A weighted-average expense ratio of 0.36% sure eases some of my fears about spiraling costs, Jim. But some of your data seems to run counter to common sense.

As you pointed out, at first glance, it would seem the average ER should be closer to 0.57%. But it only makes sense to weight such a general number by the average expense ratio per dollar - the same phenomenon holds true with traditional mutual funds.

However, there is no clear pattern in the list of top 50 ETFs and revenue. Indeed, with ETFs, flows seem driven more by first-mover advantage in an asset class than by expense ratio and/or quality of the underlying benchmark of the fund.

In the traditional mutual fund world, some 90% of all new sales go to understated funds with lower-than-average expense ratios. Relatively speaking, this is happening with ETFs as well. The No. 4-ranked ETF by revenue, SPY, has roughly a third of the sales of the top dog, EEM. And yet it has 7.5 times the expense ratio of SPY (0.74% vs. 0.10%).

With both of these funds, likely the more tangible factor of expense ratio has taken a back seat to another: first mover advantage. Though IVV had a lower expense ratio than SPY out of the gate, it has taken a long time for it to make any inroads on SPY in assets, and the same has been true of the Vanguard and iShares emerging markets funds - funds that have essentially identical underlying constituents, but very different expense ratios.

The table displayed on Jim's blog points to how many different types of markets and asset classes that can now be targeted using ETFs. It seems investors are making their decisions more by access and exposure than pure pricing.

The comparison of funds also points to liquidity differences between buying a fund for emerging markets and one for the broad U.S. market. Still, how other than first-mover advantage do you explain the fact that something like VWO - again, with essentially the same coverage as EEM to emerging markets and 60% cheaper - has less than a tenth of its sales?

Perhaps there's just so much flooding the market that marketing and name brand familiarity wins out. Vanguard has certainly not counted on advisors in the past to sell its low-cost offerings. Perhaps it still has some work to do getting the attention of advisors? Most of the advisors I've talked to over the years who use ETFs know iShares and consider it and SSgA as the mainstream names. And I suspect that most of these are using ETFs to replace less-volatile traditional core funds they've used to take more control over active management of client portfolios ... hence, they're not as focused on looking under the hood of all of the different options available in each category.

And look at how many of the top 50 sorted by revenue are below that 0.36% weight-adjusted average ER ... 18.

Even though data shows more consumers are getting involved in the ETF market, I'm wondering if these sort of statistics reveal that it's still a trader's marketplace. Considering that a lot of individuals buy ETFs through advisors, and that most are still trying to add value through some sort of active management, I'm afraid ETF costs aren't going to trend down as long as the focus is on short-term profits rather than long-term returns (hence maintaining relatively low costs).

But it certainly is good news that even with all of this activity, the average ER is really well short of half a percentage point. And it still indicates there are enough long-term investors buying ETFs to keep pressure on sponsors from hiking prices for every hot commodity they can find.

Very interesting stuff ... it's kind of a bummer, though, to work with people who post this type of stuff on a weekend and make your head hurt before a new week even starts! I'm learning that, give Jim or Matt a spreadsheet and some spare time, and they can be quite dangerous!