown in these tables. This data will be discussed in Part 10 of this series (Secondary Market Cycles). Secondary bull markets will be discussed in Part 8 and secondary bear markets in Part 9. Following Tables 1 – 7, there is a discussion about how the number of legs in a primary market do or do not relate to gain/loss and duration of a primary move.

Table 1. Secondary Market Moves (1900 – 1914)

Table 2. Secondary Market Moves (1915 – 9/3/1929)

Table 3. Secondary Market Moves (9/4/1929 – 2/27/1933)

Table 4. Secondary Market Moves (2/28/1933 – 5/29/1946)

Table 5. Secondary Market Moves (5/30/1946 – 1987)

Table 6. Secondary Market Moves (1/12/1973 - 1999)

Table 7.Secondary Market Moves (2000 – 11/20/2008)

In the following two sections we examine if there are any patterns discernible for such things as correlations between the number of secondary legs and the duration and/or gain (loss) for the primary bull and bear markets since 1900. After that, a third section looks at the distribution of secondary market moves over time for primary bull markets and bear markets combined.

PRIMARY BULL MARKETS

The following two tables show the distribution over time of the primary bull markets since 1900, identified by number of legs in the primary move, and the gains for each number of legs.

Table 8. Distribution Over Time Primary Bull Markets With Number of Legs

Table 9. Gains for Primary Bull Markets For Each Number of Legs

Table 10. Durations for Primary Bull Markets For Each Number of Legs

We have color coded the entries in Table 9 and Table 10 and in other tables that follow (Tables 12 and 13). The peach color identifies gains for primary bull markets up to and including the market top of 9/3/1929. The purple color shows markets following the market top of 9/3/1927 through 1940. The light blue color indicates markets which topped since 1/1/2000.

There are some observations that stand out in Table 8, Table 9 and Table 10.

1. There is a tendency for larger gains as the number of legs increase in a primary bull market.
2. There is no trend evident for durations of primary bull markets as a function of number of legs for one, three and five legs. However, the four primary bulls with seven or more legs all have much longer durations. The average for all the bulls with one, three and five legs is 374 days (standard deviation = 371). The other four primary bulls (seven legs and more) are all obviously not part of the same distribution (one nearly three standard deviations and the other three much more away from the average of the other 28 data points).
3. A high percentage of primary bull markets with only one leg (9 out of 11) have occurred in a small segment of the time span: from the 1929 crash through 1940 and the almost 8 years since 2000 (18% of the time).
4. Nine of the one leg primary bull markets had returns between 20.8 and 31.2 and the other two look like outliers: 93.9% (market top 9/7/1932, duration 42 days) and 69.8% (market top 7/16/1990, duration 659 days). What is significant about these two outliers is that they followed two of the three most traumatic bear events since 1900, the post 1929 crash bottom of 7/8/1932 and the 1987 crash. The third trauma, the 1974 bear bottom was followed by a similar substantial bull market (75.7%) that took three legs and an intermediate time: 402 days.
5. The nine one leg primary bulls that occurred in the two short time frames (1930-1940 and post 2000) had an average duration of 64 days with a standard deviation of 38 (ratio = 1.7, fair “tightness”). The other two one leg primary bulls had much longer durations (255 and 659 days), obviously not part of the same distribution.
6. There does not seem to be any consistent pattern for the three market tops preceding the three traumatic bears. The 9/3/1929 top was for a primary bull market of 17 legs covering 2169 days and gaining a record 471%. The 8/25/1987 top followed three legs over1269 days and a gain of 250.4%. The third event is different. The 1/11/1973 top followed three legs over 665 days and a gain of 66.6%.

PRIMARY BEAR MARKETS

The following two tables show the distribution over time of the primary bull markets since 1900, identified by number of legs in the primary move (Table 11), the gains (Table 12) and durations Table 13) for each number of legs. As in Tables 9 & 10 (above), we have color coded the entries in Tables 12 & 13.

