Written and copyright © 2007-2013 by Thomas N. Bulkowski. All rights reserved.
For superb price performance over time, pick stocks from the industry showing the best relative strength. The most recent update to the industry list can be found by clicking here.
Industry Relative Strength Background
Industry relative strength has nothing to do with the Wells Wilder relative strength index (RSI, an indicator).
As I use the term "industry relative strength" in this document, it means the performance of stocks within the same
industry. Each industry is compared to other industries daily and ranked for performance based on the change in closing
prices of each stock in the industry as measured from the current day to a day 6 months ago.
Below I discuss a few tests that gave good results but you will want to do your own research.
Industry Relative Strength Methodology
I created a database of industry relative strength beginning from 1/1/1995 to the current date (11/9/2007), for each trading day. The starting date
was arbitrarily chosen, but I wanted to include the 2000 to 2002 bear market (as represented by the S&P 500).
I used 512 stocks and placed them into 44 industries (with between 5 and 23 stocks each, but most had about 10
companies represented. I selected each industry because it interested me and those are the ones I follow on a daily basis.) then measured the close-to-close price change over the prior 6 months
chosen, but 6 months is the length I use in my daily analysis and it works well (I have daily, weekly, monthly, 3 months and 6 months options on my screen and have played with them).
Other periods may work better but you can test this.) for each stock, each day. I used
the total price changes for the stocks to find the average percentage change for the industry and ranked each
Once the database was built, I split the industry into four groups of industries: those ranked 1 to 11, 12 to 22,
23 to 33, and 34 to 44. For performance over time, I looked at three industries within each group: the best
performing, the middle industry, and the worst performing. Then I looked at future performance. For example, if the
drug industry was the topped ranked one on January 2, 1995, I computed the average price change for the industry
from that date to 1 month, 2 months, and so on, into the future. I did this for the four groups and the table shows
the results (see below). Each cell in the table shows the percentage change of the industry over time.
For example, the best performing industry (Group 1, top rank) gained 8% in a month, 15% in 2 months, and 98% in
two years. The industry ranked 11 (the bottom ranked industry of Group 1), made 2% in a month, 4% in two months, and
31% in two years. This compares to the bottom ranked industry in Group 4 with a 6% decline after one month, 11% drop
in 2 months, and a decline of 25% in two years.
Industry Relative Strength Results
The results are startling. If you want to make money in the stock market over the longer term (as short as a month to as long as 2 years, but probably longer), pick stocks from the top industries, preferably from the industry showing the best relative strength. Avoid selecting stocks from the industry showing the worst relative strength, unless you are looking for stocks to short. Since some stocks within an industry can be hurting, you will want to be selective and chose the strongest relative strength among them, too.
|Performance||Group 1 (best)|
RS rank 1-11
|Group 2 RS|
|Group 3 RS|
|Group 4 (worst)|
RS rank 34-44
|Top rank, 1 month||8%||2%||1%||0%|
|Middle rank, 1 month||3||2||1||-1|
|Bottom rank, 1 month||2||1||0||-6|
|Top rank, 2 months||15||4||2||-1|
|Middle rank, 2 months||7||3||1||-3|
|Bottom rank, 2 months||4||2||-1||-11|
|Top rank, 3 months||23||7||3||-1|
|Middle rank, 3 months||10||5||1||-3|
|Bottom rank, 3 months||7||3||-1||-15|
|Top rank, 6 months||57||13||5||-3|
|Middle rank, 6 months||21||9||2||-8|
|Bottom rank, 6 months||14||6||-2||-28|
|Top rank, 9 months||69||16||8||-1|
|Middle rank, 9 months||25||11||4||-6|
|Bottom rank, 9 months||18||8||0||-32|
|Top rank, 1 year||80||19||10||1|
|Middle rank, 1 year||30||14||6||-4|
|Bottom rank, 1 year||20||10||2||-33|
|Top rank, 2 years||98||30||20||12|
|Middle rank, 2 years||45||23||16||6|
|Bottom rank, 2 years||31||20||13||-25|
Industry Relative Strength Results 2
I conducted another series of tests using data from January 1, 2000 to May 18, 2008 on 550 stocks but not all stocks covered the entire period. The test was used to determine
the best parameters for a model portfolio, and you can view an Excel spreadsheet of the results here. I found that using a
volatility stop significantly hurt performance.
The tests showed that you get the best performance by buying the three top ranked stocks in
the top ranked industry and holding them (meaning almost never sell them). This result may be due to the bear market coming first in the test followed by a bull market.
The best test sold the stocks in the industry when the rank dropped to 45 or below (out of 48, so it is very near the bottom). That gave an average hold time of 610 days,
or about 1 year and 8 months. A number of trades remained open from year 2000, so I do not view this as the optimum configuration. Also, the maximum loss (the largest drop
below the purchase price) was huge, up to 95%. Thus, this test showed when to buy and not when to sell.
The series of tests performed showed that buying three stocks per industry worked best, followed by 4 and 5 stocks. One and 2 stocks were farther down the list.
From a money management perspective, it says to diversify in an industry but owning too many stocks in the industry is bad. Incidentally, the stocks chosen in each industry
were the ones with the best stock relative strength. Thus, each additional stock was performing less well when added to the database.
This link gives more information about the model portfolio.
Industry Relative Strength Duration
How long does an industry remain top ranked? I looked at the rank of all industries and determined how long they remained within the group, as an average. The best and worst ranked groups showed stocks remaining in the group nearly double groups 2 and 3. Why this is the case is unknown, but the best appear to keep performing and the worst keep suffering. The actual duration is shorter than I expected, between 21 and 47 days.
Bull market duration (From 1/1/1995 to 10/9/2007, excluding the bear market from 3/24/2000 to 10/10/2002)
Average length: 47 days of industry ranks 1 to <= 11
Average length: 25 days of ranks 11 to <= 22
Average length: 25 days of ranks 22 to <= 33
Average length: 45 days of ranks 33 to <= 44
Bear market duration (3/24/2000 to 10/10/2002)
Average length: 39 days of industry ranks 1 to <= 11
Average length: 21 days of ranks 11 to <= 22
Average length: 22 days of ranks 22 to <= 33
Average length: 43 days of ranks 33 to <= 44
Last year (10/10/2006 to 10/9/2007)
Average length: 38 days of industry ranks 1 to <= 11
Average length: 22 days of ranks 11 to <= 22
Average length: 25 days of ranks 22 to <= 33
Average length: 41 days of ranks 33 to <= 44
Industry Relative Strength Trading
I tested a system that used the industry relative strength data and found the following:
If you buy all of the stocks in the industry ranked 1 for performance and hold them until the rank fell below 14 (and sell them), you would have made an average of 28.1% from 1/1/1995 to 11/9/2007. The S&P 500 index, over the same hold time, gained an average of just 2.3%. There were 114 trades made during that period.
As I mentioned, the method assumes you buy all of the stocks in the industry, not just the best performers in the group. For the list of stocks I used in each industry, see Relative Strength Rank. The test was not an exhaustive examination of a trading system, so you will want to do your own research to see what works for your markets.
-- Thomas Bulkowski
Copyright © 2007-2013 by Thomas N. Bulkowski. All rights reserved. Anything not worth doing is not worth doing well.