Three Approaches to Index Weighting

There are three key approaches to indexing that are competing for investor attention:  Market-Cap Weighting, Equal Weighting, and Fundamental Weighting.  Let’s explore how they differ.

 

Market-Cap weighting is the traditional and predominant method of index weighting today.  The S&P 500 index (SPY) or (IVV) is an example.  Companies are weighted in the proportion that their free-float market-cap has to the total free-float market-cap of all the companies in the index.  Giant companies dominate the index and its performance. This tends to be momentum biased.

Criticism - this method systematically overweights overvalued stocks and underweights undervalued stocks.

Equal weighting is less widely used, with fewer investment fund opportunities.  The S&P 500 Equal Weight index (RSP) is an example.  Each company is represented in equal amounts in the index. Each company contributes equally to index performance.  This alternative method removes the systematic overweighting and underweighting features of the market-cap approach. This method is neither momentum nor value biased.

Criticism, this method has higher turnover (lower tax efficiency), higher expenses, and typically higher volatility (due to greater role of the smaller-cap stocks).

Fundamental weighting is a fundamental rules driven method based on factors such as sales, book value, dividends, and earnings.  It likely prevents participation in bubbles and subsequent crashes.  Some people believe that it is a new label for an old method, namely “quantitative value” investing.  Examples, are the FTSE RAFI US 1000 index (PRF) and the S&P 500 Value index (IVE). This tends to be value biased.

Criticism - this method is subject to higher turnover and expenses.

Price weighting, is a fourth and minimally used method that is the basis of the Dow Jones Industrial Average (DIA).  We mention it for completeness, but are not reviewing that method here.

RETURN COMPARISONS

click images to enlarge

Multiple Years

 

One Year

  

Ultra Short-Term

The ultra short-term returns (1-week, 4-weeks, 13-weeks, 26-weeks, and 52-weeks) are posted and undated weekly on our website.  You can access the updated figures here.

S&P COMMENTS ABOUT EQUAL WEIGHT vs MARKET-CAP WEIGHT

S&P published “Equal Weight Investing - Five Years Later” in April 2008.  This is their bullet point summary:

• The introduction of the S&P 500 Equal Weighted Index [EWI] in January 2003 pioneered the subsequent development of non-capitalization weighted indices catering to investors who question market efficiency.
• Equal weighting is factor indifferent. It randomizes factor mispricing and is thus an attractive option for proponents of the theory that the market is inefficient and at times, misprices factors.
• The S&P 500 EWI has different properties than the S&P 500,including a lower concentration of individual stocks and slower-changing sector exposures.
• The S&P 500 EWI has outperformed the S&P 500 since its inception.  The level of outperformance has varied considerably under different market conditions.
• The outperformance of the S&P 500 EWI is a result of the differing weighting and rebalancing processes. In terms of risk factor exposures, a complex and dynamic combination of size and style risk factors have contributed to the difference in returns. It may be difficult to replicate S&P 500 EWI return outcomes through a simplistic combination of style and sector indices.
• Equal weighting also demonstrates long term outperformance internationally.
• Criticism of equal weighted indices has centered on additional turnover and increased capacity constraints relative to market capitalization weighted indices.  While true in abstract theory, neither is a serious hurdle in practice.

They provided these two charts showing the comparative volatility of the market-cap and equal weight indices, and also the correlation of returns between them.

click images to enlarge

 

They go on to say:

The S&P 500 EWI will at any time have different sector exposures than the S&P 500.  The S&P 500 is designed to have sector weightings close to those of the large cap market.  The weight of each sector in the index at any time is dependent on the total market cap of the stocks in that sector relative to the market cap of the entire index.  On the other hand, the sector weights of the S&P 500 EWI will be determined at each rebalancing by the number of stocks in each sector in the S&P 500.  Therefore, the S&P 500 EWI will be overweight relative to the S&P 500 in sectors that contain stocks that are on average, smaller than the average stock in the S&P 500, and will be underweight sectors that contain larger than average
companies.

Similar to the S&P 500 EWI, the S&P International 700 EWI outperforms relative to its market weighted equivalent, has somewhat higher volatility, particularly in recent years, and over time has become increasingly correlated to its market weighted equivalent.

The S&P International 700 EWI has significantly outperformed the S&P International 700 by a greater margin than the S&P 500 EWI has outperformed the S&P 500.  In fact, while the S&P 500 EWI has outperformed in certain market cycles and underperformed in others, the S&P International 700 EWI has consistently outperformed.

Often the most powerful investment ideas are simple. … sophisticated investors have realized that equal weighting creates a different set of risk factor exposures than market capitalization weighting that seem to work over the long term.

Further, the concept randomizes factor mispricings in the market. As trading costs shrink globally, and as investors realize that turnover of equal weighted indexing is only about a fifth of active managers, we expect the concept to gain ground. … we would not be surprised to see interest in equal weighted international products.

EXPENSES

Expenses are one of the key investment dimensions that investors can control, and that produce important long-term impact on total return.  The three index weighting approaches generate different levels of expense in different ways.

Expenses include not just sales loads, management fees and administrative expenses, but also the cost of trading (higher in high turnover funds) and market impact (if the size of trades by the fund moves the market when buying or selling), and the tax cost associated with distributions, including any realized gains that were unrealized when you bought into the fund.

The study done by Srikant Dash and Keith Loggie of Standard and Poor’s in 2008 analyzed the performance of the S&P 500 and the S&P 500 EWI. Since 1990, the EWI has outperformed by 1.5% per year, but not consistently. It lagged the S&P 500 for six consecutive years from 1994 to 1999, but outperformed  the S&P 500 for seven years through 2006. The study suggested that the equal weighted index appeared to underperform the S&P 500 during strong markets, but held up better during bear markets. The result also suggest that when value stocks outperform growth stocks, the equal weighted index will outperform the S&P 500.

Note that since that study, through 2006, we have a current bear market in which the performance of the EWI is also superior on a 52-week and ultra short-term basis (as of September 12, 2008).  You can monitor if that continues to be true weekly here.

OUR VIEW

We think equal weighting is an interesting approach that deserves serious consideration on a portfolio-by-portfolio basis.

The period of EWI underperformance from 1994 - 1999 coincides with the dot.com bubble era, which is an unusual period that favored momentum companies.  We don’t appear to be facing a similar bubble in the US markets any time soon.

We wonder since equal weighting of stocks has outperformed market cap weighting, whether equal weighting of sectors would outperform too. And, if we go that far, what if a portfolio had equal weighting of sector funds that themselves are equal weighted for the stocks within them. We don’t know the answer. Perhaps, we’ll develop the data someday, but we hope someone saves us the effort and does it first.

Comments

[bigmoney:]

I deployed into equities, using equi-weight, equi-sector for convenience. I much prefer doing that, than getting into mutual funds / ETFs that are top-heavy financials. This likely would have been a much better formula for getting through the 2000 tech bubble. Going into a bubble, its very difficult to determine if "this time is different", so the equi helps me sleep better at night.

Unfortunately the deflating credit bubble seems to be taking down all stocks across the board, so its hard to pick "winners". The only refuge seems to be federal reserve notes, backed by faith & trust and a very low interest rate.

2008 Sep 17 01:51 PM