Table 11. Distribution Over Time Primary Bear Markets With Number of Legs

Table 12. Gains for Primary Bear Markets For Each Number of Legs

Table 13. Durations for Primary Bear Markets For Each Number of Legs

There are some observations that stand out in Table 11, Table 12 and Table 13.

1. Unlike primary bull markets, primary bear markets do not show any strong trend of gain (loss) with increasing number of legs. (The 1, 3 and 5 leg averages +/- one standard deviation are all significantly overlapped.)
2. There is a tendency for the duration of primary bear markets to increase with the number of legs, at least up to seven legs, in contrast with the lack of trend for primary bull markets.
3. One-half of primary bear markets with only one leg (7 out of 14) and more than 50% with three legs (4 out of 7) have occurred in a small segment of the time span: from the 1929 crash through 1940 and the almost 8 years since 2000 (18% of the time).
4. In contrast with primary bull markets, a majority (2 out of 3) of the bear markets with more than seven legs occurred during the Great Depression.
5. There is much more uniformity in primary bear market losses than for gains across all the primary bull markets. This was previously pointed out in Part 2 of this series.

SECONDARY MARKET MOVES IN BOTH PRIMARY BULLS AND BEARS

In the Table 14 are shown the distribution of secondary market moves over time since 1900.

Table 14. Distribution of Secondary Moves Over Time

The graphs below are informative. The first graph emphasizes the fact that for all the decades up to the 1940s the number of secondary moves in primary bear markets significantly outnumbered secondary moves in primary bull markets, with the exception of Roaring 20s. In contrast, until the current decade, all the decades starting with the 1950s saw the number of secondary moves in primary bull markets equal or exceed the number in primary bear markets. This is emphasized in Table 14 where the ratio of the number of secondary legs in primary bear markets to those in bull markets is shown in red whenever the bears had more secondary moves. In the current decade the ratio has returned to pre-1950 behavior, with the second highest ratio in history (since 1900).

The second graph, shown below, shows the variation of all secondary market moves by decade. Superimposed on the bar graph is the linear trend line from 1900 to the present. Three decades show the number of secondary market moves above the trend line: the 1930s, the 1970s and the 2000s. One may consider the number of secondary market moves as a macro measure of volatility. Clearly, these three decades, which are generally considered to be the three decades with the greatest financial stress since 1900, also showed the greatest market distress. The two completed super bear market bottoms were seen in the 1930s and the 1970s. It remains to be seen if the third super bear market bottom is reached in the 2000s decade. Whether the super bear is underway at present or not, the last few months have seen one of the most precipitous market declines in history.

SUMMARY

Some summary observations regarding the changes observed for different numbers of legs in primary markets are shown in Table 15:

Table 15. Summary of Results With Different Numbers of Legs

Other summary observations are:

1. Almost 82% of the primary bull markets with only one leg occurred during two time intervals (1931 through 1940 and March, 2001 to date) comprising only 16% of the time span studied.
2. One-half of the primary bear markets with only one leg (and more than one-half of the primary bears with only three legs) occurred in the same 16% of the time.
3. Only three decades (1930s, 1970s and 2000s) have a number of secondary market moves greater than the 107+ year trend line. There is a correlation between a large number of secondary market moves and periods of severe economic stress.
4. Several points about the ratio of secondary legs in primary bear markets to the number in primary bull markets:
   1. Four of the five decades before 1950 had more primary bear market legs than primary bull market legs.
   2. The five decades starting in 1950 had fewer primary bear market legs (or the same number for the 1980s) than primary bull market legs.
   3. So far for the decade starting in 2000 there have been more than twice the number of legs in primary bear markets as in primary bull markets.

Comments

[sleepless_o...:]

This may be too academic for some SA readers, but I find it a valuable reference work. Hope you will continue to contribute comprehensive analysis pieces to this web site.

2008 Nov 24 11:43 AM

[nym:]

Table 6 is missing.

2008 Nov 25 04:24 PM

[sleepless_o...:]

nym - - -

Please check again and let me know if there is a problem. I find Table 6 right between Tables 5 and 7.

Thanks

2008 Nov 29 12:22 AM