In writing my blog, I've put together some very interesting data about the actual size, revenue-wise, of the U.S. ETF industry.

I started out with a discussion of the global target market funds Matt and Murray have been discussing, and the further I delved the more interesting the tangents got, so I'll drop those in right here, and then I'll continue on to my response panning various assertions of Mr. Hougan and Mr. Coleman.

There are 664 ETFs in our database (with loads of fully sortable data points downloadable to an Excel spreadsheet for those of you not acquainted). I split them right down the middle and looked at fund number 337 sorted by expense ratio. That fund would be the iShares UK (EWU) - and it has an expense ratio of...drum roll...57 basis points. You do that same list of 664 and take weight by average ER per dollar, and that number goes down significantly.

And I've had to spend a little time on this project.

But I put all the ETFs we have on a spreadsheet. I multiplied all the assets for each ETF by its respective expense ratio. Then I divided that number by the total assets under management number of all ETFs, and VOILA, I get the average-dollar-weighted expense ratio for the U.S. universe of ETFs.

And what is it?

35.4995 bps

Soon I realized - oh man, I'm on to something here with this spreadsheet. Give me a guess as to what the total revenues are in the ETF business (this is based on the $560 billion in assets in the 603 funds I had left after hacking out all the HOLDRs and funds that were missing ERs for whatever reason).

It's $1.994 BILLION. And here are the top money earners in the ETF business:

These are the top 50 ETFs sorted by revenue per fund. Bear in mind this is just a quick snapshot for the blog, so while I've cleaned the data a bit, I have not done it thoroughly, though this should give you a good idea of the lay of the land.

Ah - if I only had more time, I could do this sort of thing all day and be happy as a pig in...a very dirty pen.

And now on to my regularly scheduled blog attacking certain laughable assertions of the aforementioned Mssrs. Hougan and Coleman.

The Real Significance of the iShares, Vanguard & Northern Trust Moves

The (at last) focus on diversification is great, but I do agree with Murray's point that Matt glosses over: these things probably cost too much.

I don't see the revolution that Hougan does, but I do think this industry is making a bit more sense every day. I am a BIG fan of the concept of having the broadest total market funds available, with all the pieces below them allowing you to tilt at will for whatever reasons.

The funny thing is, though, Matt sort of presents these new funds as a revolution in investing; I don't quite see it that way. But what you don't have is the prototypical asset allocation portfolio of products. For me the ideal is that you've got the lead total market index at the top, and then all of the interlocking indexes below it that match together to make up the whole.

I guess you've got the biggest global brand name with the MSCI ACWI index. SSgA had already launched the ex-US version (CWI) (which might be the most logical first launch given the proliferation of U.S. index ETFs - has $241 million btw). SSgA also has the SPDRs S&P World Ex-US ETF (GWL). Ironically (because it's only got $18 million in assets), this ETF may make the most "sense" from an asset allocation aspect of all the fund families because of SSgA's heavy emphasis on S&P.

So SSgA, what's stopping you - you might as well come out with the all-world with U.S. and join that (new) party, right? As with iShares, both of the SSgA funds already trading are at 35 bps.

The issuer who should have the ACWI as their total market index is actually VANGUARD, who uses MSCI indices across the board. And iShares, who has always modeled themselves as the perfect asset allocation tool set, should actually be using Russell, or maybe DJ Wilshire, where most of their other indices are. And finally, NETS is using, though as yet they have not fleshed out, the whole family of funds below it and have cobbled together all the big international exchange brand names.

We've got the Russell global out there still, and that would seem a natural for iShares, since Vanguard left the Russell party.

And now the point on cost. Incredibly, Matt, the father of the sub 15 bps portfolio, just brushed off Murray's key point, which was that perhaps you could get the same basic exposure of these global funds by mixing and matching a bit. In fact, Mr. Hougan has done it, for 13.65 basis points.

So personally, I'm not going to give up 20 basis points to the glorified concept of having it all in one fund, if I can do it in 5 or 6 at less than 15 bps. That said, all those 35 basis point funds are WAY better than the majority of the garbage being pushed at individual investors out there, and I think you'd be hard pressed to do much better than loading all your equity into one of those and leaving it there for 30 years.

All that from Murray about total market vs. slice and dice, please, no. These are all ultimately battles at the margins. If you want to tilt against how the market has weighting everything and have a bit more value or small or EM or whatever your poison is, go get ‘em. I certainly do it.

So we have multiple total global funds at 35 bps, and now Vanguard comes in at 25 bps, and I'm sure would say that this represents a price that has no eye at all on the competition, but certainly represents Vanguard's cost of running the fund. I don't believe it for a second, particularly if that fund gets a little scale. Now with BGI, Vanguard and SSgA having a rather limited history of reducing the expense ratios on funds, it looks like it's up to Steven Schoenfeld and Northern trust to come in with THEIR total global DJ Wilshire fund (you know, the one everyone else is responding to before it even comes to market) at lower basis points yet.

The fund managers may not greatly relish the "race to the bottom" on fees, but we rather like it as investors.