[ Glossary menu ]
Except for the Dow, which is price-weighted, most major market averages
and industry/sector indexes are capitalization-weighted. This means
that the daily impact of a stock's price change upon a given market
index is weighted by multiplying its price by the number of shares
outstanding. The result is that a handful of large-cap stocks will
exert the most influence on the index. For example, at the beginning
of 2001 the top 25 (5%) of S&P 500 stocks accounted for 40% of
the total market cap of the index, and the top 50 (10%) for 57%, and
the top 100 (20%) for 72%. So the premise that investing in an S&P
500 index fund will provide diversification is clearly not true.
I decided to create unweighted versions of the major indexes so that
we can get an idea of how the index would look if each stock were given
equal weight. This will give us a clearer assessment of market internals,
particularly market breadth, and we can see when money is moving into
or out of the smaller-cap components of the index. One way to think
of it is as a price-weighted advance-decline line.
The way I have chosen to construct our unweighted indexes is with
a simple arithmetic calculation -- calculate the daily average percentage
change of each stock in the index, and increase or decrease the prior
day's index value by that percentage. This is the way Marty Zweig calculates
his Zweig Unweighted Price Index or ZUPI. (See his book "Winning On
Wall Street".) And in his book, "Stock Market Logic", Norman Fosback
says of this method: "This is an eminently fair and reasonable method
of calculating stock market indexes. Every stock receives equal treatment,
regardless of price or capitalization."
According to their March 25, 1988 supplement introducing the Value
Line Arithmetic Index, this is the method Value Line uses to calculate
it. Their purpose in establishing this index was to provide "a good
estimate of the price performance of an equal-dollar portfolio of the
stocks reviewed by Value Line". This statement brings up an important
point about arithmetic indexes: We have to assume that the theoretical
dollar value of the index is equally distributed among each issue in
the index at the beginning of each market day.
Arithmetic averaging is not universally accepted because of the assertion
that it has a distorting positive mathematical bias that adds false
percentage points of performance to the index, but it seems to me that
a theoretical portfolio of stocks in which the total dollars were distributed
equally to each stock each day would perform exactly as the arithmetic
calculations say it would (in a theoretical world in which there were
no taxes or transaction costs). Alternatives to arithmetic averaging
require the use of logarithms or "price relatives" in order to account
for the previous performance of an individual stock, which in effect
causes the resulting index to become price-weighted. This is not what
I wanted to accomplish.
The bottom line is that different methods of calculating market indexes
will render different results, so it is important to understand how
an index is derived and what it intends to convey. For example, when
Value Line began its Arithmetic Index "it was set equal to the Value
Line Composite [Geometric] Index prior to the opening of business on
February 1, 1988. At that time the value of both indexes was 210.75." As
of December 31, 2000 the Arithmetic Index was 1124.76 and the Geometric
Index was 393.47. Quite a difference, but the purpose of the Geometric
is to represent the results of a median price change, whereas the Arithmetic
represents the performance of an equal-dollar portfolio.
In January 2001 Decision Point began publishing unweighted index
charts for the S&P 500, S&P 100, Nasdaq 100, and Dow Jones
Industrial Average. We began the project in November 2000 and we constructed
the prior two year's historical data using the components of the indexes
at the time we started the project. Since we didn't account for index
component changes during that prior two years, the indexes for that
period are not 100% accurate; however, we believe that the unweighted
nature of the indexes minimizes the effects of these inaccuracies.
Data subsequent to November 2000 incorporate ongoing component changes
in the indexes.
Our unweighted index charts also feature a Ratio Index that is derived
by dividing the unweighted index by the weighted index. When the ratio
is rising, it indicates that the unweighted index is stronger than
the weighted index. If it is falling, the weighted index is stronger.
EQUAL-WEIGHTED EXCHANGE TRADED FUND (ETF)
On April 30, 2003 the new Rydex SP Equal Weight ETF (RSP) began trading.
It is based on the S&P EWI (Equal Weighted Index), an index composed
of the S&P 500 stocks but calculated in a way to give each stock
an equal weighting. The index is rebalanced quarterly. While RSP is
a relatively new security, we fortunately have data prior to 4/30/2003
, which has been derived from back calculations of the S&P Equal
Weight Index performed by Rydex and Standard & Poors. It is an
accurate depiction of how RSP would have performed, and it allows
us to examine RSP performance over a wide range of market conditions.
Now, why do I think this new product is so important? Because
an equal weighted index: (1) provides investors with true diversification;
(2) is not subject to the fortunes of a handful of stocks with very
high market caps; (3) has the potential to out-perform the cap-weighted
version; and (4) could help technical indicators provide better forecasting
Capitalization-weighted indexes do not provide true diversification
For example, about 70% of the S&P 500 weighting is in the 50 (10%)
largest stocks, so claims by many financial advisors that investing
in an S&P 500 Index fund diversifies your portfolio over 500 stocks
is simply not true. With investment products that track the S&P
EWI, you truly can attain broad diversification.
Perhaps the best argument for this equal-weighted index is that there
is a direct relationship between price performance and the derivative
indicators. Virtually all market indicators, like the Advance-Decline
Line, are unweighted indicators -- each stock in the index carries
an equal weight. Unfortunately, using them to analyze their cap-weighted
price indexes is not always effective because the movement of the price
index is determined by the weight of a just a few stocks. Now the relationship
of the price and internal indicators should prove much more helpful
to our analysis.
Performance is an important issue, so how does RSP perform relative
to the S&P 500 Index? On the chart below we compare the two indexes
going back to 1990. At the bottom of the chart is a ratio index that
divides the RSP by the SPX. When that index is rising, it means RSP
is stronger than the SPX.
It is interesting to note that there was an extended period between 1995
and 2000 where RSP underperformed the SPX. This was the period when
the mass market discovered stock investing, and indexing (investing
in index funds) became extremely popular. Because of this, money became
concentrated in large-cap stocks, causing them to appreciate at a faster
rate than smaller-cap stocks. This divergence reversed after the SPX
topped in 2000, caused primarily by the fact that large-cap stocks
began to tank. RSP did not begin a serious decline until mid-2002,
which is when stocks, large-cap and smaller-cap, were being abandoned
across the board.
There are some interesting statistics regarding comparative performance.
From the 1990 bear market low to the bull market highs in 2000 (SPX)
and 2001 (RSP) the SPX rallied +428% versus +408% for RSP. The bear
market decline into the October 2002 low took the SPX down -50% versus
only -40% for RSP. Finally, as of 9/16/2003 the rally from the October
2002 low has taken the SPX up only +33% versus +54% for RSP. Surprisingly
RSP is only -8% off its all-time high versus -33% for the SPX. The
chart below gives us a closer look at the more recent time frame.
The Rydex SP Equal Weight ETF (RSP) has characteristics that make
it superior to capitalization-weighted versions of broad stock market
indexes. I think it tends to be more stable, and will generally perform
better except during times when indexing is dominant preference of
Decision Point has a Straight Shot series on RSP, which presents
a series of indicator charts derived from the stocks in the S&P
500 Index. This permits the useful comparison of equal-weighted indicators
against an equal-weighted price index.
Additional information regarding RSP can be found on the Rydex