What about the potential for trading problems in ETFs? Miller points out that many ETFs track highly liquid indices, though Smith posed a slightly different question, focusing on the potential problems that could arise if those heavily short on an ETF tracking less liquid small-cap stocks (for example) all needed to cover their short positions at once.
To me, both men have a point here. Miller argues that ETFs have generally tracked well, in silver, for example, throughout the recent, highly volatile period. But Smith is worried about ETFs’ promise of real-time liquidity, something that may be difficult to achieve in practice when less liquid underlying areas of the market are a fund’s focus. Here, Smith is restating the concerns voiced by regulators a few weeks ago. And, as we wrote in March, even in major equity markets ETFs may not guarantee the continuous trading they promise. Are these funds misrepresenting the real liquidity they can offer investors?
This is a fiendishly difficult question to answer. Sudden investor withdrawals could pose a problem, as Miller argues, in actively managed funds investing in smaller stocks, even if redemptions can only take place daily. In 2008 we saw that even the monthly “gating” mechanism of hedge funds was insufficient to quell panic. Meanwhile, in last year’s flash crash, some ETFs went from trading with tiny spreads and apparently healthy volumes to drying up completely in a millisecond.
Weighing all this up, and if we can try to divorce the question of fundamental liquidity from questions of market structure (which, I believe, caused the flash crash), all we can really say is that an ETF can’t make the underlying asset more liquid than it is by nature. And with ETFs moving into areas of the market that are less liquid and more difficult to track, this is potentially a much bigger concern than it was when such funds were based only on the S&P 500 and the Nasdaq 100.
Finally, who’s right on the risks of ETFs operating with high levels of short interest? Given the open-ended nature of the exchange-traded fund structure there shouldn’t be a problem in high short interest levels per se, though it’s clear that operational risks (the risk of a chain of settlement failures, for example) go up when more shorting is based on a single ETF issue, given the multiple layers of ownership involving all the shorts and the longs.
But when Smith argues that “you may have thought you bought an ETF but what you think you own may actually have been lent to hedge funds” (or words to that effect) he’s surely being disingenuous. After all, this is just what may be happening when you buy an actively managed mutual fund. That fund’s shares may have gone on loan to a short seller too, being replaced (temporarily) by some form of collateral. Come to think of it, that £100 you put in the bank last year may have been lent out by the bank 50 times over as well.
Returning to the question of an ETF with a high short interest, if this led to a price squeeze in the stocks the ETF is tracking (this risk, which Smith points out, is undoubtedly a real one), this would actually benefit owners of the fund (though it would cause losses for those short on the ETF).
Some people may be uncomfortable with short selling as a concept, or with the idea that a single fund or security can be lent out many times over. But these concerns should be addressed within debates over short selling and securities finance (or financial system leverage in general); ETFs are merely offering one illustration of the fact that such practices exist.
There was an excellent discussion last year of the whole topic of ETF short interest, creation and redemption on the FT Alphaville and Kid Dynamite blog sites. Some of the well-reasoned comments on those blogs ultimately convinced me that the scare stories on this particular issue are overblown.
But the current, furious debate over ETF risks must surely be seen in the context of a much broader discussion. And that’s about the future of the fund management business. How are costs to be measured and what underlying risks are being incurred to earn a fund’s returns? What about the asymmetric risk/return profile of the performance fees that many active managers charge? How do regulated funds compare with the myriad other ways that people can access markets—structured products and spread betting, for example? All these are ways of constructing what are, ultimately, vehicles for people’s hard-earned savings. Making a like-for-like comparison between them is a very challenging task. It’s also an ever more urgent one.