Tracking Errors on US and International ETFs

April 27, 2011

The table below shows the tracking error of U.S. and international equity ETFs in 2010.

To make sure you understand these numbers in their proper context, see yesterday’s post about tracking errors on Canadian equity ETFs.

Fund Index Tracking
US equity (hedged) Ticker return return error
iShares S&P 500 XSP 13.5% 13.5% 0.0% (1)
TD US Index Currency Neutral – e TDB904 12.6% 13.5% -0.9%
RBC US Index Currency Neutral RBF558 12.6% 15.1% -2.5%
Altamira US Index Currency Neutral NBC856 13.0% 15.1% -2.1%
Claymore US Fundamental CLU 16.7% 18.3% -1.6% (2)
BMO US Equity ZUE 11.2% 11.8% -0.7%
BMO Dow Jones Industrial Average ZDJ 11.9% 12.8% -0.9%
Horizons S&P 500 Index HXS 6.7% 6.5% 0.2% (3)
US equity (non-hedged)
Claymore US Fundamental CLU.C 11.8% 14.1% -2.3%
TD US Index – e TDB902 8.4% 9.2% -0.8%
RBC US Index RBF557 8.0% 9.2% -1.2%
Altamira US Index NBC846 7.5% 8.3% -0.7%
International equity (hedged)
iShares MSCI EAFE XIN 4.6% 4.6% 0.0% (4)
TD Int’l Index Currency Neutral – e TDB905 4.0% 4.8% -0.8%
RBC Int’l Index Currency Neutral RBF559 3.3% 4.8% -1.5%
BMO International Equity ZDM 1.1% 5.7% -4.6% (5)
Claymore Japan Fundamental CJP -1.9% 0.4% -2.3% (6)
International equity (non-hedged)
Claymore International Fundamental CIE -0.2% 1.9% -2.1% (2)
TD International Index – e TDB911 1.7% 2.1% -0.4%
Altamira International Index NBC839 0.9% 2.3% -1.4%
Emerging markets equity
iShares MSCI Emerging Markets XEM 10.0% 12.7% -2.7%
BMO Emerging Markets Equity ZEM 11.9% 15.3% -3.4% (5)
Claymore Broad Emerging Markets CWO 16.2% 19.2% -3.1% (7)
Claymore BRIC CBQ 8.8% 11.6% -2.9%
Miscellaneous
Claymore Global Advantaged Dividend CYH 11.1% 13.9% -2.8%
Claymore Global Real Estate CGR 13.1% 14.9% -1.7%
iShares MSCI World XWD 5.7% 5.9% -0.2%

Notes:

1. The first four funds in this list all track the S&P 500 with currency hedging. However, you’ll notice that the iShares S&P 500 Index Fund (XSP) and TD’s U.S. Index Currency Neutral Fund report the index return as 13.5%, while the RBC and Altamira funds measure their performance against an index return of 15.1%. Why the difference?

The S&P 500 returned earned 15.1% in US dollars last year, and the latter two funds use this benchmark. However, XSP and the TD fund track the S&P 500 Hedged to Canadian Dollars Index, which factors in the currency hedging, reset once a month. That makes it a more realistic benchmark for the fund managers.

However, as S&P explains: “It is important to remember that since only beginning-of-period balances are hedged, the index does not assume a perfect hedging of currency movements.” That means Canadian investors should not expect the same returns from the S&P 500 that US investors enjoy. Indeed, here’s how XSP performed compared with the S&P 500 in US dollars over the last six years:

S&P 500 XSP
in USD in CAD Difference
2010 15.1% 13.5% -1.6%
2009 26.5% 23.0% -3.5%
2008 -37.0% -40.3% -3.3%
2007 5.5% 3.2% -2.3%
2006 15.8% 14.3% -1.5%
2005 4.9% 3.5% -1.4%

2. Claymore’s US Fundamental ETF (CLU) and International Fundamental ETF (CIE) had enormous tracking errors in 2009. Both improved greatly in 2010, but they still lagged their indexes significantly. Until recently the two funds held only a sampling of the stocks in their indexes. However, in 2010, both funds filled in those gaps, increasing their holdings from fewer than 400 stocks to about 1,000. Going forward, the tracking errors on these funds should be much smaller.

3. The Horizons S&P 500 Index ETF (HXS) was launched in late  November 2010, so this tracking error covers only a few weeks. However, the structure of HXS, which uses a swap to deliver the total return of the S&P 500, hedged to Canadian dollars and with no withholding taxes, should guarantee  that the tracking error will not exceed 0.5%. That takes into account its 0.17% MER and the additional cost of the swap.

4. The iShares MSCI EAFE Index Fund (XIN) tracks an index that has the currency hedging strategy built in, following the same principal as XSP (see note 1). The competing international index funds from TD and RBC track the MSCI EAFE index with returns measured in their local currencies. Here XIN’s benchmark really does make more sense, since the EAFE index includes many different currencies, so no investor could possibly achieve local returns in all of them.

5. BMO’s two international equity funds (ZDM for developed markets, and ZEM for emerging markets) both used representative sampling in 2010. The managers were not very successful in this respect, and the tracking errors were large. However, this will probably improve as the fund grows in size.

