The third and final installment of Larry Swedroe’s analysis of the small-cap premium.
Small-cap growth stocks have been euphemistically called “The Black Hole of Investing.” The reason is that they have produced the lowest returns of any of the major
The following table shows both the annualized returns and the annual standard deviations for the period 1927-2012:
|Asset Class||Annualized Return (%)||Annual Standard Deviation (%)|
|Fama-French US Small Value Index (ex-utilities)||13.6||34.9|
|Fama-French US Small Cap Index||11.8||30.4|
|Fama-French US Large Value Index (ex-utilities)||10.2||27.0|
|Fama-French US Large Cap Index||9.6||19.9|
|Fama-French US Large Growth Index (ex-utilities)||9.1||21.7|
|Fama-French US Small Growth Index (ex-utilities)||8.8||33.7|
Basically, the data lines up the way we would expect—the riskier the asset class (using volatility as the measure of risk), the higher the returns, with one really glaring exception.
Small-cap growth stocks should produce higher—not lower—returns than large-cap growth stocks and large-cap stocks in general. Yet they’ve produced the lowest returns, and did so while exhibiting much higher volatility. In fact, while other subsets of small-cap stocks have outperformed large-caps, the performance of small-cap growth stocks has been so poor that the small-cap premium can disappear completely.
For example, their performance has been so poor that for the 44-year period 1969-2012, U.S. small-cap growth stocks returned just 6.2 percent, underperforming small-cap stocks by 4.2 percent a year, and small value stocks by 6.6 percent a year, and even underperforming long-term Treasury bonds (20-year) by 2.4 percent a year and five-year Treasurys 1.5 percent a year. Like the momentum effect, this is an anomaly and a problem for those who believe in market efficiency.
Now let’s take a look at the results in the post-Banz era, specifically the 21-year period 1982-2002 when, as we showed, large stocks outperformed small stocks. During this period, while the CRSP 1-5 Index returned 12.75 percent, the stocks within the Fama-French small growth (ex-utilities) index returned just 5.95 percent. However, the stocks in the Fama-French small value (ex-utilities) index returned 14.78 percent, outperforming large stocks by more than 2 percentage points a year.
We now turn to the results from a recent study by Dimensional Fund Advisors. DFA sorted stocks independently on the book-to-market (BtM) ratio scale, the earnings-to-price (E/P) ratio scale and cash flow-to-price (C/P) ratio scale, and defined extreme small-cap growth stocks as small-cap stocks that are in the bottom 25 percent of the small-cap market by BtM as well as in the bottom 25 percent of the small-cap market by either E/P or C/P.
For the period 1975-2011, while small-caps stocks returned 16.7 percent, the most extreme small-cap growth stocks returned just 8.5 percent. Excluding these stocks, which represent only 12 percent of the small-cap universe, the return to small-caps would have been 17.6 percent. Thus, while sacrificing little in the way of diversification (maintaining exposure to 88 percent of the small-cap market), returns would have increased by 0.9 percentage point.
The evidence was similar in international markets. For the period 1995-2011, small-caps returned 7.2 percent. The most extreme small-cap growth stocks (about 12 percent of the overall international small-cap market) returned just 1.1 percent. Excluding these stocks would have raised the return of small-caps to 8.1 percent. The results were similar in every country—the most extreme small-cap growth stocks produced the worst returns.
Before summarizing the evidence, there’s one more important point we need to cover.
Investors who expose their portfolios to the size factor benefit from a diversification effect.
From 1927 through 2012, the annual correlation of returns between the size factor and the market factor (beta) was just 0.41, and the correlation with the value (or price) factor was an even-smaller 0.10. The low correlation with the other factors helps to dampen the volatility of the overall portfolio. We’ve already seen how asset classes and factors can underperform for long periods—that’s the nature of risky assets. With that in mind, the following is a good example of the benefits of diversification across factors and asset classes.
For the period 1984-1990, the CRSP 6-10 underperformed the CRSP 1-5 by more than 9 percentage points a year, returning just 4.8 percent per year versus 14.0 percent. However, during the same period, the Fama-French International Small Index outperformed the MSCI EAFE Index by 6.5 percentage points per year—28.1 percent per year versus 21.6 percent per year.
The small premium seems alive and well, especially if we exclude the extreme small growth stocks from portfolios. We see further proof of this in looking at the data on three live DFA small funds. The data is for the period since the fund’s inception through June 2013. The S&P 500 is used to represent the returns to large stocks. I also provided the returns of the other DFA funds for the same periods.
- The DFA Micro Cap fund (DFSCX) inception was January 1982.
1982-2012: DFA Micro Cap 12.1 percent versus S&P 500 11.4 percent.
- The inception of the DFA Small Cap fund (DFSTX) was April 1992.
April 1992-2012: DFA Small Cap 10.7 percent and DFA Micro Cap 11.6 percent and 8.9 percent for the S&P 500.
- The DFA Small Value fund (DFSVX) was April 1993.
April 1993-2012: DFA Small Cap 10.2 percent, DFA Micro Cap 11.4 percent, DFA Small Value 12.3 percent, and 8.6 percent for the S&P 500 Index.
The bottom line is that if there is any story about the disappearance of the small premium, it’s really a story about the small growth anomaly, not about small stocks in general. Another takeaway is that you can improve the expected performance of a portfolio while still maintaining broad diversification within the small-cap space by excluding extreme small growth stocks. This is another case where passively managed “structured portfolios” can add value over pure indexing strategies (which include all stocks within their benchmark index to avoid tracking error).
Larry Swedroe is director of Research for the BAM Alliance, which is part of St. Louis-based Buckingham Asset Management.