Comments on: Chart of the day, ETF size edition

By: jusha

jusha — Tue, 24 Jan 2012 08:58:27 +0000

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.

By: Danny_Black

Danny_Black — Mon, 23 Jan 2012 22:11:24 +0000

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

By: TFF

TFF — Mon, 23 Jan 2012 21:05:38 +0000

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

By: Danny_Black

Danny_Black — Mon, 23 Jan 2012 20:01:00 +0000

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.

By: dWj

dWj — Mon, 23 Jan 2012 16:52:11 +0000

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