The past year was a stellar one for almost all global equity markets, but it was the world’s emerging economies that truly stole the show. As investors regained their appetite for risk, those markets that have historically been deemed to be the most risky surged. But the impressive gains for emerging markets in 2009 weren’t solely attributable to investor perceptions: developing economies have clearly taken the lead in the global recovery efforts, bouncing back from the recent recession and now expanding at impressive rates.
For investors reallocating portfolio assets from developed economies to emerging markets, ETFs were clearly a popular choice. The 35 ETFs included in our emerging markets equities ETFdb Category saw aggregate cash inflows of more than $17 billion, a portion of which came out of domestic equity funds.
While there are a number of ETF options that offer emerging markets exposure, the majority of assets are concentrated in two funds: the iShares MSCI Emerging Markets Index Fund (EEM) and Vanguard Emerging Markets ETF (VWO), which have market capitalizations of about $41 billion and $20 billion, respectively. The parallels between these funds go far beyond the similar names: both track the MSCI Emerging Markets Index, a free float-adjusted, market capitalization benchmark that includes equities from nearly two dozen developing economies. ETFdb Pro members can read more about emerging markets investing in the ETFdb Category Report (if you’re not a Pro member yet, sign up for a free trial or read more here).
While investors in EEM were no doubt pleased with the overall results in 2009, many have been bothered by a perplexing question: why did VWO, which maintains an identical investment objective, gain 83.0% on the year, while EEM added only 75.2%? It’s hard to complain about a 75% gain, but it’s also difficult to watch a direct competitor leave you in its dust.
Part of the surprisingly large performance gap between VWO and EEM is easy to explain. Vanguard’s emerging markets ETF charges an expense ratio of 0.27%, while EEM charges 0.72%. This means that if the two ETFs maintained the exact same basket of underlying stocks throughout the year, VWO should outperform EEM by about 45 basis points. For relatively small short-term investments, this difference in costs may be immaterial. But when compounded over a longer period, the difference in fees begins to add up. Assuming a ten year holding period and an annual pre-expense return of 10%, the gap in expense ratios can translate into a 10% difference in bottom line results.
So part of the gap between EEM and VWO can be explained by varying expense ratios. But a 0.45% gap in costs only accounts for a small portion of the nearly 8% difference in returns in 2009. So what’s to blame for the rest?
EEM’s Tracking Error
While EEM and VWO appear almost identical on the surface, a look under the hood of these ETFs reveals that they’re not nearly as similar as they seem. According to the EEM fact sheet, as of September 30 the index underlying the fund has about 750 components. But at that point EEM held only 372 stocks, less than half the companies included in the index it purports to track. EEM now holds about 435 individual holdings.
EEM uses a sampling strategy to achieve its investment objective, meaning that the ETF generally won’t hold all the components of the underlying index, but rather will attempt to construct a smaller portfolio that will closely correspond to the results of the complete benchmark. Because the index underlying EEM is a cap-weighted benchmark, the overlap between the fund’s holdings and those of the index was much greater than 50% as of September. EEM held most of the companies that have the largest weighting in the index, and passed up some of the smaller constituents.
According to the iShares Web site, EEM returned about 6.7% less than its benchmark during 2009, a huge amount of tracking error for a single year. This implies that the components of the MSCI Emerging Markets Index that aren’t found in EEM outperformed those that are included by a wide margin over the last year. Gains of 80% or more in several of these components magnified the gap considerably.
It should be noted that the tracking error for EEM in 2009 appears to be an aberration. Between its inception in April 2003 and the end of 2008, EEM had actually outperformed the MSCI Emerging Markets Index by almost 11%. The implied tracking error of about 0.90% per year is still high, but is far below 2009 results.
VWO’s Full Replication Plus
The alternative to the sampling strategy of EEM is full replication of the underlying index — essentially holding every component of the benchmark is the same allocation. But VWO went even beyond full replication in 2009. As of September, this ETF had 790 individual holdings, or about 45 more than the MSCI Emerging Markets Index. Recently, this number had climbed above 820.
|Top Ten Holdings|
|Taiwan Semiconductor||3.1%||Gazprom OAO||2.0%|
|Pertoleo Brasileiro||2.6%||Pertoleo Brasileiro||1.4%|
|Pertoleo Brasileiro||2.3%||Teva Pharmaceutical||1.4%|
|Posco||2.2%||America Movil de CV||1.4%|
|KB Financial Group||1.9%||Pertoleo Brasileiro||1.3%|
|China Mobile||1.9%||China Construction Bank||1.2%|
|Chunghwa Telecom||1.8%||Vale SA||1.2%|
|Banco Bradesco||1.8%||I & C Bank of China||1.2%|
|Total Top Ten||24.6%||14.7%|
So during 2009, VWO held more than twice the number of stocks EEM did, and even invested in equities that weren’t included in its underlying index.
Since EEM had far fewer holdings than VWO, it makes sense that the iShares fund would generally have a higher concentration in the shares that it does hold. But, as shown in the adjacent table, the top ten holdings of these ETFs are very different, both in terms of the companies covered and the allocations given to each.
Through September, the top ten holdings of EEM accounted for 24.6% of total assets, nearly 10% more than the ten biggest holdings in VWO.
But the differences go far beyond the allocations. Anyone glancing at the adjacent table would likely be shocked to learn that EEM and VWO are linked to the same benchmark.
VWO and EEM aren’t the only funds that track identical indexes but maintain dramatically different holdings. The total bond market funds offered by the same issuers, AGG and BND, are both designed to track the performance of the Barclays Capital U.S. Aggregate Bond Index. AGG has about 275 individual holdings. BND has more than 4,100, and still includes less than half of the benchmark.
The material discrepancies between these funds shows some of the confusing issues surrounding ETF investing and highlights the importance of diligent research into potential investments. “Buying an ETF is like buying a used car,” says Andy Hagans, principal at Carolean Capital Partners. “You should always take a long, hard look under the hood before driving off.”
For more on emerging markets investing, visit the Emerging Markets ETF Center.
Disclosure: No positions at time of writing.