6. Claymore’s Japan Fundamental ETF (CJP) returned –1.88% in terms of net asset value, but if you held this fund in your portfolio, its market price went down an alarming –5.49%. It’s unusual for these two returns to differ so widely. What happened here was that the Claymore International Fundamental ETF (CIE) used to hold CJP for its exposure to Japanese stocks. But in April 2010, CIE sold off this holding—which was over $15 million—and started buying up the individual stocks (see note 2). The market impact of this move helped drive down the price of CJP.

In addition, later in the year CJP itself increased its holdings by about 100 stocks in order to replicate its index more closely. It seems reasonable to expect the funds tracking error to improve in the future.

7. Claymore’s CWO had the highest returns of the emerging markets ETFs in 2010. However, the outperformance was due to the ETF’s unusual strategy of hedging against the US dollar in order to match the returns of Vanguard’s Emerging Markets ETF (VWO). In this respect, it underperformed by more than 3%.

{ 14 comments… read them below or add one }

Greg April 27, 2011 at 12:26 pm

Any info on how the CIBC Emerging Markets index fund stacks up? I know CIBC’s MERs for investors under $200,000 is significantly higher than the MER for index funds offered by some of the other banks, but it is the only emerging market index mutual fund offered in Canada.

Canadian Couch Potato April 27, 2011 at 12:40 pm

@Greg: You can check the perfromance of any fund by getting the Management Report of Fund Performance from http://www.sedar.com/search/search_form_mf_en.htm

In this case, the fund returned 10.5% compared with the benchmark of 13%.

Jason April 27, 2011 at 1:26 pm

Very informative post. You did a lot of work! Thanks.

If a fund “sampled” the index as Preet suggested yesterday, that would be form of active management, wouldn’t it? Presumably, you’d pick the best or heaviest stock(s) in each sector. Even then, they can rarely beat the index! The whole idea of a “tracking” error would become moot.

Canadian Couch Potato April 27, 2011 at 1:38 pm

@Jason: Sampling is not active management; it’s just a way of dealing with a practical problem. If an index has 1,000 stocks, it may be cost-prohibitive to buy them all, so the managers choose the most influential, and they keep the overall sector weighting of the fund the same as the index. Their goal is still to match the index returns with the same risk, not to choose a sample they think will outperform.

The major bond indexes are literally impossible to replicate fully, so all the managers use sampling. The tracking error measures how well they did.

James April 27, 2011 at 2:41 pm

How do the US and international fundamental index funds like those from Powershares stack up against those from Claymore?

Chris April 28, 2011 at 2:28 am

Since you include Vanguard ETFs in a number of your model portfolios, I’m surprised that you haven’t included their tracking errors here.

Alan April 28, 2011 at 9:32 am

Great educational info. Would be great if you could add printer-friendly format for printing.

Canadian Couch Potato April 28, 2011 at 10:04 am

@James: I was unable to find annual reports showing the 2010 tracking errors for the PowerShares ETFs. However, the rolling perfromance numbers on their website suggest that they track the RAFI indexes better than Claymore’s ETFs, likely because PowerShares fully replicates the indexes and has greater economies of scale.

@Chris: Here are the 2010 tracking errors for the Vanguard ETFs (based on NAV). For VWO and VSS, the year-end is October 31:

VTI: -0.02%
VEA: +0.72%
VWO: +1.51%
VB: -0.07%
VTV: -0.09%
VSS: +1.02%

gsp April 28, 2011 at 2:03 pm

Notes 6 and 7 are reversed.

Thanks for adding the major Vanguard ETFs. It’s incredible they performed so well, any explanation for the 3 outperformers?

James April 28, 2011 at 2:07 pm

Dan, thanks for the info.

Canadian Couch Potato April 28, 2011 at 2:19 pm

@gsp: Thanks for pointing out the error, which I’ve fixed.

I’ve come to expect extremely low tracking errors from Vanguard, though I’m not sure the reason for the outperformance. With international funds, this is often the result of anomalies that disappear when you choose different start and end dates, so I wouldn’t read too much into it. I did a post on this last year:
http://canadiancouchpotato.com/2010/04/23/international-tracking-error-part-1/

gsp April 29, 2011 at 8:09 am

Thanks for the link, enjoyed the 3 part series. The explanation by Vanguard was very informative.

gibor May 1, 2011 at 1:51 am

@CCP, any info on PRF and ZQQ?

Paul T March 22, 2012 at 12:48 pm

I’ve looked at a few of the Claymore products, and as of Aug 2011, here are the numbers that I calculated:

Name Symbol Index Return Fund Return (NAV) Distribution MER Tracking Error

Canadian Fundamental CRQ 3.10% 2.34% $0.14 0.71% 0.05%
US Fundamental (hedged) CLU 5.95% 2.18% $0.09 0.72% 3.05%
US Fundamental (unhedged) CLU.C 5.95% 5.36% $0.07 0.73% -0.14%
International Fundamental CIE 1.90% 1.53% $0.12 0.73% -0.36%
1-5 Yr Laddered Gov’t Bond CLF 1.70% 1.61% $0.45 0.17% -0.08%

From what I can see, they are tracking much closer to their indexes than in the past. The only major outlier is the hedged US Fundamental (CLU). Understandable seeing it is hedged to the CAD. The unhedged version, CLU.C, tracked very closely to its index.

This makes me much more comfortable in picking up additional shares of the Claymore (now iShares Canada) ETFs.

Leave a Comment

{ 2 trackbacks }

Previous post:

Next post: