Take a look at the 2 year chart comparing the performance of the three US REIT ETFs. The three funds follow 3 distinct indices: the Wilshire REIT Index (NYSEARCA:RWR), the Morgan Stanley REIT Index (NYSEARCA:VNQ) and the Dow Jones U.S Real Estate Index (NYSEARCA:IYR).
The underlying holdings of the latter two are almost identical, so it's not surprising to see their performance has been just about the same. They have both been outperformed by the Wilshire index, which has a somewhat different weighting of various REITs.
In terms of income, the IYR yields 3.65%, the VNQ 3.78% and the RWR 3.74% (These are ETFConnect numbers at the time of writing - other sites quote other numbers, and Vanguard says the VNQ's unadjusted effective yield is 3.91%).
The big difference seems to be in the expense ratios. While the VNQ costs 0.12% per annum, the RWR charges 0.25% and the IYR a rather expensive-looking 0.48%.
As we've noted in the chart, the RWR actually outperformed the other two, so arguably it's worth an extra 0.13% over the VNQ cost, which isn't a very big difference anyway. But the iShares IYR fund has almost exactly the same performance as the Vanguard VNQ, almost exactly the same holdings, and four times the expense ratio. That just doesn't make sense.
I wonder how efficient market theorists might explain this: $1.2 billion in assets are sitting in the IYR fund, when the VNQ presents the same performance with 0.36% less expense, or extra yield. At the risk of repeating myself: It just doesn't make sense.