Domestic small cap stocks are a popular investment choice set in retirement plans, and many long-term investors have a strategic allocation to this asset class. After all, small cap stocks have outperformed their large cap brethren over long-time intervals, and can enhance long-term portfolio returns for investors with a tolerance for small cap's higher risk. There are two common domestic small cap benchmarks - the Russell 2000 (IWM) and the S&P SmallCap 600 (IJR). This article discusses why one of these indices has outperformed over the trailing twenty years. Note: It is not the index that has an additional $10 billion of assets under management in its related ETF, signaling that there is a broad universe of investors who could benefit from reading this article.
While the return profile of two of the broadest large cap benchmarks - the S&P 500 (SPY) and Russell 1000 (IWB) - have been highly correlated (Figure A below), the cumulative returns of the two most oft used small cap benchmarks - the S&P SmallCap 600 and Russell 2000 (Figure B) have diverged sharply over time.
Figure A: Cumulative Return - Russell 1000 vs. S&P 500
(click to enlarge)
Figure B: Cumulative Return - Russell 2000 vs. S&P SmallCap 600
The S&P SmallCap 600 outperformed the Russell 2000 by 177bps per annum from 1994-2012. Over that time frame, a dollar invested in the S&P index would have totaled a cumulative figure forty-seven percent higher than that of the Russell 2000. Part of the reason for this divergence is the difference in index mechanics between the two indices. The Russell indices are reconstituted each June with the top 1000 companies by market capitalization joining the Russell 1000, and numbers 1001-3000 joining the Russell 2000. The S&P SmallCap 600 uses less of a mechanical approach. In the S&P index, constituents must be profitable (four consecutive quarters of positive earnings), liquid (30% of shares trading annually), and have at least half of their shares publicly floated.
The July Effect
The rebalancing of the Russell indices each June leads to a market anomaly. Because the Russell index changes are rules-based and predictable, have featured large number of index changes annually (roughly one-quarter historically), and given the level of indexing to the Russell 2000 is so high, arbitrageurs are able to bid up the prices of future constituents who will have increased future sponsorship and sell/short exiting constituents before indexing participants are forced to sell these now off-index positions. In a 2005 paper by Honghui Chen of the University of Central Florida, Gregory Noronha of Arizona State University-West, and Vijay Singal of Virginia Tech entitled "Index Changes and Unexpected Losses to Investors in S&P 500 and Russell 2000 Index Funds," the authors estimate that the performance drag related to this rebalancing is between 1.30% and 1.84% annually. Comparatively, the paper estimates 0.03% to 0.12% is lost due to index changes in the S&P 500. Given the return differential between these two small cap indices has been 177bp historically, this performance drag could account for the entire return differential.
An examination of the monthly returns of the two indices demonstrates that the Russell 2000 has indeed lagged the returns of the S&P SmallCap 600 in July likely due to selling pressure on constituents who were deleted from the index in the preceding June. Roughly forty percent of the annual relative underperformance of the Russell 2000 has occurred in July, a result that is statistically significant.
Source: Bloomberg, Standard and Poor's, Russell
The Earnings Requirement
The S&P SmallCap 600 requirement that constituents be profitable for four trailing quarters before inclusion produces the second largest source of differentiation between the two indices. In Standard and Poor's September 2010 small cap research piece "A Tale of Two Benchmarks," the index provider detailed a study that tested the impact of this earnings criteria. The population was companies with a market capitalization between $250 million and $2 billion from year-end 1993 to year-end 2009. This set was then subdivided into two groups - those that successfully met the earnings threshold and those who did not. Companies that met the earnings requirement outperformed those that failed by 531bp over the twelve month forward holding period. The higher return of the S&P SmallCap 600 is likely due in part to the fact that there are less firm failures than amongst the Russell 2000 constituents, which include more unprofitable firms. An examination of the current multiples of the two indices below demonstrates the higher quality nature of the small cap constituents in the Standard and Poor's index.
The Russell 2000 is trading at a fifty-two percent premium to the S&P SmallCap 600 as a ratio of trailing earnings. This large disconnect is due to the much higher level of negative earnings in the Russell 2000, which as discussed, is a direct result of the differing constituent inclusion rules.
The S&P SmallCap 600 has outperformed the Russell 2000 since inception, and has done so with less variable returns. The annualized standard deviation of monthly returns for the S&P SmallCap 600 has been roughly 1% less than the Russell 2000 annualized. Holders of the S&P SmallCap 600 have gleaned alpha since 1994, generating higher average returns with less variable returns.
The wealth transfer from those who index to the Russell 2000 to arbitrageurs who profit from the annual rebalancing has led to a significant performance drag. The S&P's requirement that firms be profitable in each of the four trailing quarters prior to induction leads to a higher quality constituent base, and research has proven that unprofitable small cap firms have weighed on the Russell 2000 results historically. For Seeking Alpha readers looking to gain exposure to small cap domestic stocks, replicating the oft overlooked S&P SmallCap 600 should provide meaningful outperformance relative to the Russell 2000 over forward periods.