August 29, 2014

Comparing U.S. Small Cap Stock Indexes

It generally looks like selection of a particular large cap index doesn't matter much. The top holdings are pretty much the same across each, and having 500 or 750 stocks in a large cap index doesn't mean much when the top 100 make up well over 50% of the value of all of them.

So how about small cap indexes? Does the same conclusion hold true when we look at the smaller end of the market? Large caps can be valued well in to the hundreds of billions, whereas their smaller counterparts average just a tiny fraction of that value. For a little perspective, the Russell indexes, from mega cap to micro cap, cover $5.2 trillion in benchmarked assets. The Russell 3000, the largest 3,000 U.S. stocks, cover 98% of that value. The Russell 2000 small cap index is 9% of that 98%, so around 8.8% of the investable U.S. market. The total market cap of U.S. listed companies is over $18.6 trillion. With Apple on its own equally about 3.5% of the total S&P 500, it would only take about three Apples to equal the entirety of the 2,000 companies in the Russell 2000 (assuming my math is right).

To give you an idea of how big a typical Russell 2000 company is, the weighted average company value in the index is worth $1.7 billion, with an overall median of just under $700 million. The largest companies in the index, currently Puma Biotechnology and InterMune Inc., are each worth nearly $8 billion, but overall the constituents in the Russell 2000 are worth between $500 million and $2 billion.

While the Russell 2000 is the most well-known of all U.S. small cap indexes, it's not the end-all be-all and a number of ETFs are based on other small cap indexes, not limited to:
  • S&P 600
  • CRSP US Small Cap Index
  • Dow Jones U.S. Small-Cap Total Stock Market Index
There are others, including the Defined Small Cap Core Index from S&P which is a subset of the S&P 600, or small-mid indexes like the FTSE RAFI US 1500 Small-Mid Index. But the Russell 2000 and S&P 600 are by far the biggest names. I included the CRSP and Dow Jones indexes I named above primarily because the Vanguard Small-Cap ETF (VB) and Schwab U.S. Small Cap Fund (SCHA) are based on each index respectively and round out the four largest small cap ETFs, each with billions in assets.

How does each index differ?

  • The Russell 2000 is made up of the bottom 2,000 stocks from the Russell 3,000 index (in other words, excluding the 1,000 largest public companies in the U.S.)
  • The S&P 600 covers about 3% of the total U.S. investable market, in contrast with the S&P 500 covering the large segment, S&P 400 covering mid caps, and combined the three mutually exclusive indexes make up the S&P 1500. The S&P 600 is interesting in that it uses some basic criteria to screen for eligible companies, including a >50% public float, consistent profitability, and a minimum threshold of liquidity. According to the S&P site, the median market cap is just over $1 billion, and average is mean is nearly $1.2 billion.
  • The CRSP U.S. Small Cap Index, like its larger cousins, selects constituents based on a percentile range, in this case companies that range from the bottom 2% to 15% of market cap. The Vanguard Small-Cap ETF (VB) is based on the CRSP flavor and holds 1,465 stocks, with a median market cap of $3.1 billion.
  • Lastly, the Dow Jones U.S. Small-Cap Total Stock Market Index is made up of stocks ranked 751-2500 in the Total Stock Market Index. This is interesting, because as far as I can tell, the Dow Jones U.S. Large-Cap Total Stock Market Index (ranked 1-750) is mutually exclusive to the small cap variant, but the mid-cap version of this index uses stocks ranked 501-1000, so there appears to be overlap in the mid segment. The other three indexes seem to be more distinct within their index families.
Knowing how each index differs is interesting, but how do the four largest small cap blend ETFs (one based on each index) perform? The iShares Russell 2000 Index Fund (IWM) is the largest small cap ETF, followed by the iShares Core S&P Small-Cap ETF (IJR), then the Vanguard Small-Cap ETF which uses the CRSP index, and finally the Schwab U.S. Small Cap Fund (SCHA) based on the small cap slice of the Dow Jones U.S. Small-Cap Total Stock Market Index.

Total returns from each, assuming dividend reinvestment, are below. The date range is slightly under five years, based on the inception date of the younger Schwab fund.

U.S. Small Cap ETF Total Returns - IWM, IJR, VB, SCHA

The results are interesting. Large cap indexes all seem to track quite closely to one another, and three of the four small cap indexes behave very similarly. From November, 2009 to the beginning of August, 2014, IJR, VB, and SCHA all ended up within a couple percent of each other, more than doubling, up around 130%. But IWM, based on the Russell 2000 ended up a bit over 114%. Spectacular performance, but noticeably behind the others.

The Russell 2000 reconstitution effect appears to be a main cause of the index's under-performance. The index is a bit more strict and rules-based compared to the others. As a result, traders try to anticipate additions and deletions from the index and trade in advance to take advantage of this predictability, potentially dragging down returns to the tune of around 1% per year. I've seen other sources saying the effect could be even great - in the neighborhood of 2%, but don't quote me on that.

There were apparently changes to the Russell 2000 in 2007 that partially addressed this arbitrage issue, but as you can see in the chart above, it has still lagged its fellow small cap indexes since then. As the most widely-followed small cap index, it likely draws the most attention from traders trying to game the system. Additions and deletions from funds based on the less famous indexes may not move the market as much, and consequently might not be as attractive to try and game. This leads me to wonder, is it potentially better to target ETFs based on less well-known indexes for your portfolio?

No comments:

Post a Comment