The weighting methodology for an ETF can have a profound difference in the structure of its portfolio and in its performance. There are various ways the same portfolio of stocks can be weighted in an ETF. I want to lay out an overview of some of the most common, and try to develop some insights into how weighting tactics can affect the ways that a fund is structured.
As I go through this exercise here in Part 1 of the series, I am going to emphasize large-cap funds for each weighting style. My reason for this is not because I think the large-cap funds are the most interesting or have the most potential for investors who are interested in exploring alternative weighting tactics. Just the opposite, in fact. I think that weighting tactics have more impact on investor returns in categories other than the large caps. The reason for focusing on large caps in this article is that I think it's easier to get a sense of the differences and similarities the tactics bring to portfolio structure by using familiar names from the top of the market which is the point I want to focus on here In Part 1.
For part 2, I plan to zoom out to cover the fuller range, putting more emphasis on performance differences and asset classes. If the subject interests you, you'll really want to look for Part 2.
I will be covering 5 styles: Market Capitalization Weighted, Equal Weighted, Dividend Weighted, Earnings Weighted and Revenue Weighted. Let's begin with a couple of tables summarizing the composition of the various Large Cap funds from each of the weighting categories I'll discuss.
Market Capitalization Weighted
I haven't actually analyzed all ETFs, but I strongly suspect that Market Cap weighting is, by far, the most common, so let's start there. The S&P 500, perhaps the most widely recognized index, is cap-weighted. So are S&P's mid-cap (S&P Mid Cap 400) and small-cap (S&P Small Cap 600). The Russell indexes, benchmark standards for the broad market, are also cap weighted. The Russell 3000E index includes the 4000 largest US stocks by market capitalization. This index is sliced and diced in any number of ways to generate the Russell 3000, the top 3000 of those 4000 stocks; the Russell 2500, the bottom 2500, thus a mid to small cap index; the Russell 1000, the large caps comprising the top 1000, and so forth. There are, of course, ETFs that attempt to track each of the Russell indexes. All of them are, necessarily, market cap weighted.
What precisely is a market cap-weighted ETF? As the name implies, holdings are weighted by their market capitalizations. Sticking to the familiar names that comprise the mega-cap universe, a cap-weighted S&P 500 fund will have its largest holding in XOM, the largest of the large-caps, followed by AAPL, MSFT, and so on. Such funds have the reputation of being extremely top heavy. Cap-weighted SPDR S&P 500 ETF (SPY), for example, has a portfolio of 500 stocks where the top 10 account for 18% of the fund's total assets under management (AUM) and the top 50 (10% of the holdings) account for about 50% of the fund's AUM. While it does seem excessive, as the table above shows, this is not an unusual situation for several of the weighting tactics. Three of the four other styles have even higher percentages for their top 10 holdings in the large cap funds.
For mid-caps, the SPDR S&P 400 MidCap ETF (MDY) has 7.4% of its 400 holdings in its top 10. And the comparable small-cap fund, SPDR S&P 600 Small Cap ETF (SLY), has 5.26% of its holdings in the top 10. SLY, the least heavily weighted toward its largest holdings at 5.26% for the top 10, has its 10 largest holdings weighted at 3.2 times what equal weighting would produce. For the large cap SPY, it's a factor of 9.
In virtually every asset class, the dominant ETFs are cap-weighted. Emerging Markets? VWO with $41B AUM and EEM with $32B. EAFE? EFA, $34B AUM. Russell 2000? IWM, $13B AUM.
Cap weighting has its advantages. Such funds provide investors with a representative piece of the market. But that slice is skewed toward the largest sectors such as Technology and Financial for the large cap funds. They tend to perform better in momentum-driven markets, but less well under other conditions. Cap-weighted funds also tend to be cheaper and easier to manage. This is the style of the classic passive index fund.
The second most common weighting tactic has to be equal weighting. As is obvious from the name, each holding receives an equal weight in the fund on a dollar basis.
Returning to the widely recognized large caps to facilitate comparisons, let's look at Guggenhiem's S&P 500 Equal Weight ETF (RSP). The fund's target is to have 0.2% (1/500) of its AUM in each individual holding. Obviously, between rebalancings high-flyers will move above that level and laggards will fall below. Currently their number 1 holding is ETFC with 0.28% of AUM. At the time of this writing, none of SPY's top 10 holdings make it into RSP's top 10.
Guggenheim Investments dominates the equal-weight ETF space where they offer such funds in multiple asset classes. In addition to the S&P 500 fund, there are funds covering Emerging Markets (EWEM) plus the Russell 1000 (EWRI), MidCap (EWRM), and 2000 SmallCap (EWRS) indexes. There is also an equal-weighted fund for each of the S&P 500 sectors. Other sponsors offer equal-weighted funds as well. It's becoming a common style for certain specialized portfolios, such as the shareholder value (SYLD) and buyback (PKW, TTFS) ETFs.
Characteristics of equal weighting include a disciplined selling of winners and buying of losers, in effect an institutionalization of a contrarian "sell high; buy low" philosophy. They tilt toward smaller cap stocks when compared to cap-weighted funds. They also tend toward a more balanced sector allocation, something that is more obvious in the funds with larger number of holdings. One fund has taken this approach as its guiding principle: ALPS Equal Sector Weight ETF (EQL) tracks equal weights in each of the nine S&P sectors.
Equal-weight funds require more intensive management due to their disciplined, rules-based rebalancing approach. This leads to higher fees. And, their tilt toward smaller cap sizes tends to make them more volatile than their cap-weighted counterparts.
Dividend weighting offers an interesting spin on the structuring of an ETF. Here the focus is on dividends and the weighting may take one of two forms: Weighting by cash dividends paid, or the less common weighting by dividend yield.
One obvious difference between dividend-weighted funds and the others I've discussed is the limited universe the dividend-weighted funds draw from. An S&P 500 dividend-weighted fund is limited to those S&P 500 stocks that pay dividends. Note that not all "dividend index" funds are dividend weighted. The best known, including SDY, VIG and VYM are cap-weighted funds.
WisdomTree dominates in the dividend-weighted fund space with their Total Dividend (DTD), MidCap Dividend (DON), SmallCap Dividend (DES), Global ex-U.S. Dividend (DLN), and several international funds.
The most direct comparison to the large-cap S&P 500 funds examined above is DTD. It holds 931 positions. For this fund, the top 10 slots account for 22.89% of the fund's AUM, greater than the cap-weighted SPY on a percentage basis. Much greater, in fact, when one considers that the fund has nearly twice as many positions in its portfolio. Sector allocations skew toward Consumer Defensive, Utilities and Real Estate. Dividend yield is near 3%.
Earnings weighing is certainly not common. The only earnings weighted ETFs I am aware of come from WisdomTree. (If you're starting to get the idea that WisdomTree is a bit more innovative than other fund sponsors, you may be right.) Once again, the name says it all. These funds are weighted by company earnings in December.
Clearly there is a close correlation between earnings and market cap - size matters. So, unsurprisingly, there are familiar names heading the list of the large-cap fund. One important distinction between earnings weighting and market-cap weighting is that companies may have similar sizes, but generate quite different earnings. Some generate no earnings at all. Obviously, those do not find a home in the earnings-weighted universe. The focus on earnings tends to tilt the funds toward value stocks.
WisdomTree replicates its domestic fund lineup for dividend weighted fund in this category. There are funds that cover the total market (EXT), the top 500 (EPS), MidCaps (EZM), SmallCaps (EES) and LargeCap Value (EZY).
The Earnings 500 Fund's portfolio of 500 stocks is headed by familiar names: XOM, AAPL, CVX and MSFT. The top 10 comprises 23.8% of the portfolio. Worth noting that when compared to SPY's top holdings, EPS drops JNJ, GOOG and PG, and picks up JPM, WMT and IBM in their place. The sector mix skews toward Financials and Technology.
Revenue weighting focuses the fund on topline revenues. This category is the province of RevenueShares, the sponsor of revenue-weighted ETFs in the LargeCap (RWL), MidCap (RWK), and Small-SmallCap (RWJ) categories as well as a Financial Sector (RWW) fund. They draw their pool of companies from those that make up the S&P 500 (large), S&P 400 (mid), and S&P 600 (small) indexes. This makes it very straightforward and informative to do a direct comparison to funds that use the same indexes with different weighting tactics. See my article here for just that analysis.
Revenue weighting shifts the mix strongly toward lower-margin businesses and, to some extent, toward high-debt companies. The strategy leads to rebalancing from high P:Sales to low P:Sales.
The top 10 names in the large cap fund's holdings include some of the same names we've seen above. Energy giants XOM and CVX are here, as are AAPL, BRK.B, GE from SPY, but the others are new: WMT heads the list. Two more energy sector companies, PSX and VLO, appear along with F and CVS. None of these last four appear on any of the other top 10 lists. The top 10 account for 20.28% of the fund's 500 holdings. The sector allocations skew toward Consumer Cyclical and Energy at the expense of Real Estate and Technology.
I want to focus on performance across the spectrum in Part 2, but I can't leave this now without at least an overview look, so I'll close with a summation of the large-cap funds we've been following.
The revenue- and equal-weighted (RWL and RSP) funds had the best trailing twelve months based on stock price. Earnings- and market cap-weighted (EPS and SPY) filled in the middle about 600 bps behind the leaders. And the dividend-weighted fund brought up the rear another 400+bps behind.
Plotting 3 year return against 3 yr standard deviations shows that there is a volatility associated with the increased returns and that market cap weighting reduces volatility, at least in the case of these large cap holdings.
Notice in the table below that the non-traditional weighting styles come at some cost. Fees are higher by 3 to 4 fold compared to market cap weighted SPY. Part of this is the cost of managing funds that require more intensive management. This is evident from the turnover percentages, which are 4 to 6 times that of the SPY.
It's clear there are multiple alternatives to the default market-cap weighting. Changing weighting tactics produces quite different portfolios. The various portfolios may meet the needs of different investors more or less effectively. The portfolios also could provide better fits for certain market conditions. Momentum-driven market conditions would seem to favor market-cap weighting while more defensive conditions could favor dividend- or earnings-weighted funds.
The marked differences that the various weighting styles manifest are less clear in the large-cap sector I've focused on here. Remember, the purpose was not to see how the funds perform as much as it was to get a picture of the portfolios that result from each style. Part 2 should provide a more interesting picture of performance differences among these tactics.