Chart of the day, ETF size edition

By Felix Salmon
January 23, 2012
Devin Riley's excellent post when I was writing about ETFs on Friday. Here's his chart:

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

I’m sorry I missed Devin Riley’s excellent post when I was writing about ETFs on Friday. Here’s his chart:


What you’re seeing here is a y-axis ranking assets under management: the biggest funds are lower down. The x-axis ranks launch date: the earlier the fund launched, the further it is to the left. The correlation could hardly be more obvious. If you want to be a hugely successful ETF, by far the best thing you can do is to launch early.

Here’s Riley:

Launch rank explains up to 81 percent of an ETF’s rank in its segment. More to the point, in 71 percent of all segments, the first-mover had the most assets.

To me, that was surprising.

Given that ETF issuers compete fiercely on expense ratio, index tracking, and marketing materials to win investors, it’s a little disheartening to learn that so much of an ETF’s success is tied to its launch date.

This is surely good for consumers, at least so long as people continue to launch new ETFs. Because when they do so, they’re going to have to compete aggressively on fees, if only because that’s the only way that they’ll ever be able to gain any kind of market share. In turn, the new low-cost competitors will keep fees on the market leaders low and falling.

In theory, eventually, the supply of new ETFs will dry up, and the less successful competitors will drop out of the market. At that point, the market leaders might in theory be able to start raising their fees. But I’m not so worried about that: the best way to increase fee income is to keep your fees low and constant, while increasing your assets under management. Any hint of fees going up is only going to attract new competitors, or incentivize existing competitors to lower their own fees.

So for the time being it’s fine to just buy whatever the biggest ETF is in any given asset class, and sleep easily at night. There’s a small risk that in future you might end up invested in something suboptimal, but you’ll probably hear about it if that happens. Investing isn’t normally this easy, so let’s just celebrate the idea that this strategy seems to work so well, and concentrate entirely on asset allocation rather than spending lots of time second-guessing which of many ETFs to choose from.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

Some of this might be survivorship bias. (It’s quite possible there’s a real effect in there as well, though.)

Posted by dWj | Report as abusive

At the risk of making an embarassing left-right mistake, surely you mean to the right here, not left:

“the earlier the fund launched, the further it is to the left”

Also, given that the sheep tend to pile into funds with a track record, it cannot come as a surprise that people are piling into funds with a track record. I wonder what happens to the correlation if you factor in ETFs that failed?

It is hard to see a genuine first-mover advantage in a fund, unless your investment model is based on being very large.

Posted by Danny_Black | Report as abusive

Read the sentence before that, Danny Black. The biggest funds are lower down, the older funds are farther left. Described accurately, if plotted oddly.

Posted by TFF | Report as abusive

TFF, yeah knew it would be embarassing…. At least it is not me getting left and right confused!

Posted by Danny_Black | Report as abusive

There is an odd mixture of doubtful points in that argument. Firstly, as dWj points out, I suspect a heavy survivorship or, more to the point, selection bias. What about all the ETFs that went out of business over that time? The very early ones would rank high on the Launch scale (i.e. left) but at the bottom of the AUM scale (i.e. top). So there’s a potentially big bunch of observations missing in the upper left end of the chart.

Also, “Given that ETF issuers compete fiercely on expense ratio, index tracking, and marketing materials to win investors”, it’s a little dishartening that he doesn’t take all these variables into account when making such a strong statement. R2 of 81 percent raises more doubts than anything else in general.

Posted by jusha | Report as abusive