I had a reader question the other day where they mentioned buying both XIU (iShares Cdn Large Cap 60 ETF) and XDV (iShares Cdn Dividend Index Fund ETF) for their portfolio. I had responded that although I wasn’t sure, I suspected that might be a lot of duplication in the two funds since XIU has all the biggest public Canadian companies – a lot of which are good dividend stocks and would probably also be in XDV.
Duplicate holdings is a common problem in mutual funds – especially in a market like Canada where there are not a lot of different companies to buy for the larger funds.
I decided to do a bit research and find out if there was as much duplication as I suspected in the two funds. The question I want to answer is if it is worthwhile to own both funds for diversification purposes or will just one do.
Number of companies in common
The first and simplest criteria was how many companies are in both ETFs. This isn’t necessarily all that meaningful since one ETF might have a lot of XYZ company whereas the other might only have a small holding.
XIU 60 has 61 holdings (can’t they count?), XDV dividend has 31 holdings, there are 15 companies that they have in common. This seems like quite a bit since it means that half of the companies in the dividend ETF are also in the XIU ETF.
Amount of market cap in common
What I did here is take the companies that are in both ETFs and compare the percentage holdings and add up the smaller number. For example if CIBC was 9% of the dividend fund and 5% of the XIU then I counted that as 5% in common (by market cap). This totalled up to 31%. This was a smaller number than I expected which means that a good portion of the dividend ETF is not represented in the XIU 60.
Measuring correlation between the ETFs
The next test I did, which should have been the first and only test since it is the only one that has any real meaning is to measure the amount of correlation between the two ETFs. Correlation is a measure of the relationship between the prices of the two ETFs.
A measure of 1 means that they always move in price exactly the same way, a measure of 0 means they are completely uncorrelated and a measure of -1 means they always move in price in exactly the opposite direction. One of the main concepts behind building a portfolio is to try to find different assets that are not correlated with each other.
To accomplish this I needed some historical price data which I managed to find at Yahoo Finance. To figure out the correlation I used the Excel correl function (is there anything Excel can’t do?). XDV dividend has only been around since the end of 2005 so the data is only for a bit less than 4 years. Not being a stats guy I’m not sure if this is a long enough period to be meaningful but it’s all I’ve got. Regardless, the correlation “r” number was 0.72 which implies some benefit for diversification but not a whole lot.
The last thing I looked at was performance. Since the time period is fairly short I’m not looking to see which ETF did better but rather to look at the difference in performance. Ishares.ca website has a handy calculator just for this purpose. I choose the last 3 years since the next category was 5 years which wouldn’t work for XDV dividend.
3 year total return
- XIU Large Cap 60 = -12.98%
- XDV Dividend = -18.19%
From what I’ve read the XDV dividend has a higher ratio of financials than the XIU 60 which is probably one of the reasons for the big performance difference. The XDV dividend has a higher mer (0.5%) than XIU 60 (0.17%) which would account for about 1% of the 5% difference.
I looked at 4 categories to see how different XIU and XDV are:
- Similar companies – half of the XDV dividend companies are in XIU.
- Similar companies by stock market capitalization – 31% of the companies market cap are in both ETFs.
- Correlation – over the last 4 years the correlation is 0.72.
- Performance – the two ETFs were about 5% off in terms of total performance over 3 years.
What does it all mean? Hard to say – there are much better ways to diversify your portfolio – REITs, small cap, foreign holdings would likely all have correlations that are less than 0.72. I’m also not crazy about the higher mer of the dividend ETF.
I think if you want to have most of your equity in Canada then buying partially overlapping ETFs might be the only way to diversify without getting into individual stocks. Personally I like to be diversified over the whole world so for me, the XIU Large Cap 60 by itself is good enough – in my case adding XDV would not increase my diversification enough to make the higher mer worthwhile. XIC (TSX 300) is also a good choice.