A Mixed Bag From Kauffman

October 25, 2011

The Kauffman Foundation’s recent theories on ETFs range from common sense to persuasive to incomprehensible


[This blog previously appeared on our sister site, IndexUniverse.eu.]


Harold Bradley and Robert Litan of the Kauffman Foundation raised some valid concerns about ETFs in their testimony to the US Senate Banking Committee last week. But other criticisms are off the mark.

What are the key points in their report?

ETFs are preventing companies from going public

To me this is the authors’ weakest argument. Bradley and Litan state that over the last 12 years the number of exchange-traded stocks in the US has fallen from 6,200 to 4,300, while the number of ETFs has risen from 95 to over 1,100.

“ETFs are undermining the fundamental role of equities markets in pricing securities to ensure that capital is efficiently allocated to growing businesses,” they conclude.

But this ignores the chilling effect on the US IPO business of the compliance costs introduced by the 2002 Sarbanes Oxley Act, as well as the growing recognition by corporate treasurers worldwide that US banks’ underwriting fees are a rip-off. There are other places in the world to list, in other words.

Twelve years back—1999—also takes us back to the peak of the internet bubble, when hundreds of companies with non-existent earnings and zero prospects did get listed. Remember eToys, webvan and pets.com? A drop in the number of listed stocks from those days and a tightening up of underwriting rules is hardly a loss to the investing public, in my opinion.

0/10 for Kauffman on this one.

ETFs cause high co-movement of index stocks

In their testimony, Bradley and Litan reproduce a chart from Ned Davis Research, which shows that the median two-month correlation between S&P 500 index stocks and the index itself has risen to 0.86, a four-decade high.

To me it’s common sense that the more investors buy and sell equities via index trackers like ETFs and futures, the higher you’d expect internal index correlations to be. And since ETFs have been growing in importance, the greater you’d expect their impact on index stocks’ co-movements to be.

Researchers at JP Morgan came to similar conclusions last year, arguing that high-frequency trading is also playing a role in boosting internal index correlations.

There’s nothing forcing investors, by the way, to purchase those index trackers where such herd effects are most prominent. You don’t have to buy the S&P 500—you could easily choose an index put together in different ways and featuring different stocks. Arguably, avoiding “overowned” indices should be near the top of a list of any investor’s priorities.

Nor are herd effects any less pronounced among active investment managers, all of whom tend to track and try to outperform the same benchmarks.

7/10 for pointing out the existence of such “tight coupling” and asking questions about its significance.

ETFs make certain areas of the market appear more liquid than they really are

I agree with Bradley and Litan’s view that certain ETFs “may have outgrown the market maker’s ability to buy component securities” and that “perceived easy to trade ETFs cannot ever make hard to trade stocks easier to buy or sell”.

That’s relevant not just to the smaller cap stocks they highlight in their testimony, but to other relatively illiquid areas of the market: large sectors of the bond markets and emerging market equities, for example.

And I believe Bradley and Litan are correct to argue that sponsors should be forced to describe how high levels of short interest in an ETF are aligned with the liquidity of ETF constituents. How long, in other words, would it take to cover a short via the ETF creation process, given limited trading volumes in the underlying stocks?

7/10, again, for this section.

ETF sponsors should pay companies for index inclusion

I simply don’t understand this suggestion, made by Robert Litan in an interview on our site. If you list your company’s shares on a stock exchange, you’re allowing third parties to buy and sell them at will.

If you don’t want the volatility that goes with a public listing, you’re under no obligation to allow your shares to be traded on an exchange.

If someone wants to put together a portfolio that includes your company’s shares, whether on the basis of company size, capitalisation, business activity, valuation, share price volatility, that’s up to them.

There are undoubtedly economic incentives that go hand in hand with inclusion in some of the more popular indices, given the weight of passive money following them. If certain index variants become inefficient or too popular from an investment perspective, other indices will emerge to compete with them.

I can see no reason for payments of this type. Am I missing something?

2/10 for this proposal.



Find your next ETF