Comparing Market Cap ETF vs Dividend ETF – How Much Duplication?

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. 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.


REIT ETF? Or REIT Investment Trusts?

I’ve been pondering buying some REITs in order to increase the real estate allocation of my portfolio which is currently at zero. Several options have popped up:

  1. XRE – iShares REIT ETF which is made up of Canadian REITs.
  2. Buying just the top REIT (RioCan) in XRE.
  3. Buying the top three REITs in XRE.
  4. Buying the top eight REITs in XRE.
  5. American REITs most notably Vanguard REIT ETF (VNQ).

Today I will look at the first three options – feel free to suggest any others.

XRE – iShares Canadian REIT Sector Fund Index is my first choice for REITs. It replicates the S&P/TSX Capped REIT Index and fits very well with my (mostly) passive investment choice. The only problem with this particular ETF is that it has a high MER of 0.55%. Now 0.55% is not totally outrageous but it seems a bit high considering that you can buy the top three REITs in the index and get 50% of the market capitalization of XRE. Another option is to buy the top eight REITs and get 85% of the capitalization of XRE.

The amount I’m planning to own for REITs is about $20,000 which is about 8% of our portfolio. For the purposes of this comparison I’ll assume a 10 year investment time horizon.

  1. Paying 0.55% management fee of XRE per year will result in $110 per year in fees.
  2. Buying just RioCan (REI-UN.TO) will result in a $4.95 purchase fee which works out to $0.50 per year in fees. Problem is that this stock only covers 25% of the REIT index so diversification might not be good enough.
  3. Buying the top three Reits (RioCan, H&R, Can) will result in about $1.50 in fees per year. This will give me about 50% of the capitalization of the index which isn’t bad.
  4. Buying the top eight Reits (RioCan, H&R, Can, Boardwalk, Calloway, Chartwell Seniors Housing, Canada Apartment Properties, Primaris Retail) will result in about $4.00 in fees per year. This will give me about 85% of the capitalization of the index which is pretty darn good.

When I look at the numbers it’s pretty obvious that as long as I don’t care about following the REIT index too closely, buying XRE is a bit of a ripoff. I could probably buy every single REIT in the index and still save money on it’s MER. Buying the top REIT, the top three and the top eight are all pretty much the same cost given that they are all chump change. I think that the extra diversification of buying the top eight is worth the extra couple of bucks per year. With 85% of the index covered, that’s good enough for me. I’ve ignored rebalancing costs but since I’m not too concerned about following the index exactly, I don’t think those will be very much.

I still have to consider US REITs but that will be for another day.


For the amount of REITs I want to buy ($20k) and my trading fees ($4.95) I think that buying the top eight REITs in the REIT index will give me adequate diversification (85% of index) and low costs ($4 per year).

If anyone has any thoughts on my analysis then I’d love to hear them.