Small capitalization stocks are often seen as providing two major benefits: diversification and greater returns. This article will focus on small capitalization ETFs from around the world to see if they do provide these benefits. This review will not be exhaustive since there are countless small capitalization ETFs, let alone the additional funds and individual securities that a casual investor can use. The following ETFs were considered:
|SPY||SPDR S&P 500 Trust ETF||81,230||U.S.||Large Cap Index|
|IJR||iShares S&P SmallCap 600 Index||5,550||U.S.||Small Cap Index|
|PDN||PowerShares FTSE RAFI Dev Markets ex-US Small/Mid||63||ex - U.S. Developed||Small Cap|
|LATM||Market Vectors LatAm Small-Cap Index ETF||15||Latin America||Small Cap|
|ILF||iShares S&P Latin America 40 Index||1,560||Latin America||Large Cap Index|
|BRF||Market Vectors Brazil Small-Cap ETF||595||Brazil||Smalll Cap|
|EWZ||iShares MSCI Brazil Index Fund||9,110||Brazil||Large Cap|
|VBK||Vanguard Small Cap Growth ETF||6,970||U.S.||Growth|
|VBR||Vanguard Small Cap Value ETF||5,670||U.S.||Value|
|XLU||Utilities Select Sector SPDR||6,600||U.S.||Large Cap Utilities|
|PSCU||PowerShares S&P SmallCap Utility ETF||43||U.S.||Small Cap Utilities|
|EWX||SPDR S&P Emerging Markets Small||909||Emerging||Small Cap|
|EEM||iShares MSCI Emerging Markets Index||27,470||Emerging||Large Cap|
|JSC||SPDR Russell/Nomura Small Cap Japan ETF||99||Japan||Small Cap|
|EWJ||iShares MSCI Japan Index Fund||6,620||Japan||Large Cap|
|GDX||Market Vectors Gold Miners ETF||8,950||Various||Large Cap|
|GDXJ||Market Vectors Junior Gold Miners ETF||2,090||Various||Small Cap|
Some additional larger capitalization ETFs from different sectors and geographies were also included to provide a comparison with respect to the diversification benefits. The benchmark portfolio will be SPDR S&P 500 Trust ETF (SPY). The first analysis was to check correlations to see if small cap ETFs indeed offered lower correlations than corresponding large capitalization ETFs. I reviewed this over several different time frames since some ETFs had more limited history.
|Ticker||15 Month Correlation||24 Month||36 Month|
Source: Author Calculations
In most cases, it appears that the small cap ETF version had slightly but not significantly lower correlations. The two counter examples were for gold mining (GDX, GDXJ) and utilities (XLU, PSCU). The counter example to being just slightly lower was Japan where JSC had a 40-60% correlation while EWJ had a 65-80% correlation over the different time frames.
I would not place much, if any, emphasis on the magnitude of the difference, but rather note that in almost all cases the smaller cap variation showed a lower correlation to SPY. However, does this help with portfolio diversification? While the correlation might be lower, if the volatility is substantially higher than the overall portfolio, volatility might increase. The next table looks at volatilities over two time frames.
|Ticker||15 Month||36 Month|
Source: Author Calculations
This table largely confirms intuition that small cap ETFs have higher volatilities than there larger cap counterpart. However, this is again not true for Japan where JSC had lower volatility than EWJ and it also was not consistent for emerging markets in looking at EWX and EEM. These two tables make it evident that except for Japan, the addition of small cap ETFs will not lower portfolio volatility relative to their larger cap counter parts. The slight correlation benefit is trumped by the often substantially larger volatility.
One additional comment might be around looking at the correlations between the small cap ETF and large cap ETF. While it is very solid to compare the small cap U.S. ETFs to SPY, does it make as much sense to compare both BRF and EWZ to SPY. What about the correlation between BRF and EWZ? The following table answers these questions and shows that the two variants are usually highly (90+%) correlated. However, XLU and PSCU show a relatively lower correlation, which I will address later.
Source: Author Calculations. Time frames were varied to use at most 36 months or whatever history was available.
The final question is to look at returns to see how the ETFs compare. I pulled returns over a few different time frames. Again, unfortunately, not all ETFs had sufficient history for each time frame.
|Ticker||March 2009- present||12 month||6 month|
Source: Author calculations
The synthesis here is that the small cap version tends to provide more extreme returns - greater declines and greater increases. The one exception was JSC which outperformed EWJ in all three time frames.
Small Cap ETFs Are For Returns, Not Diversification
The rationale for purchasing smaller cap ETFs is for return benefits, not diversification. This appears to be quite true in U.S. markets and probably extrapolates to other developed markets. It seems to be true in many developing markets as well. It is also clear that small cap ETFs often carry more risk as noted by their higher volatilities and more extreme returns.
However, there seem to be exceptions to the rule. First, there does not seem to be a compelling reason to purchase PSCU over XLU. This analysis shows it to have a higher risk profile with worse returns. PSCU also has relatively few assets and is thinly traded. Furthermore, it tends to be more focused on gas utilities than electric or integrated which dominate XLU.
For investors interested in Japan with a relatively passive approach, it seems that one should definitely look at small cap ETFs as well as and possibly to the exclusion of large cap ETFs. While it is not clear why they should be superior on both return and risk, the numbers support it. One caution is that it is smaller and more thinly traded.
Finally, EWX could be an interesting option for investors looking at emerging markets. While it is more thinly traded than EEM, which is highly liquid, it still has reasonably volume (better than either PSCU or JSC).
Disclaimer: This article is for informational and educational purposes only and shall not be construed to constitute investment advice. Nothing contained herein shall constitute a solicitation, recommendation or endorsement to buy or sell any security.